Seasonal changes in susceptibility of Quercus suber to Botryosphaeria stevensii and Phytophthora cinnamomi

Authors

  • J. Luque,

    1. Departament Protecció Vegetal, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Centre de Cabrils, Ctra. de Cabrils s.n., E-08348 Cabrils, Barcelona, Spain
    Search for more papers by this author
  • J. Parladé,

    1. Departament Protecció Vegetal, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Centre de Cabrils, Ctra. de Cabrils s.n., E-08348 Cabrils, Barcelona, Spain
    Search for more papers by this author
  • J. Pera

    Corresponding author
    1. Departament Protecció Vegetal, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Centre de Cabrils, Ctra. de Cabrils s.n., E-08348 Cabrils, Barcelona, Spain
    Search for more papers by this author

*To whom correspondence should be addressed.

Abstract

Monthly inoculations of both intact plants and excised shoots of Quercus suber with the pathogenic species Botryosphaeria stevensii and Phytophthora cinnamomi were performed to investigate seasonal changes in susceptibility of this forest tree species in relation to environmental parameters and plant water status. Infection symptoms were mainly detected on seedlings inoculated from spring to autumn (April through October) with either pathogen. Mean canker sizes also showed a seasonal pattern, the higher values being recorded in the same period as above. Lesion lengths were significantly (P < 0·001) related to environmental minimum temperature. Mean daily minimum temperatures within the range of 5–12°C clearly inhibited lesion development of P. cinnamomi, whereas B. stevensii showed a less pronounced decrease in canker expansion at the same temperature range. In excised shoots of Q. suber inoculated monthly with B. stevensii, a negative linear relationship was found between the studied range of plant relative water content (81–91%) and canker length. In contrast, the lesions caused by P. cinnamomi were not significantly (P = 0·32) related to any seasonal change in water content. Some control measures for the diseases caused by both pathogens are discussed on the basis of the seasonal changes in host susceptibility observed in this study.

Introduction

Botryosphaeria stevensii[anamorph: Diplodia mutila] and Phytophthora cinnamomi are two frequent decline-associated pathogens in the Mediterranean cork oak forests (Luque & Girbal, 1989; Brasier et al., 1993; Bakry & Abourouh, 1995; Marras et al., 1995; Santos, 1995; Muñoz et al., 1996; Robin et al., 1998). Botryosphaeria stevensii, which causes dieback and trunk canker in trees harvested for cork, is responsible for serious economic losses in managed cork oak forests of north-east and south-west Spain. Extreme pathogenic effects of B. stevensii usually become apparent after cork extraction, through the development of trunk cankers that lead either to a decrease in tree cork production or even to plant death (Luque & Girbal, 1989). Phytophthora cinnamomi acts as an important contributory factor in oak decline in the south-western Iberian peninsula, having been isolated from cork oak roots both in slow declining areas and from trees affected by ‘sudden death’ (Brasier et al., 1993; Brasier, 1996).

Seasonal changes in host susceptibility to pathogens have been reported for many plant diseases. Research on susceptibility of some forest tree species to various pathogens has demonstrated the existence of such seasonal variations in eucalyptus (Shearer et al., 1987; Tippett et al., 1987; Tippett et al., 1989), Norway spruce (Horntvedt, 1988), oak (Hunter & Stipes, 1978; Robin et al., 1994), pine (Paine, 1984; Stephen & Paine, 1985; Cook et al., 1986; Swart & Wingfield, 1991), poplar (Wood & French, 1965) and walnut (Matheron & Mircetich, 1985). Seasonal changes in susceptibility have sometimes been related to climatic factors, such as temperature and water-related parameters, emphasizing the importance of environmental conditions for infection and colonization (Matheron & Mircetich, 1985; Shearer et al., 1987; Robin et al., 1994).

Research on two important aspects of pathogenesis, viz. the period of maximum host susceptibility and the optimum environmental conditions for pathogen development, is essential for successful disease management. The knowledge of environmental and host conditions for infection development by B. stevensii and P. cinnamomi on cork oak could be useful in designing new strategies for cork extraction, improving fungicide application (Fernández-Escobar et al., 1999) and increasing precision in disease epidemic models (Brasier & Scott, 1994; Brasier, 1996).

The purpose of this work was to investigate the seasonality of the lesions caused by B. stevensii and P. cinnamomi in cork oak through artificial monthly inoculations. Pathogen infection, characterized by symptom development and canker expansion, was studied in both living plants and excised shoots. Main meteorological data and relative water content of plants were related to the size of canker necroses.

Materials and methods

Plant material and pathogens

Seedling inoculations

Stem inoculation trials were performed with 1-year-old cork oak seedlings obtained from field-collected acorns. Acorns were sown in 1000-mL containers filled with a peat:vermiculite mixture (peat: Floratorf, from Floragard, Oldenburg, Germany; vermiculite: Termita 2, from Asfaltex, Valldoreix, Barcelona, Spain) in a 1:1 ratio (v/v). The substrate was amended with Osmocote Plus® (Grace-Sierra Spain, Tarragona, Spain) at 2·5 g L−1 (final pH 5·0). The seedlings were maintained in a glasshouse and irrigated twice a week to field capacity. Three months before the beginning of inoculations, a batch of 150 seedlings were transplanted to 5-L pots filled with a peat:sand 1:1·5 (v/v) mixture, fertilized with Osmocote Plus® (final pH 5·5), and moved to a shadehouse (50% sunlight transmission). Irrigation was scheduled to ensure adequate water availability to the plants throughout the experimental period.

Excised shoot inoculations

Several branches from a single adult cork oak tree were collected monthly and stored in plastic bags at 4°C for further processing later the same day. Branches were cut into straight, nonramified segments 15 cm in length and 3–6 mm in diameter and immediately stored in a moist chamber to prevent water loss before inoculation. At the same time, 10 additional shoot samples were randomly selected and used to calculate the relative water content percentage (RWC).

Pathogens

The strains used in the experiments were obtained from cork oak trees surveyed in Catalonia, north-eastern Spain (B. stevensii, Vallgorguina, UTM 31TDG6008, isolated in January 1992; P. cinnamomi, Sant Sadurní de l’Heura, UTM 31TDG9239, isolated in June 1995). Botryosphaeria stevensii was maintained in potato dextrose agar (PDA) plugs in tubes with sterile distilled water at 4°C, whereas P. cinnamomi was maintained on cornmeal agar (CMA) slants at 25°C. In preparation for monthly inoculations, a mycelial plug of each strain was transferred to PDA and incubated at 25°C for 7 days.

Inoculation methods and experimental variables

Seedling inoculations

Seedling inoculations were performed in two independent 1-year-long assays (B. stevensii in 1995 and P. cinnamomi in 1996). Monthly inoculations started in early January and continued at the beginning of every month. At the time of the inoculations, 10 plants were randomly selected and a superficial wound (15 × 4 mm) was made with a sterilized scalpel on the bark of each plant 10 cm above ground level. Either a mycelial plug obtained from the margin of a 7-day-old colony or a sterile PDA plug (control) was placed in the wound. Eight plants were inoculated with each pathogen and the remaining two plants were used as controls. All the wounds were finally wrapped with Parafilm® (American National Can Co., Greenwich, CT, USA). Inoculated plants were maintained in the shadehouse for either 8 or 14 days (for B. stevensii or P. cinnamomi, respectively) to allow for colonization and symptom development. Incubation periods were determined empirically in earlier independent tests in order to establish an optimal period for symptom readability and to avoid secondary colonization of saprophytes.

At the end of each inoculation period, the lengths of canker lesions were recorded by removing the bark from the stem and measuring the necrotic lesions upwards and downwards from the point of inoculation. Additional disease symptoms, such as chlorosis and wilting, were also recorded. Reisolation was attempted by transferring to PDA several surface-sterilized (3 min in 70% ethanol) wood pieces taken from the necrotic tissues at intervals of 0·5 cm from the point of inoculation. Cultures were grown at 25°C with a 12-h photoperiod until full colony development and identification.

Air temperature and relative humidity inside the shadehouse during the experimental period were recorded continuously using a hygrothermograph (Wilh. Lambrecht, Göttingen, Germany). The same external parameters were also recorded by an automatic weather station located 100 m away from the shadehouse.

Excised shoot inoculations

Monthly inoculations were carried out during 2000 similarly to seedling inoculations, with 10 shoots per pathogen used as replicates. Inoculated shoots were kept in two moist chambers (one per pathogenic species) and incubated in darkness at 25°C. After the incubation periods (4 days for B. stevensii and 5 days for P. cinnamomi), shoots were processed for necrosis length measurement and reisolation of pathogens using the procedures described above.

Relative water content (RWC) of the plant material used in the inoculations was calculated according to Weatherley’s procedures (Ritchie, 1984). Ten intact, current-year shoots were cut into 5-cm segments and immediately weighed (fresh weight, FW). Samples were brought to full turgidity (at 5°C, in darkness) by soaking the cuttings in distilled water until they ceased to gain weight, usually 24–30 h (turgid weight, TW). Shoots were then oven-dried (90°C) for a minimum period of 72 h and weighed again (dry weight, DW). RWC was calculated as follows: RWC (%) = (FW − DW)/(TW − DW) × 100.

Data analyses

Data were analysed using the systat statistical package (SYSTAT, 1992). Necrosis lengths were checked for normality and equal variance distributions, transformed if necessary, and analysed using the one-way anova procedure. After anova, means of each treatment were compared with Tukey’s HSD test. Regression analyses were performed to detect any potential relationship between necrosis lengths and either meteorological data during the incubation period or plant RWC. Tested models included multiple and simple linear, logarithmic and quadratic equations.

Results

Seedling inoculations

Control plants grew normally during the assays, showing new leaf production and continuous branch elongation from April to September. No abnormal symptoms, such as chlorosis, wilting and canker formation, were detected in control seedlings, and all reisolations were negative. Necrotic lesions in control plants were minimal throughout the experimental period, rarely greater than 0·1 cm and always less than 0·3 cm in length.

Plants inoculated with either B. stevensii or P. cinnamomi showed most external infection symptoms between April and October (Table 1). Eighteen plants inoculated with B. stevensii wilted during the experiments, and 21 additional plants showed symptoms of partial wilting. Total and partial wilting numbers for seedlings inoculated with P. cinnamomi were five and eight plants, respectively. However, frequencies of stem canker formation were similar for both pathogenic species (B. stevensii, 53; P. cinnamomi, 50). All plants inoculated with B. stevensii in May died, and progressively lower mortality was detected in the following months, until September. The mycelium of B. stevensii was successfully reisolated from all the seedlings except for four plants inoculated in January and four plants inoculated in February (Table 1). Plant mortality caused by P. cinnamomi was lower than that caused by B. stevensii and was restricted to the period of June to August. The mycelium of this pathogen was reisolated in all inoculation periods, but frequencies were lower during autumn and winter months (Table 1).

Table 1.  Number of seedlings showing external symptom development and mycelium recovery of inoculated Quercus suber plants (n = 8) with either Botryosphaeria stevensii (after 8 days) or Phytophthora cinnamomi (after 14 days)
Inoculation monthBotryosphaeria stevensiiPhytophthora cinnamomi
Total wiltingPartial wiltingStem cankerMycelium recoveryTotal wiltingPartial wiltingStem cankerMycelium recovery
January00060006
February00060002
March00080004
April02780157
May80880077
June34883188
July43781288
August25781288
September14880288
October03880067
November00080004
December00080004

The cankers caused by B. stevensii and P. cinnamomi varied in length during the experimental period, showing a significant effect of month of inoculation (P ≤ 0·01) (Fig. 1a). Longer necroses caused by B. stevensii were detected from April to November, with two peaks, in May and August (5·9 and 6·5 cm, respectively). Plants inoculated with P. cinnamomi showed longer lesions from June to October, with a drastic increase in June (7·6 cm) and the annual maximum in August (10·8 cm).

Figure 1.

Seasonal changes in canker necrosis length of Quercus suber seedlings (n = 8) inoculated either with Botryosphaeria stevensii (after 8 days) or Phytophthora cinnamomi (after 14 days). (a) Necrosis length. Mean values with the same letter are not statistically different according to Tukey’s multiple range test (P < 0·05). Original data were log-transformed before anova. (b) Meteorological data recorded during the monitoring periods of monthly inoculations. MT, mean temperature; Mx, maximum temperature; MMx, mean daily maximum temperature; Mn, minimum temperature; MMn, mean daily minimum temperature; RH, air relative humidity. (c) Relationship between air temperature and necrosis length. Plotted equations correspond to regression fittings with the highest r2 values.

The meteorological parameters recorded inside the shadehouse were representative of the Mediterranean climate, with a hot summer and warm winters (Fig. 1b). The maximum differences between external and shadehouse temperatures were never greater than ±3·3°C (data not shown). Summer recordings were usually greater inside the shadehouse, whereas winter temperatures were lower. Air relative humidity (RH) showed an irregular trend during the experimental period, with mean monthly values ranging from 68 to 93%. Differences between inside shadehouse and external RH mean values ranged from −7 to 14%, and RH values were usually greater inside than outside the shadehouse.

Multiple regression analyses showed no correlation between RH data and necrosis lengths but a high correlation among the different temperature parameters in both experiments was observed. Therefore, simple regression analyses were performed to detect a possible relationship between lesion length and each one of the temperature variables. The best fit obtained for infection by B. stevensii corresponded to a power equation model, whereas a quadratic was the best for P. cinnamomi (Table 2). The percentage of variation explained by the models of each regression equation ranged from 82·7 to 93·9% for B. stevensii and from 93·9 to 98·0% for P. cinnamomi. The best-fitting equations, as shown in Fig. 1(c), were obtained when the mean minimum daily temperature (MMn) was considered as the independent variable. The intervals of fitted MMn values ranged from 1·1 to 19·7°C and from 5·5 to 21·1°C for the experimental inoculation of B. stevensii and P. cinnamomi, respectively.

Table 2.  Regression data for the relationship between several meteorological variables (x) and canker necrosis length (y) of Quercus suber seedlings inoculated with either Botryosphaeria stevensii (n = 8) or Phytophthora cinnamomi (n = 8)
Pathogen, regression model and independent variablesRegression dataa
abcr2P(F)
  • a

    a, b, c, regression coefficients; r2, proportion of variation explained; P(F), anova probability for the regression equation.

Botryosphaeria stevensii (y = bxa)
Mean temperature1.8850.01760.929< 0.001
Maximum temperature3.0850.00020.827< 0.001
Mean daily maximum temperature2.5870.00110.872< 0.001
Minimum temperature
Mean daily minimum temperature1.1560.19780.939< 0.001
Relative humidity0.4470.36600.0020.882
Phytophthora cinnamomi (y = ax2 + bx + c)
Mean temperature0.037−0.6272.5680.954< 0.001
Maximum temperature0.044−1.55313.8540.939< 0.001
Mean daily maximum temperature0.039−1.0937.7280.946< 0.001
Minimum temperature0.048−0.2790.4310.947< 0.001
Mean daily minimum temperature0.041−0.3450.1750.980< 0.001
Relative humidity0.012−2.03891.9490.1540.079

Excised shoot inoculations

Mean necrosis lengths caused by B. stevensii ranged from 2·0 to 6·0 cm, whereas those of P. cinnamomi ranged from 1·2 to 3·0 cm (Fig. 2a). Lesions caused by both pathogens varied significantly in length during the experimental period (P ≤ 0·01). The colonization of excised shoots by B. stevensii followed a seasonal pattern, with progressively longer necroses during spring and summer, and lower values starting from August. Minimum development of lesions took place in the winter months (especially in January and February). However, no distinguishable seasonal pattern was observed for P. cinnamomi, in which necrosis lengths fluctuated irregularly (Fig. 2a).

Figure 2.

Seasonal changes in canker necrosis length of Quercus suber shoots (n = 10) inoculated either with Botryosphaeria stevensii (after 4 days) or Phytophthora cinnamomi (after 5 days). (a) Necrosis length. Mean values with the same letter are not statistically different according to Tukey’s multiple range test (P < 0·05). Original data were log-transformed before anova. (b) Relationship between plant relative water content and necrosis length. Plotted equations include sample correlation coefficient (r) and significance level of the regression analysis.

Mean necrosis lengths in excised shoots inoculated with B. stevensii correlated significantly with RWC values (P < 0·05). A negative linear relationship between RWC and necrosis lengths was observed, although the percentage of the variation explained by the regression equation was only about 31% (Fig. 2b). In contrast, a nonsignificant correlation was obtained for the same variables in the P. cinnamomi inoculation experiment (Fig. 2b).

Discussion

Quercus suber plants inoculated with either B. stevensii or P. cinnamomi showed similar symptoms despite different monitoring periods after inoculations. Infection symptoms were mainly detected from spring to autumn (April through October). Canker sizes also showed a seasonal pattern, with the higher values being recorded in the above period. Irrespective of the pathogen inoculated, two annual peaks in lesion lengths were detected, one in late spring (May–June) and the other in August. Whereas the greatest value in August coincided with the highest mean temperature, the peak of late spring coincided with the period of maximum active plant growth (i.e. leaf expansion and branch elongation). Increased pathogenic effects in plants during active growing periods are widely reported. Matheron & Mircetich (1985) detected an increase in susceptibility of Juglans hindsii seedlings inoculated with Phytophthora citricola during the period of maximum plant growth (May and June). A similar phenomenon was observed by Jeffers & Aldwinckle (1986) in late spring inoculations of apple trees with five Phytophthora spp. El-Hamalawi & Menge (1995) periodically inoculated plants of Persea americana with P. citricola, observing two annual peaks in the size of necroses (May and November), which coincided with the two major vegetative flushes of annual growth.

Robin et al. (1994) specifically studied changes in susceptibility of Quercus rubra to P. cinnamomi. They observed one maximum peak in necrotic lesions corresponding to June (during active branch growth) and a progressive decrease afterwards, until the absence of external symptoms in the winter months. Therefore, seasonal symptom development and changes in the sizes of lesions caused by P. cinnamomi were similar to those observed in Q. suber in the current study, with the exception of the second maximum value detected in August for cork oak. Possibly, this difference might be explained, not only by the difference in oak species, but also by differences in local climate parameters.

Both survival and colonization of P. cinnamomi in inoculated plants were low in January–March and December 1996, when mean temperatures were near 10°C and the average daily minimum temperature ranged from 5 to 10°C. Zentmyer (1980) and Shearer et al. (1987) reported optimal growth for P. cinnamomi in the range 20–32°C, but minimal growth between 5 and 10°C. In addition, Benson (1982) observed a decrease in survival of this pathogen at temperatures lower than 0°C. Matheron & Matejka (1993) observed a highly significant linear relationship between the number of days in winter with a temperature below 10°C and the length of lesions on root segments of Citrus aurantium inoculated with two Phytophthora species. Robin et al. (1994) also observed a high correlation between minimum temperatures and the size of lesions caused by P. cinnamomi on Q. rubra. The results of the current study agree with all those mentioned above, since no or small necroses were observed in the range 5–12°C (Fig. 1c).

No data regarding the effects of seasonal infection of Quercus spp. by B. stevensii have been reported earlier. In contrast with the response of P. cinnamomi to low temperatures, colonization by B. stevensii was not clearly inhibited in the range of minimum temperatures recorded, as indicated by the regression equation of necrosis length and MMn (Fig. 1c). Therefore, necroses were recorded over a longer period than those caused by P. cinnamomi.

The results obtained in the seedling inoculation experiment indicated the lack of a relationship between air relative humidity and the size of cankers, irrespective of the pathogen. Monthly inoculations of excised shoots of Q. suber with B. stevensii showed a negative linear relationship between RWC and necrosis length, but for P. cinnamomi this was not statistically supported. However, significant positive correlations between RWC and the size of lesions caused by P. cinnamomi have been reported earlier for Eucalyptus marginata and Q. rubra (Tippett et al., 1987, 1989; Shearer & Tippett, 1989; Robin et al., 1994). The lack of a relationship between canker size and tissue water content in the current study could be due to differences in either host or pathogen characteristics, as well as to procedural details, which remain to be further investigated.

The changes observed in lesion incidence in Q. suber inoculated with B. stevensii and P. cinnamomi showed the seasonal variation in susceptibility of this forest tree species. Additionally, monthly inoculations of seedlings showed that colonization by pathogens was significantly related to environmental temperature, especially to minimum values. The existence of a peak in canker size, prior to that observed in the hottest month (August) and chronologically coincident with the active plant growing season (late spring), suggested the possibility of a factor other than temperature being responsible for the increase in susceptibility. Despite several studies showing that plant water content is frequently related to seasonal changes in susceptibility, the experiments reported in this paper found that relative water content of excised shoots only partially accounted for the colonization of plant tissues by B. stevensii, while the results obtained for P. cinnamomi were not consistent.

Quercus suber showed a high susceptibility to B. stevensii throughout the year despite the existence of seasonal changes that were related to environmental temperature and plant water content. Since the period of maximum susceptibility coincides with the season of cork extraction in most Mediterranean cork oak forests (June through August), it seems necessary to carry out some control measures against B. stevensii. Direct application of fungicides (Benomyl®) on the surface of recently decorticated trunks has demonstrated its effectiveness in controlling the disease (J. Luque, unpublished data). On the other hand, the increased susceptibility of Q. suber to P. cinnamomi during the active growth of the tree suggests that treatments (by trunk injections with phosphonates, as indicated by Fernández-Escobar et al., 1999) should be applied before the first flush occurs in spring. This operation would be feasible in low-density forests, like the Iberian ‘dehesas’, but is economically impracticable in more dense areas.

Forestry of cork oak results in areas of high disturbance, especially in the south-western Iberian ‘dehesas’. Agroforestry activities in these areas include additional land uses, such as cattle rearing and other agricultural practices. Hence, the development of epidemiological models on the exclusive basis of natural biotic and abiotic factors (including host susceptibility), but neglecting human activity, would be a serious error. Robin et al. (1998) detected P. cinnamomi more frequently in managed forests than in undisturbed ones, which highlights the importance of human activities in the spread of this pathogen. Brasier (1996) also pointed out the effect of changing land-use patterns on the reduction of tree vigour and spread of P. cinnamomi, mainly due to overgrazing and neglect of traditional pasture management. In addition, cork harvesting is a crucial factor in the development of trunk cankers caused by B. stevensii (Luque & Girbal, 1989). High susceptibility periods of Q. suber to pathogens, which occur mainly in late spring and summer as shown by this study, may increase the negative effects of forest management on disease epidemics of both pathogens.

Acknowledgements

J. Luque was the recipient of a fellowship (FI-PG/94-9·806) provided by the Direcció General de Recerca de la Generalitat de Catalunya (1994–97).

Ancillary