SEARCH

SEARCH BY CITATION

Keywords:

  • fitness;
  • linear mixed-effect model;
  • Malus × domestica;
  • non-linear mixed-effect model;
  • scab;
  • Vf major resistance gene

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

For sustainable management of scab-resistant apple cultivars, it is necessary to understand the role of aggressiveness in the adaptation of Venturia inaequalis populations and particularly the costs to the organism of acquiring additional virulence. The aims of the present study were (i) to identify the quantitative variables that are most important in determining the differences in aggressiveness among groups of V. inaequalis isolates, and (ii) to ascertain whether virulent and avirulent isolates of V. inaequalis differ significantly in aggressiveness. The aggressiveness of eight isolates that differed in their virulence to the major resistance gene Rvi6 was compared on the non-Rvi6 apple cv. Gala. Three components of aggressiveness, namely lesion density, the number of spores per square centimetre of leaf area, and the number of spores per lesion, were evaluated 21 days after inoculation, and the kinetics of lesion density over time were analysed in terms of maximum lesion density, length of latent period and rate of lesion appearance. On the second youngest but fully developed leaf at the time of inoculation, maximum lesion density in the virulent group was 20% lower and the latent period 7% longer, than in the avirulent group. However, the alternative hypothesis, namely that isolates had adapted to quantitative resistance present in cv. Gala depending on their cultivar of origin, could not be rejected. The analysis of the kinetics of lesion density by a non-linear mixed-effect model proved useful in the assessment of aggressiveness.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Selection for quantitative traits related to pathogenicity (the components of aggressiveness) largely determines the composition of populations of fungal plant pathogens (Pariaud et al., 2009a). The dominance of particular races (Brodny et al., 1988; Pariaud et al., 2009b) and the emergence of new populations of a pathogen (Brasier, 1991; Miller et al., 1998) can be explained in terms of the greater aggressiveness of those races or populations. The selection of aggressive isolates can be mediated by the climate (Milus et al., 2009) and by the host (Lehman & Shaner, 1997; Cowger & Mundt, 2002; Andrivon et al., 2007). On the other hand, low aggressiveness can slow down the development of isolates that are virulent against a major resistance gene (Vera Cruz et al., 2000; Huang et al., 2006).

Because disease control in the Venturia inaequalis– apple tree pathosystem is based mainly on spraying fungicides, resistant apple cultivars have been developed to reduce the chemical input. Most of the resistant cultivars carry the major resistance gene Rvi6 (Vf in the old nomenclature), which has resulted in the selection of virulent isolates (Parisi et al., 1993). Approaches are now being developed that aim at a choice of more durable genes for resistance and at sustainable management of these genes, which are considered a fragile genetic resource. Both these objectives require a better understanding of the pathogen’s capacity to adapt through changes in both qualitative and quantitative characters. The deployment of quantitative resistance in a breeding programme for apples (Calenge et al., 2004) requires improved ways of characterizing the aggressiveness in V. inaequalis populations to evaluate the durability of the resistance. Identifying fitness penalties – the cost to the organism of acquiring the required virulence (virulence cost) – can also be particularly useful in deploying the available resistance gene more strategically (Sapoukhina et al., 2009). The hypothesis that isolates that are virulent to the major resistance gene Rvi6 may be less fit than avirulent isolates on non-Rvi6 cultivars (Guérin et al., 2007) is yet to be tested experimentally.

Aggressiveness can be evaluated in terms of disease severity or can be dissected into the components of the infection cycle, namely the length of the latent period, infection efficiency, lesion growth, and the rate of sporulation. However, measuring those components is both technically difficult and system-specific, i.e. the techniques or the results of such measurements cannot be transposed directly from one pathosystem to another, because of differences in biological features of the systems. Several components of aggressiveness of different isolates of V. inaequalis to a susceptible apple cultivar are known to vary: infection efficiency, latent period, lesion size, and the number of spores per square centimetre of lesion (Lalancette et al., 1987; Parisi et al., 2004). These components were shown to be significantly different between benomyl-sensitive and benomyl-resistant subpopulations of V. inaequalis (Lalancette et al., 1987). Population studies require testing a large number of isolates (Pariaud et al., 2009b), which limits the number of aggressiveness components that can be measured at one time. Better characterization of the components and their relationships is therefore necessary.

The aims of the present study were (i) to identify the quantitative variables that are most important in determining the differences in aggressiveness of different isolates of V. inaequalis, and (ii) to test, on a limited number of Venturia inaequalis isolates, the hypothesis that isolates that are virulent to the major resistance gene Rvi6 are less fit than avirulent isolates on a non-Rvi6 apple cultivar. Because no apple cultivar can be considered as universally susceptible (Barbara et al., 2008), the alternative hypothesis, namely that the isolates may be adapted to other potential resistance genes present in the non-Rvi6 cultivar depending on the host from which they originated, was also analysed.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Plant material

Gala (X4712 from the INRA depository, France) was chosen as a non-Rvi6 cultivar. It carries at least two QTL (quantitative trait loci) of resistance to scab (Soufflet-Freslon et al., 2008), but is often used as the susceptible reference standard in pathogenicity studies (Bus et al., 2009). Gala was grafted on rootstock MM106 (Pirard, France) and grown in 1·5 L pots filled with a potting mix comprising 60% leaf mould, 30% peat and 10% disinfected garden soil. The pots were placed in a glasshouse (15–25°C). Only actively growing plants with uniform growth were chosen and transferred to a climate chamber. The youngest fully expanded leaf was labelled F0, the second youngest leaf below F0 was labelled F1, and so on (the higher the number, the older the leaf). At the time of inoculation, each plant had 8–13 leaves.

Experiments

Two experiments were performed: Experiment 1 in April 2005, 6 weeks after grafting, and Experiment 2 in September 2005, 4 weeks after pruning the same set of plants. Each plant was pruned back to retain 2–3 buds. For each isolate, nine plants were inoculated in Experiment 1 and eight in Experiment 2.

Isolates of Venturia inaequalis

Eight monoconidial isolates collected in 1998 or 1999 from six apple cultivars and from four European countries (European project Dare ‘Durable Apple Resistance in Europe’ 1998–2002, Table 1) were chosen from 39 isolates that had been tested on a set of eight differential hosts (Parisi et al., 2004), that, among others, carried three major resistance genes, Rvi6, Rvi1 and Rvi13 (Vf, Vg and Vd in the old nomenclature, respectively). This choice was based only on whether the isolates were avirulent or virulent to the major resistance gene Rvi6, although they may have differed in their virulence to other major genes. Four isolates were virulent to Rvi6 (Group V) and four were avirulent to Rvi6 (Group A). Three of the eight isolates (EU-B04, EU-D42 and EU-NL24) have been used for testing Rvi6 cultivars every 2 years since the first test of pathogenicity; the same virulence profile has been observed each time (F. Laurens, INRA, Angers, France, personal communication). No isolate originating from cv. Gala was chosen in order to avoid any bias due to direct adaptation of such isolates to cv. Gala compared to that of isolates originating from other cultivars. Because the isolates originated from cultivars that differed in the degree to which they were related to cv. Gala, their genetic proximity to cv. Gala was calculated (Table 1) based on their known genealogies using Wright’s coefficient of relationship (Wright, 1922). The isolates were stored on malt agar and maintained at 4°C. Conidial suspensions were obtained from cultures grown on cellophane overlaid on malt agar in Petri dishes after 10–12 days of incubation at 17°C with 16 h of light and inoculated onto seedlings of cv. Golden Delicious × cv. Granny Smith to obtain sufficient inoculum for the tests of aggressiveness. This multiplication was performed under standard conditions (Parisi et al., 1993). Scabbed leaves were harvested 10–11 days after inoculation, dried and stored at –20°C.

Table 1.   Description of the isolates of Venturia inaequalis used in this study
IsolateaSampling dateCountry of originCultivar of originRelationship coefficientbAvirulence/virulence to:c
Rvi6 (=Vf)Rvi1 (=Vg)Rvi13 (=Vd)
  1. A = avirulent and V = virulent.

  2. aIsolates from the European Dare project (1998–2002).

  3. bGenetic proximity between the host of origin and the test cv. Gala computed using Wright’s coefficient of relationship.

  4. cData from Calenge et al., 2004; Parisi et al., 2004; Bus et al., 2005; L. Parisi, INRA, Gotheron, France, personal communication.

EU-B041998BelgiumGolden Delicious1/2AVA
EU-B271999BelgiumPrésident Roulin0AVA
EU-GR031998GreeceStarking Delicious1/4AVA
EU-NL191998The NetherlandsGolden Delicious1/2AVA
EU-D221999GermanyFlorina1/8VVA
EU-D421999GermanyPrima1/8VVA
EU-NL051998The NetherlandsMalus floribunda 8210VAV
EU-NL241998The NetherlandsPrima1/8VVA

Inoculations

Spore suspensions were prepared from the scabbed leaves and adjusted to a concentration of 110 000–130 000 spores mL−1 (Coulter counter, Beckman, 6·2 μm < mean diameter < 12·8 μm). Three droplets of the suspension were deposited on malt agar to measure the germination rate, which varied from 61% to 92% depending on the isolate and the experiment. The concentration of viable inoculum (Ct) was calculated as the spore concentration of the inoculum suspension multiplied by its germination percentage. For each isolate, 20 mL of suspension was applied with a chromatography sprayer (mainly on the upper side of the leaves) to eight or nine plants of cv. Gala enclosed in a plastic chamber to avoid cross-contamination of isolates during inoculation. For 44 h after inoculation, the leaves were kept wet with an atomizer and the plants kept in darkness at 18°C. When the leaves were dry, the plants were placed at randomly assigned positions in the climate chamber, and incubated under 16 h of light at 18°C and relative humidity fluctuating between 70% and 90%. During incubation the leaves remained dry, which prevented secondary infections.

Measurements of aggressiveness

Aggressiveness was evaluated on leaves F0 and F1, which were the most susceptible to scab, and developed full and well-defined lesions within 3 weeks.

Number of lesions

Seven days after inoculation, each plant was inspected to confirm the absence of symptoms. From the 8th to the 21st day, the number of new lesions on each leaf was recorded daily until the 11th or 12th day and every 2 or 3 days thereafter. Each lesion was marked with a permanent marker to avoid recounting.

Sporulation rate

Twenty-one days after inoculation, each leaf was shaken in distilled water with 0·01% Tween 20 in order to collect the spores in a suspension. This was done for three plants per isolate in Experiment 1 and for eight plants per isolate in Experiment 2. After diluting the suspension a 100-fold, the number of spores per millilitre of the suspension was calculated after counting the spores with a Coulter counter.

Leaf area

At the end of the experiment, the length (l) and width (w) of leaves were measured for estimating leaf area (La) based on the following equation established for cv. Gala under controlled conditions (F. Didelot and L. Parisi, INRA, Angers, France, personal communication):

  • image(1)

where 0·6997 and 0·3886 are the coefficients of linear regression (r² = 0·9837, < 0·001).

Plant growth

As the stage of plant growth is known to affect the development of scab through changes in ontogenic resistance (Schwabe, 1979), the number of inoculated leaves from the youngest (F0) to the oldest (Li) and the number of new leaves that had appeared after the inoculation (Ln) were counted for each plant 21 days after inoculation.

Data analysis

Twenty-one days after inoculation, three components of aggressiveness were evaluated for leaves F0 and F1, namely lesion density (ld21), the number of spores per square centimetre of leaf area (sps21), and the number of spores per lesion (spl21). Lesion density (ld21) was expressed as the total number of lesions per square centimetre of leaf area. As the number of conidia originally deposited on the leaves at the time of inoculation was not known precisely, the efficiency of infection could not be ascertained, and lesion density was considered an acceptable substitute.

For each of these variables, denoted as Y hereafter, the following linear mixed-effect model (Lme) was used with the random factor ‘isolate’ nested in ‘group’ V or A (Pinheiro & Bates, 2000):

  • image(2)

where β0 is the general mean; αe is the effect of the ‘experiment’ factor, for e = 1, 2; γg is the effect of the ‘group’ factor, for g = 1, 2 (avirulent and virulent, respectively); β1,β2,β3,β4 are parameters associated to co-variables Ct, La, Li, Ln; bi is the random effect associated with the ‘isolate’ factor, supposed to be normally distributed bi ∼ N (0, inline image); and εegij is the error associated with the jth observation.

Two models were tested for the error term εegij. The first one assumed that εegij ∼ N (0, inline image), with the variance inline image depending only on the ‘experiment’ and ‘isolate’ factors. The second model assumed that εegijN (0, inline image), with the variance inline image depending only on the ‘experiment’ and ‘group’ factors. The better of the two models was selected using AIC (Akaike’s Information Criterion), and each co-variable was included in this model based on an ascendant selection using AIC (Pinheiro & Bates, 2000). Parameters were estimated by the maximum likelihood algorithm with constraints ∑α= 0 and γ1 = 0, for the parameters concerning the ‘experiment’ and the level ‘avirulent’ of the ‘group’, respectively.

For each date of observations, the lesion density (ldt) was calculated as the cumulative number of lesions at time t per unit of the final leaf area (La). The kinetics of lesion density are the repeated measurements for each individual plant and can be represented by a logistic function, commonly used to describe the evolution of plant disease over time (Zadoks & Schein, 1979; Holb et al., 2005). The logistic function f(β, t) is given by the following equation:

  • image(3)

where β is the vector (k, lat, r) of three parameters: k is the asymptote, representing the maximum lesion density; lat is the latent period, the time t for 50% of the lesions to appear; and r is the relative rate of lesion appearance.

A non-linear mixed-effect model (Nlme) with link function given by equation (3) was used (Davidian & Giltinan, 2003; Holb et al., 2005). The three parameters (k, lat, r) were linearly modelled depending on co-variables Ct, La, Li, Ln, fixed factors ‘experiment’ and ‘group’, and the random effect of ‘individual’. Furthermore for the parameters k and lat, the random effect of ‘isolates’ nested in ‘group’ factor was included in the model. The variance inline image of the error term depended on the ‘experiment’ and ‘isolate’ factors. The covariance matrix of the residual error was defined by a first-order autoregressive model, which took into account the dependence of residuals (Pinheiro & Bates, 2000). The best model was selected using the same criteria as those for the model (2) described above.

Correlations among aggressiveness components were tested by Spearman tests (Sprent & Smeeton, 2007) based on the data for each plant. For each plant, maximum lesion density and length of the latent period were fitted on the logistic curve with the non-linear least squares method (Bates & Watts, 1988). The Spearman tests were based on the assumption that correlations were not affected by the isolate and data were pooled over all isolates. Analyses were done separately for leaves F0 and F1 and for Experiments 1 and 2.

All statistical analyses were performed using r version 2.8.0 (R Development Core team, 2008) with the ‘nlme’ package (Pinheiro et al., 2008).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Co-variables

Most variables were significantly affected by the co-variables linked to the growth of the plant (Table 2). The leaf area (La) had a significant effect on ld21, sps21, k and lat; the number of inoculated leaves (Li) had a significant effect on ld21 and lat; and the number of new leaves (Ln) had a significant effect on ld21. There was no significant effect of the concentration of the viable inoculum (Ct) on any of the variables.

Table 2.   Fixed effects of Lme modelling of three components of aggressiveness 21 days after inoculation of apple cv. Gala with eight isolates of Venturia inaequalis pooled in two groups (avirulent or virulent to the major resistance gene Rvi6), and fixed effects of Nlme modelling of the kinetics of lesion density (leaf F1)
VariableModelFactor or co-variableaNobValueStandard Errord.f.P-valuec
  1. aCo-variables: La = leaf area, Li = number of inoculated leaves, Ln = number of new leaves 21 days after inoculation.

  2. bNumber of individuals.

  3. cIn bold, P-value < 0·05 for Group V (virulent to Rvi6).

  4. dModelling of variance of the error term according to ‘isolate’ and ‘experiment’ factors.

  5. eIntercept represents Group A (avirulent to Rvi6).

  6. fModelling of variance of the error term according to ‘group’ and ‘experiment’ factors.

  7. gNull data were excluded from the data set.

ld21 (no of lesions per cm2 of leaf area)Lmed(Intercept)e1320·850·171210·0000
La −0·020·001210·0000
Li 0·030·011210·0002
Ln −0·030·011210·0174
sps21 (no of spores per cm2 of leaf area)Lmef(Intercept)e 7530477453139650·0000
Exp1 3391615368650·0309
La −60821626650·0004
spl21 (no of spores per lesion)Lmef(Intercept)e 7526867421698650·0000
GroupV −13289922936650·0012
Exp1 9148121698650·0001
GroupV:Exp1 −9614522936650·0001
Kinetics of lesion densityNlmedk.(Intercept)e661g1·530·185300·0000
k.GroupV −0·300·115300·0048
k.La 0·020·015300·0002
lat.(Intercept)e 13·820·605300·0000
lat.GroupV 0·930·305300·0022
lat.La 0·050·025300·0033
lat.Li −0·170·045300·0000
r.(Intercept)e 1·130·055300·0000

Comparison of the two experiments

The number of leaves at the time of inoculation was higher in Experiment 1 than in Experiment 2 (12·5 and 8·9, respectively), although the plants had been growing more actively during Experiment 2 (4·5 new leaves and 2·2 new leaves, respectively, on day 21). The mean leaf area (La) was similar in the two experiments. Twenty-one days after inoculation, sporulation (sps21 and spl21) was significantly lower in Experiment 2 than in Experiment 1, but the effect of ‘experiment’ on ld21, k, lat and r was not significant (Table 2).

Comparison of the two groups of isolates

The eight isolates were divided into two groups, one virulent to Rvi6 and the other avirulent to it. For leaf F0, there was no significant effect of ‘group’ for any of the variables (data not shown). For leaf F1, there was a significant effect of ‘group’ on k and lat (Table 2). In group V, the maximum lesion density k was 20% less, and the latent period 7% longer, than group A; the differences are described by two different estimated logistic curves (Fig. 1).

image

Figure 1.  Kinetics of measured lesion density on leaf F1 after inoculation of apple cv. Gala with eight isolates of Venturia inaequalis, pooled in two groups according to their avirulence (x) or virulence (o) to the major resistance gene Rvi6. Solid line: logistic curve estimated by Nlme modelling for the avirulent group; dashed line: logistic curve estimated by Nlme modelling for the virulent group. Experiments 1 and 2 were pooled.

Download figure to PowerPoint

As there was a significant effect of ‘experiment’ on sps21 and spl21 (Table 2), variables evaluated 21 days after inoculation were analysed separately for each experiment (Table 3). There was no significant effect of ‘group’ on ld21 in either of the experiments. However, there was a significant effect of ‘group’ on sps21 and spl21 in Experiment 1; in Group V, the number of spores per unit leaf area was 30% less, and that per lesion was 40% less, compared to Group A. There was a significant effect of interaction between the isolates and the experiments (data not shown), largely due to isolate EU-NL19, which was relatively more aggressive in Experiment 2 than in Experiment 1. In both experiments, isolate EU-NL05 was the least aggressive isolate.

Table 3.   Means and standard deviations of three components of aggressiveness 21 days after inoculating apple cv. Gala with eight isolates of Venturia inaequalis
ExperimentIsolateGroupld21sdaTukey testbsps21sdTukey testspl21sdTukey test
  1. A: avirulent to Rvi6; V: virulent to Rvi6.

  2. aStandard deviation.

  3. bDifferent letters indicate significant differences between isolates or between groups of isolates (< 0·05).

1EU-GR03A1·130·26a344525112886ab298730117721a
EU-B04A0·980·62a295921132220ab370837127353a
EU-B27A0·970·37a49871080321a452971203008a
EU-D22V0·890·37a221334217506ab22330271715a
EU-NL19A0·840·63a169787260007ab439083302873a
EU-D42V0·650·37ab12821311835b12823718623a
EU-NL24V0·570·29ab7732665095b13979266645a
EU-NL05V0·210·16b2500414313b255081288322a
Mean of the groupA0·980·47a327 236146 359a390 405187 739a
V0·580·30a112 96977 187b186 603111 326b
2EU-NL19A1·730·52a30411072341a18584949915a
EU-D22V0·950·53b14881196936b13242535245a
EU-GR03A0·900·59b11273274757b170731126458a
EU-B04A0·580·41cd8099635723b14932196243a
EU-NL24V0·570·39cd7596048997b16671194654a
EU-B27A0·440·27cd6569343226b173859142369a
EU-D42V0·350·34cd5099139274b210369237152a
EU-NL05V0·030·05d235729314b24766713671a
Mean of the groupA0·910·45a140 88356 512a169 940103 746a
V0·480·33a74 83348 630a189 29395 180a

Correlations between variables

The results for F0 and F1 were similar, and only those for F1 are presented here (Table 4). As expected, lesion density 21 days after inoculation (ld21) was highly correlated with maximum lesion density (k). These two variables were correlated with sps21 but not with spl21. Latent period was moderately correlated with ld21 and k.

Table 4.   Spearman coefficients of correlation (rho) among components of aggressiveness for eight isolates of Venturia inaequalis inoculated on apple cv. Gala (leaf F1)
VariableExperimentksps21spl21lat
rhoP-valuerhoP-valuerhoP-valuerhoP-valuea
  1. aBold characters indicate a significant correlation (< 0·05).

ld21 (lesion density)10·820·00000·660·00200·280·2251−0·430·0007
20·980·00000·850·00000·050·7755−0·480·0022
k (maximum lesion density)1  0·430·05930·240·29780·120·3389
2  0·810·00000·020·9097−0·430·0067
sps21 (no of spores per cm2 of leaf area)1    0·860·0000−0·480·0351
2    0·520·0009−0·640·0000
spl21 (no of spores per lesion)1      −0·330·1543
2      −0·420·0083

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Aggressiveness of V. inaequalis isolates was tested under controlled conditions on the non-Rvi6 apple Gala and analysed in relation to their virulence or avirulence to the major resistance gene Rvi6 or in relation to the cultivar from which the isolates had originated. On leaf F0, no significant differences were observed, probably because this leaf was very susceptible to scab, as shown by rapid scab development and a latent period 1·5 days shorter than that on F1 (data not shown). On leaf F1, the group comprising virulent isolates (group V) and that comprising avirulent isolates (group A) differed significantly in two parameters of the kinetics of lesion density: maximum lesion density and length of the latent period; in the virulent group maximum lesion density was 20% less and the latent period 7% longer. However, the hypothesis that isolates may be differentially adapted to quantitative resistance genes present in cv. Gala depending on the host from which isolates originated, could not be rejected.

There was a good fit between the three-parameter logistic function and the kinetics of lesion density. The non-linear mixed-effects model of kinetics gave information on two components of aggressiveness: lesion density and length of the latent period, which are known to vary in populations of V. inaequalis (Lalancette et al., 1987). These two parameters are usually analysed separately. The non-linear mixed-effects model is noteworthy because it compares the kinetics by taking into account both the parameters simultaneously. The model proved more sensitive than Lme for the detection of significant differences between the two groups of isolates. For instance, differences between the two groups in lesion density were not significant when analysed by Lme, but significant when analysed by Nlme.

Maximum lesion density and length of the latent period, as analysed by Nlme, were not dependent on the experimental effect, whereas disease severity and the rate of sporulation in apple scab are reported to be strongly influenced by the experimental factor (Lalancette et al., 1987). Furthermore, with regard to maximum lesion density and length of the latent period, differences between the two groups were similar for Experiment 1 and Experiment 2, although conditions at the time of the second experiment were less favourable to the disease. Greater differences in aggressiveness among isolates may be expected to express more clearly under less favourable conditions (Pariaud et al., 2009a); however, the present study did not bear out that expectation.

A good correlation was found between lesion density and the number of spores produced per unit of leaf area 21 days after inoculation, a result that is contrary to the findings from other pathosystems, such as those involving rust fungi (Kardin & Groth, 1989; Riméet al., 2005) or powdery mildews (Rouse et al., 1984), in which sporulation reaches a plateau at higher lesion densities. The good correlation observed in the present study may be due to the fact that the number of spores produced per lesion was not affected by lesion density, at least for the range of lesion density recorded in the present study. Because new lesions continued to appear over a long time, both old and new lesions were present on the plant at the same time, and, although the rate of sporulation of a lesion may decline with age (Sache, 1997), the decline may have been compensated for by the younger lesions. Analysing the kinetics of sporulation in relation to lesion density with a non-destructive method of measuring total conidial production across the entire disease span (Sache, 1997; Suassuna et al., 2004; Riméet al., 2005) may make it possible to test this explanation. A wide range of lesion density should be analysed for single isolates, to allow for the possibility that the effect may vary with the genotype of the isolate, as has been shown for bean rust (Kardin & Groth, 1989). If it is confirmed that there is a good correlation between lesion density and sporulation, analysis of lesion density could replace analysis of sporulation level, which is notoriously difficult to measure, but is important for understanding the epidemiology of apple scab.

The co-variables related to plant growth improved the estimation of aggressiveness. First, leaf area had a significant effect on all the components of aggressiveness, probably because younger – and therefore smaller – leaves are more susceptible to scab or because scab reduces leaf growth. Whatever the reason, taking leaf area into account helps in estimating the aggressiveness of the pathogen in the apple-scab pathosystem more reliably. Secondly, the number of inoculated leaves and the number of new leaves produced after inoculation were observed to affect aggressiveness. This result fits in with the fact that apple scab is affected by stage of growth of the plant, which affects its ontogenic resistance (Schwabe, 1979). The concentration of viable spores, which ranged from 60 000 to 110 000 viable spores per millilitre of inoculum suspensions in the present study, did not affect aggressiveness – an observation in line with the results of earlier studies, which showed no effect despite a wider range, namely 80 000–300 000 spores mL−1 (Hartman et al., 1999; Lateur et al., 2000).

Fungal populations that are virulent against a major resistance gene may pay a fitness penalty, as for instance in Leptosphaeria maculans on oilseed rape for the resistance gene Lm4 (Huang et al., 2006), but this is not a general rule (Bahri et al., 2009). The present study provides the first results for the scab-apple pathosystem concerning the relative aggressiveness of isolates virulent or avirulent to Rvi6; the non-linear mixed-effects model of the kinetics of lesion density indicated a significant difference between the two groups of isolates pooled according to their virulence/avirulence to Rvi6, which suggests that the breakdown of Rvi6 resistance resulted in reduced fitness of the virulent population. However, the least aggressive isolate also carried a virulence factor to Rvi13. More isolates virulent to Rvi13 need to be tested to determine whether this reduced aggressiveness is due to the breakdown of Rvi6, or that of Rvi13, or of both.

As no apple cultivar can be considered universally susceptible, an alternative hypothesis would be differential adaptation of the pathogen to the QTL present in the cultivar chosen for testing. Adaptation to quantitative resistance has been shown, for instance, in Phythophthora infestans on potato (Andrivon et al., 2007) and Mycosphaerella graminicola on wheat (Cowger & Mundt, 2002), and differential adaptation to cultivars considered as susceptible has also been shown for apple scab (Sierotzki et al., 1994; Barbara et al., 2008). Isolates of the two groups (A and V) were sampled from different cultivars, without sampling any isolate from cv. Gala itself, but generally Rvi6 cultivars were more distant from cv. Gala than were non-Rvi6 cultivars (average coefficient of relationship of 0·09 and 0·31, respectively). As a consequence, the hypothesis that the isolates had adapted to quantitative resistance present in cv. Gala depending on their cultivar of origin could not be rejected. Furthermore, the least aggressive isolate came from M. floribunda and not from M. ×domestica, therefore the hypothesis of a differential adaptation related to the host species (M. × domestica or M. floribunda) should also be considered. Testing a larger number of non-Rvi6 cultivars and testing more isolates from each host cultivar or species would strengthen the interpretation of these data.

In conclusion, the present study, based on a restricted set of isolates from different cultivars and different countries tested on cv. Gala, shows significant differences in aggressiveness between two groups of isolates, a difference which may be explained either by reduced aggressiveness of virulent populations (directly in the case of a cost of acquiring virulence or indirectly in the case of a chance mutation for virulence in an isolate of low aggressiveness) or by adaptation to QTL present in the non-Rvi6 cultivar. Data on a larger number of isolates are needed to compensate for the high diversity observed in V. inaequalis populations (Gladieux et al., 2008) and data on a larger set of cultivars without major genes of resistance are needed to take into account the fact that no cultivar can be considered universally susceptible (Barbara et al., 2008). The present study showed the usefulness of a non-linear mixed-effects model of the kinetics of lesion density for comparison of aggressiveness in different groups of V. inaequalis isolates. This model will be used in future studies of the apple-scab pathosystem extending to more isolates representing different populations.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank partners of the European project Dare (1998–2002) for sampling the isolates of Venturia inaequalis, Florian Blanchet for technical help in the experiments, and Christian Lannou and Natalia Sapoukhina for helpful comments on earlier versions of this paper.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Andrivon D, Pilet F, Montarry J et al. , 2007. Adaptation of Phytophthora infestans to partial resistance in potato: evidence from French and Moroccan populations. Phytopathology 97, 33843.
  • Bahri B, Kaltz O, Leconte M, De Vallavieille-Pope C, Enjalbert J, 2009. Tracking costs of virulence in natural populations of the wheat pathogen, Puccinia striiformis f.sp. tritici. BMC Evolutionary Biology 9. Http://www.biomedcentral.com/1471-2148/9/26 .
  • Barbara DJ, Roberts AL, Xu XM, 2008. Virulence characteristics of apple scab (Venturia inaequalis) isolates from monoculture and mixed orchards. Plant Pathology 57, 55261.
  • Bates DM, Watts DG, 1988. Nonlinear Regression Analysis and its Applications. New York, USA: John Wiley & Sons.
  • Brasier CM, 1991. Ophiostoma novo-ulmi sp. nov, causative agent of current Dutch elm disease pandemics. Mycopathologia 115, 15161.
  • Brodny U, Wahl I, Rotem J, 1988. Factors conditioning dominance of race 276 of Puccinia coronata avenae on Avena sterilis populations in Israel. Phytopathology 78, 1359.
  • Bus VGM, Laurens FND, Van De Weg WE et al. , 2005. The Vh8 locus of a new gene-for-gene interaction between Venturia inaequalis and the wild apple Malus sieversii is closely linked to the Vh2 locus in Malus pumila R12740-7A. New Phytologist 166, 103549.
  • Bus V, Rikkerink E, Aldwinckle HS et al. , 2009. A proposal for the nomenclature of Venturia inaequalis races. Acta Horticulturae 814, 73946.
  • Calenge F, Faure A, Goerre M et al. , 2004. Quantitative trait loci (QTL) analysis reveals both broad-spectrum and isolate-specific QTL for scab resistance in an apple progeny challenged with eight isolates of Venturia inaequalis. Phytopathology 94, 3709.
  • Cowger C, Mundt CC, 2002. Aggressiveness of Mycosphaerella graminicola isolates from susceptible and partially resistant wheat cultivars. Phytopathology 92, 62430.
  • Davidian M, Giltinan DM, 2003. Nonlinear models for repeated measurement data: an overview and update. Journal of Agricultural, Biological and Environmental Statistics 8, 387419.
  • Gladieux P, Zhang XG, Afoufa-Bastien D, Valdebenito Sanhueza RM, Sbaghi M, Le Cam B, 2008. On the origin and spread of the scab disease of apple: out of central Asia. PLoS ONE 3, e1455. doi:http://dx.doi.org/10.1371/journal.pone.0001455.
  • Guérin F, Gladieux P, Le Cam B, 2007. Origin and colonization history of newly virulent strains of the phytopathogenic fungus Venturia inaequalis. Fungal Genetics and Biology 44, 28492.
  • Hartman JR, Parisi L, Bautrais P, 1999. Effect of leaf wetness duration, temperature, and conidial inoculum dose on apple scab infections. Plant Disease 83, 5314.
  • Holb IJ, Heijne B, Withagen JCM, Gall JM, Jeger MJ, 2005. Analysis of summer epidemic progress of apple scab at different apple production systems in the Netherlands and Hungary. Phytopathology 95, 100120.
  • Huang YJ, Li ZQ, Evans N, Rouxel T, Fitt BDL, Balesdent MH, 2006. Fitness cost associated with loss of the AvrLm4 avirulence function in Leptosphaeria maculans (phoma stem canker of oilseed rape). European Journal of Plant Pathology 114, 7789.
  • Kardin MK, Groth JV, 1989. Density-dependent fitness interactions in the bean rust fungus. Phytopathology 79, 40912.
  • Lalancette N, Hickey KD, Cole H, 1987. Parasitic fitness and intrastrain diversity of benomyl-sensitive and benomyl-resistant subpopulations of Venturia inaequalis. Phytopathology 77, 16006.
  • Lateur M, Lefrancq B, Wagemans C, 2000. Influence of scab inoculum concentration in an apple breeding program focused on quantitative resistance. Acta Horticulturae 538, 24955.
  • Lehman JS, Shaner G, 1997. Selection of populations of Puccinia recondita f.sp. tritici for shortened latent period on a partially resistant wheat cultivar. Phytopathology 87, 1706.
  • Miller JS, Johnson DA, Hamm PB, 1998. Aggressiveness of isolates of Phytophthora infestans from the Columbia Basin of Washington and Oregon. Phytopathology 88, 1907.
  • Milus EA, Kristensen K, Hovmøller MS, 2009. Evidence for increased aggressiveness in a recent widespread strain of Puccinia striiformis f.sp. tritici causing stripe rust of wheat. Phytopathology 99, 8994.
  • Pariaud B, Ravigné V, Halkett F, Goyeau H, Carlier J, Lannou C, 2009a. Aggressiveness and its role in the adaptation of plant pathogens. Plant Pathology 58, 40924.
  • Pariaud B, Robert C, Goyeau H, Lannou C, 2009b. Aggressiveness components and adaptation to a host cultivar in wheat leaf rust. Phytopathology 99, 86978.
  • Parisi L, Lespinasse Y, Guillaumes J, Krüger J, 1993. A new race of Venturia inaequalis virulent to apples with resistance due to the Vf gene. Phytopathology 83, 5337.
  • Parisi L, Fouillet V, Schouten HJ et al. , 2004. Variability of the pathogenicity of Venturia inaequalis in Europe. Acta Horticulturae 663, 10713.
  • Pinheiro JC, Bates DM, 2000. Mixed-Effects Models in S and S-PLUS. New York, USA: Springer-Verlag.
  • Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core team, 2008. Nlme: linear and nonlinear mixed effects models. R Package Version 3.1-90. Http://www.R-project.org.
  • R Development Core team, 2008. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Http://www.R-project.org .
  • Rimé D, Robert C, Goyeau H, Lannou C, 2005. Effect of host genotype on leaf rust (Puccinia triticina) lesion development and urediniospore production in wheat seedlings. Plant Pathology 54, 28798.
  • Rouse DI, MacKenzie DR, Nelson RR, 1984. Density dependent sporulation of Erysiphe graminis f.sp. tritici. Phytopathology 74, 117680.
  • Sache I, 1997. Effect of density and age of lesions on sporulation capacity and infection efficiency in wheat leaf rust (Puccinia recondita f.sp. tritici). Plant Pathology 46, 5819.
  • Sapoukhina N, Durel CE, Le Cam B, 2009. Spatial deployment of gene-for-gene resistance governs evolution and spread of pathogen populations. Theoretical Ecology 2, 22938.
  • Schwabe WFS, 1979. Changes in scab susceptibility of apple leaves as influenced by age. Phytophylactica 11, 536.
  • Sierotzki H, Eggenschwiler M, Boillat O, McDermott JM, Gessler C, 1994. Detection of variation in virulence toward susceptible apple cultivars in natural populations of Venturia inaequalis. Phytopathology 84, 10059.
  • Soufflet-Freslon V, Gianfranceschi L, Patocchi A, Durel CE, 2008. Inheritance studies of apple scab resistance and identification of Rvi14, a new major gene that acts together with other broad-spectrum QTL. Genome 51, 65767.
  • Sprent P, Smeeton NC, 2007. Applied Non-parametric Statistical Methods. Berlin, Germany: Chapman & Hall.
  • Suassuna ND, Maffia LA, Mizubuti ESG, 2004. Aggressiveness and host specificity of Brazilian isolates of Phytophthora infestans. Plant Pathology 53, 40513.
  • Vera Cruz CM, Bai J, Ona I et al. , 2000. Predicting durability of a disease resistance gene based on an assessment of the fitness loss and epidemiological consequences of avirulence gene mutation. Proceedings of the National Academy of Sciences, USA 97, 135005.
  • Wright S, 1922. Coefficients of inbreeding and relationship. The American Naturalist 56, 3308.
  • Zadoks JC, Schein RD, 1979. Epidemiology and Plant Disease Management. New York, USA: Oxford University Press.