• Phytophthora species in general, and P. quercina in particular, have been suggested in several studies to be a contributing factor to the problem of oak decline in Europe. Although Phytophthora species are generally regarded as weak competitors, few studies of the pathogenicity of species causing root rot on oaks have hitherto been performed in natural, nonsterilized forest soils. This study describes the effects of seven southern Swedish isolates of P. quercina and one isolate of P. cactorum on root vitality of Quercus robur seedlings grown in a natural, nonsterilized, acidic forest soil.
• The pathogenicity of P. quercina and P. cactorum were tested using a soil infestation test. The climatic conditions applied were an attempt to simulate summer conditions in southern Sweden.
• Both species of Phytophthora caused a significant dieback of fine roots, and necrotic lesions on coarser roots, of Q. robur seedlings. Total and live root lengths were significantly lower in infected seedlings than in controls. No significant effects of Phytophthora on above-ground growth or leaf nutrient concentration were found.
• The results demonstrate that P. quercina and P. cactorum can cause substantial root dieback of seedlings of Q. robur in natural, acidic forest soils in competition with the inhabiting soil microflora under a mesic water regime.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In addition, the soil infestation tests previously used to examine the pathogenicity to oak of P. quercina and other Phytophthora species have usually been performed on oak seedlings grown in sterile mixtures of peat, vermiculite and sand with pH values of 6.5–7.0 and, in many cases, under favourable environmental conditions for the pathogen (Jung & Blaschke, 1996; Jung et al., 1996, 1999, 2002, 2003a, 2003b; Robin & Desprez-Loustau, 1998; Robin et al., 1998, 2001; Luque et al., 2000; Sanchez et al., 2002). Few studies have focused on the effects of the pathogens under more natural conditions, using forest soils, which have an active soil microflora and also usually have a higher acidity, lower amounts of nutrients and a lower water-holding capacity than the peat–vermiculite–sand mixtures. An exception is the pathogenicity of P. cinnamomi to Quercus rubra and Quercus ilex, where forest soils have been used to a certain extent when assessing the interaction between pathogen and host (Marcais et al., 1996; Gallego et al., 1999). In a first attempt to determine the ability of P. quercina to cause root damage in forest soils, Jönsson et al. (2003b) compared the pathogenicity of two southern Swedish isolates of P. quercina to seedlings of Q. robur in two different soils – an acidic, N-rich but otherwise nutrient-poor forest soil; and a nutrient-rich peat–sand mixture with a higher pH. By contrast to what might have been expected, the results showed that the physical and chemical environment of the forest soil did not inhibit P. quercina, and that the pathogen can have detrimental effects on oak fine-root systems in acidic, nutrient-poor forest soils under a mesic water regime. However, both soils were initially sterile, and the results gave no indication of the pathogenic abilities of P. quercina when in competition with the natural soil microflora.
The next logical step is therefore to determine the interaction between Phytophthora spp. and Q. robur in a natural, nonsterilized forest soil. The influence of other microorganisms on the pathogenicity of Phytophthora species is important in defining the role of this pathogen in European oak decline, as Phytophthora species are usually regarded as weak competitors (Tsao, 1990; Erwin & Ribeiro, 1996), and the previous use of sterile soils may have influenced the survival and root infection of P. quercina and other Phytophthora species as well as the susceptibility of the host. In particular, mycorrhizal infection has been suggested to be an efficient barrier protecting the roots against Phytophthora infection (Zak, 1964; Marx, 1973; Barham et al., 1974) but also other types of microbial competition and antagonism have been suggested to be of importance for the activity of Phytophthora species (Weste & Vithanage, 1977; Keast & Tonkin, 1983).
• P. quercina and P. cactorum can infect and induce root damage and root dieback of fine and coarser roots of Q. robur seedlings grown in a nonsterilized, acidic forest soil under a mesic water regime.
• There is no significant variation in pathogenicity between the different isolates of P. quercina.
• P. quercina is more aggressive than P. cactorum on roots of Q. robur.
• Above-ground growth and nutrient concentration in leaves of Q. robur are adversely affected by the presence of P. quercina and P. cactorum in the soil.
The climatic conditions applied were an attempt to simulate natural growth conditions for Q. robur during the summer season in southern Sweden. A mesic water regime was applied, where the soil was saturated with water for only a short period each time, a condition which is assumed to be less favourable for Phytophthora infection than long periods of high or unlimited water availability (Erwin & Ribeiro, 1996). This approach was chosen because studies where maximum infection and disease development are induced are abundant in the literature (Jung & Blaschke, 1996; Jung et al., 1996, 1999, 2002, 2003a, 2003b; Robin et al., 1998; Luque et al., 2000; Sanchez et al., 2002), while attempts to simulate more natural scenarios are largely lacking. Root vitality, as well as above-ground growth and leaf nutrient concentrations, were used as measures to evaluate the direct and indirect effects of the pathogen.
Materials and Methods
Soil infestation test
Seven different isolates of P. quercina Jung. Sp. nov. and one isolate of P. cactorum (Lebert & Cohn) Schroeter, recovered from seven declining oak stands in southern Sweden, were tested for their pathogenicity towards seedlings of pedunculate oak (Q. robur L.). The seven oak stands are located between 55.3° and 56.4° latitude, on the border between the northern nemoral vegetation zone and the southern boreal zone. The mean annual temperature and mean annual precipitation in this area ranged from 6.6 to 9.0°C and 451–728 mm, respectively, between 1991 and 2001 (SMHI, 1991–2001). For the summer period (June–August), the mean temperature and precipitation ranged from 14.6 to 17.9°C and 40–73 mm (SMHI, 1991–2001). The pH (BaCl2) of the rhizosphere soil at a depth of 10–30 cm in the seven oak stands from which samples were collected varied between 3.53 and 4.97. Further information on the eight isolates of Phytophthora is given in Table 1. The pathogenicity of the Phytophthora isolates was determined by a soil infestation test, modified after Matheron & Mircetich (1985) and Jung et al. (1996, 1999).
Table 1. Species and origin of Phytophthora isolates tested for pathogenicity towards seedlings of Quercus robur, and some soil properties for sites from which isolates were recovered
A natural forest soil, sampled from a pedunculate oak stand in southern Sweden, was used in the soil infestation test. This forest soil is acidic and silty, with a chemistry representative of southern Swedish oak stands (based on data from the National Board of Forestry). The soil was sampled at a depth of 10–30 cm in the mineral soil and sieved through a 4 mm sieve to exclude roots and large particles. Apart from low pH, this soil is rich in aluminium (Al) and nitrogen (N), but low in base cations (Table 2). Subsamples of the soil were tested, at several occasions, for the presence of Phytophthora species by an oak leaf-baiting method, performed according to Jung et al. (1996, 1999). No Phytophthora or Pythium species were found in the soil.
Table 2. Nutrient concentration of forest soil used in pathogenicity test
Concentrations of base cations Al, Mn, Fe and B were determined by inductively coupled plasma spectroscopy (Perkin Elmer, CT, USA) after extraction of 20 g soil in 100 ml 0.1 m BaCl2 for 2 h. The total concentration of C was determined using an automatic LECO instrument, and total N was analysed using the Kjeldahl technique (ICP-Forest, 1998). Analyses based on five subsamples.
27.0 µg g−1
18.6 µg g−1
5.0 µg g−1
4.0 µg g−1
116.8 µg g−1
10.6 µg g−1
1.0 µg g−1
0.0 µg g−1
1.3 mg g−1
14.7 mg g−1
Total exchangeable acidity
1.4 cmolc kg−1
1.6 cmolc kg−1
The inocula consisted of 6-wk-old cultures of individual isolates of P. quercina and P. cactorum. The isolates were grown at 20°C in the dark on an autoclaved mixture consisting of 250 cm3 vermiculite and 20 cm3 whole oat grains, thoroughly moistened with 175 ml multivitamin juice broth (consisting of 200 ml l−1 vegetable juice, 800 ml l−1 demineralized water and 3 g l−1 CaCO3). The inoculum of each isolate was rinsed with demineralized water to remove excess nutrients, then mixed with the forest soil at a concentration of 20 cm3 inoculum per 1000 cm3 soil (Matheron & Mircetich, 1985; Jung et al., 1996). Controls received only a rinsed, uninfested mixture of vermiculite, oat grains and multivitamin juice broth at the same concentration.
Quercus robur seedlings, 14–18 wk old, grown from surface-sterilized acorns with a weight of 1.6–2.6 g, were classified into four different size categories based on above-ground size. The control and the eight different isolates (= nine treatments) included an equal number of seedlings from each size category. The seedlings were transplanted into individual pots with 2.5 l soil and 50.0 cm3 inoculum. In total, 20 seedlings (five from each size category) were used for each treatment, giving a total of 180 seedlings for the whole experimental setup. All seedlings were kept in a glasshouse at approx. 20–25°C and a relative humidity (RH) of 60% during daytime, and 10–15°C and 40% RH during the night. The choice of temperature was based on calculations of summer temperature (June–August) in southern Sweden (latitude 55.3–56.4°) between 1991 and 2001 (SMHI, 1991–2001). The photoperiod was approximately 15 h. A mesic water regime was applied: seedlings were flooded with deionized water for 15 min once a day, on three consecutive days, every third week, and received 200 ml deionized water once between each flooding cycle (10 d after the third day of flooding). In total, eight flooding cycles were applied. The seedlings never reached their wilting point before watering or flooding was repeated.
After 25 wk, when the seedlings were 10–11 months old, they were harvested. Based on the yellowing and wilting of seedlings and the presence of necrotic leaf spots, the above-ground condition of each seedling was estimated on a scale from 0–3 (0 = healthy; 3 = dead). The number of leaves and extensions (number of internodes on the stem) were counted and stem length was measured. Roots were separated from the rest of the plant at the point where the dicotelydon had been attached, and the soil was sieved through a 2 mm sieve to collect remaining pieces of the root system. The roots (apart from the small fragments of fine roots used for reisolation tests) were stored in sealed plastic bags at −18°C until further processing. Leaves and stems were dried at 40°C and weighed. Reisolation from the soil of each pot and from a small number of diseased fine-root fragments of each plant was performed to confirm survival and infection of the pathogen. Reisolation from the soil was performed using the oak leaf-baiting method (Jung et al., 1996, 1999), and from the diseased fine-root fragments by direct plating onto selective PARPNH agar (Jung et al., 1996, 100 ml l−1 vegetable juice and 20 g l−1 agar amended with 3 g l−1 CaCO3, 10 mg l−1 pimaricin, 200 mg l−1 ampicillin, 10 mg l−1 rifampicin, 25 mg l−1 pentachloronitrobenzene, 62 mg l−1 nystatin and 50 mg l−1 hymexazol).
The evening before washing, the roots were removed from the freezer and incubated in a cold room (5°C) to thaw. After washing, roots were separated into dead or living based on general visible criteria, resilience, brittleness, bark integrity and colour of the stele. Live roots had an intact stele and cortex, were slightly elastic and were often white or brown in colour. Dead roots often had fragmented bark, were inelastic and brittle, and were often very dark in colour. The roots were scanned, and root length, surface area and volume were measured for different root diameter classes (0–1, 1–2, 2–5, >5 mm) using the software WinRhizo Pro 5.0 (Regent Instruments, Québec, Canada). Roots were then sorted into three different diameter classes (0–2, 2–5, >5 mm), dried at 40°C and weighed. In addition, the number of seedlings subjected to each kind of treatment with a visually detectable mycorrhizal infection in any of the root tips (a well developed mantle visible in the stereomicroscope) was noted. In the following, roots with a diameter of 0–1 mm are referred to as the finest roots, roots with a diameter of 1–2 mm as finer roots, and both diameter classes together as fine roots.
Chemical analysis of leaves
The 20 seedlings subjected to treatment with a particular isolate were sorted randomly into groups of four and the leaves bulked, giving five composite samples per treatment. The leaves were then ground and subsamples digested in concentrated HNO3. The concentrations of Ca, K, Mg, Na, B, Al, Fe, Mn, Cu, Zn, S and P were determined using an inductively coupled plasma analyser (Perkin Elmer, CT, USA). The concentration of N was determined by Kjeldahl distillation. The ratios of Ca, K, Mg and P to N by weight were calculated.
The data were analysed using parametric tests, and nonparametric data were log-transformed to achieve normal distribution and homogenous variances before statistical analyses. A one-way anova was used to test for significant differences between the means for each kind of treatment. The Tukey test was used as post hoc test for multiple comparisons between all pairs of means in the case of significant differences using one-way anova. All statistical calculations were performed using the software SPSS 10 for Macintosh (SPSS Inc., IL, USA).
Phytophthora quercina and P. cactorum were reisolated from all soil samples and from most diseased, fine-root fragments of oak seedlings that had been growing in the infested soils for 25 wk. This confirms the survival and root infection of both P. quercina and P. cactorum in the acidic forest soil. None of the pathogens was recovered from the soil samples or the fine-root fragments of the control seedlings.
There was a visible difference in the dieback of the roots, as well as in the size of the root systems, of many of the infected seedlings and the control seedlings (Fig. 1). Both Phytophthora species caused fine-root decay, with dieback of nonsuberized as well as suberized fine roots. Necroses were observed mainly on fine roots, although some necroses and dieback of coarser roots (diameter 2–5 mm) was also observed. The necroses on the coarser roots usually developed via infection of nonsuberized lateral roots. In many of the infected seedlings the tap roots were dead, and dieback of mother roots was common. A few seedlings showed an abnormal branching pattern, with extensive sprouting of fine roots where dieback of the tap root had occurred. On the coarser roots (diameter 2–5 mm and >5 mm), wounds in the bark were apparent. The lesions varied in size, but were usually around 2–5 mm in width and 5–10 mm in length. However, some of the lesions reached a length of several cm, with the longest lesion being 55 mm in length. Many, but not all, of the wounds were callusing. Lesions were also present on some of the surviving tap roots. No signs of pathogen infection could be seen on the roots of the control seedlings. Some dieback of fine roots occurred in the control seedlings, but not to the same extent as in infected seedlings. Only a few of the control seedlings showed dieback of tap roots. Both control seedlings and infected seedlings had mycorrhizal root tips, but visible infections were sparse (≤ 10% of the root tips). No significant difference in the number of seedlings with mycorrhizal root tips could be detected between treatments. The mean number of seedlings with visually detectable mycorrhizal infection in any of the root tips was 16.1 (SD ± 0.8, n = 180).
Live root length was significantly higher in control seedlings than in seedlings infected with P. cactorum and all isolates of P. quercina except for isolate 1 (Fig. 2). The difference in live root length consisted mainly of the difference in the length of live fine roots. The differences were most pronounced for the finest roots, where seedlings infected with any isolate of P. quercina and P. cactorum differed significantly from the control seedlings (Fig. 2b). For coarser roots there were no differences between control and infected seedlings.
Dead root length showed a less obvious pattern. The total dead root length differed significantly only between the control seedlings and some of the isolates. Infection with P. cactorum and P. quercina isolates 3, 5 and 7 caused a significantly higher amount of dead root length than was observed in the control seedlings, while the effects of other isolates did not differ from the control (Fig. 3). The differences in dead root length were most pronounced with regard to finer roots and roots with a diameter of 2–5 mm, where P. cactorum and all isolates of P. quercina, except for isolate 6, showed significant differences compared with the control (Fig. 3c,d).
On average, less than 7% of the total root length of the control seedlings was dead, while for seedlings infected with P. quercina the percentage of dead root length varied between 18 and 32% of total root length (Table 3). Seedlings infected with P. cactorum had, on average, 25% dead root length (Table 3). The fine roots showed similar values, with 7% dead root length in the control seedlings and between 18 and 33% dead root length for the infected seedlings. However, the variation was high within each type of treatment, and dead root length varied between 0 and 50% for infected seedlings. The percentage of dead root length in relation to total root length differed significantly for all isolates of P. quercina and P. cactorum compared with control seedlings for all roots, as well as for fine roots (Table 3). There were also significant differences between the isolates of P. quercina. None of the roots with a diameter above 10 mm was dead.
Table 3. Total root length and percentage of dead root length in relation to total root length for all root diameters together and for fine roots of Quercus robur seedlings
Values are mean ± SD (n = 20). Statistics are for one-way anova.
P ≤ 0.05;
P ≤ 0.01;
P ≤ 0.001; ns, not significant. In the case of significant differences when applying one-way anova, small letters denote statistical results of post hoc test (Tukey). Different letters indicate significant differences.
The total length of roots was significantly higher in control seedlings than in seedlings infected with P. quercina (Table 3). The exceptions are P. cactorum and P. quercina isolate 1, where total root length did not differ from that of control seedlings. The differences in total root length were mainly caused by differences in the length of fine roots (Table 3). The surface area and volume of roots followed approximately the same pattern as root length. Root biomass was found to be a less sensitive parameter (data not shown).
Phytophthora quercina and P. cactorum were equally aggressive to the oak seedlings (Figs 2, 3; Table 3). However, seedlings infected with P. quercina isolate 1 and with P. cactorum tended to have a somewhat greater live and total root length (although not significantly different from seedlings infected with the other isolates: Fig. 2; Table 3).
Above-ground condition and growth
None of the seedlings died during the experiment. The control seedlings appeared healthier than those infected with P. quercina, as many of the infected seedlings had yellowish leaves with green veins and scattered necrotic leaf spots or coherent brown areas. Most of the infected seedlings with severe root rot showed yellowing and wilting of leaves and interveinal chlorosis. However, the estimated above-ground condition (based on yellowing, wilting and necrotic spots on leaves) of the infected seedlings differed significantly from the control only for seedlings infected with P. quercina isolates 2, 4 and 7 (Fig. 4a). The seedlings infected with P. cactorum appeared healthier than those infected with P. quercina, and could not be distinguished from the control seedlings based on above-ground appearance.
Above-ground growth, as expressed by leaf biomass and stem length, showed no significant response to the loss of fine roots and the damage of coarser roots (Fig. 4c,d). Consequently there was no difference in total above-ground biomass between seedlings (data not shown). However, the number of apical buds was significantly higher in control seedlings than in those infected with P. cactorum and P. quercina isolates 2, 3, 5 and 6 (Fig. 4b). As root biomass is a less sensitive parameter than root length, the total biomass of seedlings (above- and below-ground) did not differ either (data not shown).
Leaf nutrient concentration
There were no consistent differences in leaf nutrient concentrations or nutrient ratios to N between the control seedlings and those infected with P. quercina and P. cactorum (Table 4). However, the N concentration tended to be somewhat higher in control than in infected seedlings (except for those infected with P. quercina isolate 3), and some seemingly random differences in concentration of a few elements appeared between the isolates of P. quercina (Table 4).
Table 4. Nutrient concentrations and nutrient ratios to N (by weight) in leaves of Quercus robur seedlings at the end of the soil infestation test
This study investigated the aggressiveness of seven southern Swedish isolates of P. quercina and one southern Swedish isolate of P. cactorum to seedlings of Q. robur in a natural, nonsterilized, acidic oak forest soil under a mesic water regime. The average fine root damage of about 26%, together with the reisolation of all seven isolates of P. quercina and the one isolate of P. cactorum, both from forest soil and from necrotic fine-root fragments, demonstrates the ability of the pathogens to survive, infect and induce dieback of fine and coarse roots of oaks in the presence of other soil microorganisms, and under adverse environmental conditions in the soil (acidic soils and limited water availability). This indicates that both Phytophthora species can compete well with other microbes in the forest soil. Previous studies on the pathogenicity of Phytophthora species to Q. robur have been performed in sterilized soils, and the initial lack of microbes in these soil infestation tests might have favoured the survival and root infection of the pathogen, as well as altering the susceptibility of the host, and thereby biasing the results of the tests (Jung et al., 1996, 1999, 2002, 2003a, 2003b; Jönsson et al., 2003b). However, performing the pathogenicity test in a natural, nonsterilized forest soil, harbouring a natural community of microorganisms, makes the results obtained here valuable for the interpretation of the pathogenicity of P. quercina, and to a certain extent also P. cactorum, in the field.
In particular, mycorrhizal infection of plants has been suggested to be important in protecting the roots mechanically and biochemically against Phytophthora infection (Zak, 1964; Marx, 1973; Barham et al., 1974). In the present study there was no difference between treatments at the end of the experiment in the number of seedlings with visually detectable ectomycorrhizal infection. However, the measure used is very coarse, and does not quantify the mycorrhizal infection of the seedlings. A more thorough estimate would have been to count the number of mycorrhizal root tips, or preferably to quantify the mycorrhizal infection of fine roots by ergosterol analysis (Nylund & Wallander, 1992).
The high infection rate and the considerable dieback of fine roots, together with necrosis and lesions on coarser roots in the nonsterilized forest soil, demonstrate the aggressiveness towards the host of both P. quercina and P. cactorum. In accordance with Jung et al. (1996, 1999) and Jönsson et al. (2003b), fine roots were most susceptible to infection (Fig. 3b,c; Table 3). The effects of the pathogens on the fine roots of Q. robur seedlings were expressed as a significant difference in live root length and total root length, but to a lesser degree in the amount of dead root length (Figs 2, 3; Table 3). This is likely to be caused by the presence of saprophytic soil microorganisms from the beginning of the experiment, implying that decomposition proceeded continuously. Hence the amount of dead root length in the soil at harvest is a consequence not only of dieback of roots, but also of decomposition rate. The influence of decomposition rate is most pronounced for the finest roots, as indicated by the significant difference in dead root length for all isolates compared with control seedlings for finer and coarser roots, while there are only a few significant differences for the finest roots (Fig. 3). This is in accordance with several studies on decomposition of roots of different tree species, which show an initially lower decomposition rate as a function of a larger root diameter (Berg, 1984; Fahey et al., 1988; King et al., 1997; Usman et al., 2000; Silver & Miya, 2001).
In addition to the decomposition of dead roots, the rate of root production is likely to contribute to differences in total root length and in live root length between control and infected seedlings. The infected seedlings are exposed to a stress factor that destroys an essential part of the plant, and in an advanced stage of pathogen infection their growth rate is likely to become lower than that of the control seedlings. The ratio between root death and root replacement by seedlings may easily become unbalanced, thereby reducing the amount of living roots. In addition, the adverse chemical conditions in forest soil – low pH, high Al concentration and low amounts of base cations (Table 2) – as well as competition for resources with soil microorganisms, may add to the stress caused by the pathogens. Soil chemical stress was suggested to be a possible reason for the differences in damage caused by P. quercina on root systems of seedlings grown in an acidic forest soil and a nutrient-rich peat–sand mixture (Jönsson et al., 2003b).
The quantitatively homogeneous effects of the isolates of P. quercina (Figs 2, 3; Table 3) may be due to a small genetic variation. All Phytophthora isolates were recovered from sites with similar soil conditions (loamy to clayey soil texture, mesic soil moisture and pH (BaCl2) between 3.5 and 4.0, with the exception of isolate 5 where the pH was 5.0) and they were therefore probably adapted to acidic mesic soil conditions. The tendency (although not significant) for seedlings infected with P. quercina isolate 1 and P. cactorum to have a somewhat greater live and total root length than seedlings infected with other isolates may be caused by a lower aggressiveness or competitive ability of these isolates, or possibly a slower rate of infection and disease development. For P. quercina isolate 1, a lower competitive ability or a decrease in aggressiveness during storage is likely as isolates 1 and 7 were equally aggressive in the study of Jönsson et al. (2003b). In many cases, adaptive changes or mutations may occur after prolonged culturing (Erwin & Ribeiro, 1996) and, for P. infestans and P. parasitica var. nicotianae, several studies have shown a loss of virulence when the pathogen was grown in culture or was continuously subcultured (Apple, 1957; Jeffrey et al., 1962; Pietkiewicz, 1979; Goth, 1981). Evaluation of the pathogenicity of P. cactorum is complicated by the use of only one isolate.
The climatic conditions applied in this study were an attempt to evaluate the impact of Phytophthora on oak seedlings under conditions less favourable for pathogen infection and disease development, rather than estimating the maximum potential pathogenicity of Phytophthora. The results confirm the previous findings by Jönsson et al. (2003b) that P. quercina, and also P. cactorum, can cause substantial root damage under mesic water regimes, and not only when weather or site conditions involve periods of high or unlimited water availability. This is in accordance with studies of the interaction between P. cinnamomi and Q. ilex and Q. suber, which have demonstrated that flooding did not enhance the root rot caused by P. cinnamomi (Robin et al., 2001), and that this Phytophthora species is able to infect and induce damage even when waterlogging is absent or occurs only during short periods (Sanchez et al., 2002). Periods of drought may reduce the tolerance of host to pathogen, thereby enhancing the ability of Phytophthora species to cause root damage on rewetting. The ability of P. quercina to survive droughts, and to cause higher amounts of root damage to Q. robur after re-flooding than in treatment with moist soil conditions between flooding cycles has been shown by Jung et al. (2003a). For P. cinnamomi, on the other hand, drought stress did not increase the amount of damage caused on roots of Q. ilex, Q. suber and Q. rubra (Robin et al., 2001).
The nonsignificant effects on above-ground growth (Fig. 4) and the few random differences in leaf nutrient concentration (Table 4) appear to agree well with previous studies, suggesting that Phytophthora can infect the roots of woody plants months to years before foliage symptoms are detected, and that root rot must usually be severe before significant effects can be seen on the aerial parts of the plants (Tsao, 1990; Erwin & Ribeiro, 1996). However, an early indication of a possible influence of the fine-root damage on above-ground growth is the significantly lower number of apical buds in the infected seedlings (Fig. 4). The significant amount of root rot in infected seedlings (19–33% of dead fine-root length) apparently did not influence leaf nutrient concentration during the period of this experiment. Both infected and control seedlings had leaf nutrient concentrations within the range of what is considered sufficient according to Bergmann (1988) and Linder (1995), indicating that infected seedlings must have had enough roots remaining for adequate nutrient uptake. However, the concentrations of K, Cu and N in the leaves were in the lower range of what is considered as optimum levels for oak according to Bergmann (1988), but the lack of significant differences between control and infected seedlings suggests that the low concentrations are more probably due to low availability of the elements in the soil, or strong competition for the them, rather than to the size and activity of the root system. The ratios of each of the elements to N exceeded the target values suggested by Bergmann (1988) and Linder (1995), indicating that the seedlings are N-limited. The yellowing and interveinal chlorosis of the leaves of the infected seedlings is likely to have been caused by leaf necrotic proteins (elicitins), which can be produced by P. quercina (Heiser et al., 1999; Brummer et al., 2002).
Although not significant, the N concentration in the leaves showed a tendency to be somewhat lower in infected than in control seedlings. A similar tendency was found by Jönsson et al. (2003b). It is possible that this may be an early indication of reduced nutrient uptake, but it may also be caused by a translocation of N in the infected seedlings, from the leaves to the roots, in order to sustain new root production and/or to produce defensive compounds to counteract the Phytophthora attack. Alkaloids, for example, are efficiently used as defensive agents in plants, and may be moved around within the plant to parts that need greater protection during growth and development (Crawley, 1997). Other antimicrobial substances that may be produced in attacked root cells, and which may influence the distribution of N within the plant, are pathogenesis-related proteins and possibly also N-containing enzymes necessary for the production of phytoalexins (Agrios, 1997; Lambers et al., 1998).
In conclusion, this study demonstrates that both P. quercina and P. cactorum can cause substantial root dieback of fine roots, as well as necrotic lesions on coarse roots of Q. robur in a natural acidic forest soil in the presence of the inherent soil microflora and under a mesic water regime. The stress caused by the two Phytophthora species led to a low replacement of dead roots in infected seedlings. Although severe dieback of fine roots was observed, no significant effects on above-ground growth or on leaf nutrient concentration could be detected after 6 months. These results improve our knowledge about the pathogenicity of P. quercina and P. cactorum to Q. robur, which previously has been limited to effects in sterilized soil. As a complement to the findings in this study, a thorough investigation of how mycorrhizal infections affect the pathogenicity of Phytophthora species, host susceptibility and subsequent disease severity would be desirable in order to understand the pathogenicity of Phytophthora to seedlings of Q. robur in natural forest soils.
This project was generously funded by CF Lundströms Fund, Erik and Ellen Sökjers Petersens Fund, The Consolidated Funds of the Royal Swedish Academy of Agriculture and Forestry, and The Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning. I am very grateful to Thomas Jung and Ulrika Rosengren for constructive discussions and valuable comments. I am also grateful to Håkan Wallander for reading and commenting on the manuscript. Anita Balogh and Maj-Britt Larsson helped in starting up the experiment and in harvesting the plants. Ragnhild Ohlin washed the roots, and performed the chemical analyses together with Maj-Lis Gernersson. Helen Sheppard corrected the language.