A survey was made on the occurrence of soilborne Phytophthora species in 35 oak stands on a range of geologically different sites in Bavaria. The most widespread species were P. quercina, P. cambivora and P. citricola. Seven other Phytophthora species were isolated infrequently. The fine root systems of 106 healthy and 111 declining mature trees of Quercus robur and Q. petraea were intensively investigated. The results indicate that, depending on the site conditions, at least two different complex diseases are referred to under the name ‘oak decline’. On sites with a mean soil pH (CaCl2) 3·5 and sandy-loamy to clayey soil texture Phytophthora spp. were commonly isolated from rhizosphere soil, and highly significant correlations existed between crown transparency and various root parameters. Oaks with P. quercina or other Phytophthora spp. in their rhizosphere had markedly higher levels of fine root damage than oaks without Phytophthora spp., and were subject to a relative risk of severe crown symptoms of 2·1 and 2·8, respectively. In contrast, in stands with sandy to sandy-loamy soils and a mean soil pH 3·9, Phytophthora spp. were not found. In these stands, correlations between crown transparency and various root parameters were either less significant or not significant. It is concluded that Phytophthora species are strongly involved in oak decline on sandy-loamy to clayey sites with a mean soil-pH (CaCl2) 3·5.
Since the beginning of the twentieth century, oak decline has been a serious and frequently occurring disease of European oak forests (Delatour, 1983; Oleksyn & Przybyl, 1987; Hartmann et al., 1989). Being an episodic phenomenon of local or regional importance in the past, oak decline, in its current phase, has been going on since the beginning of the 1980s, and is occurring all over Europe (Delatour, 1983; Hartmann et al., 1989; Siwecki & Liese, 1991; Luisi et al., 1993). Above-ground symptoms include dieback of branches and parts of the crown, formation of epicormic shoots, high transparency of the crown, yellowing and wilting of leaves and tarry exudates from the bark (Siwecki & Liese, 1991; Luisi et al., 1993), all symptoms indicative of water stress and poor nutrition. Mortality rates may be up to five trees per hectare per year (Hartmann et al., 1989). In Bavaria, 28% of the oaks showed severe crown damage (crown transparency > 25%) in 1999 and severely damaged oak stands with mortality due to unknown causes were found throughout all growth regions (Anonymous, 1999).
In Spain and Portugal, a strong association was shown between the presence of Phytophthora cinnamomi, interacting with prolonged drought periods, and rapid mortality and decline of cork (Quercus suber) and holm oaks (Q. ilex), which display similar disease symptoms to declining pedunculate (Q. robur) and sessile oaks (Q. petraea) in Central and Western Europe (Brasier, 1993, 1996; Brasier et al., 1993a,b; Cobos et al., 1993; Gallego et al., 1999). In South-eastern France, P. cinnamomi was also isolated from cork and holm oaks. Its pathogenicity towards both oak species was proven and its possible involvement in the decline process proposed (Robin et al., 1998). The recovery of declining cork and holm oaks following trunk injections with potassium phosphonate further confirmed that Iberian oak decline is most probably a Phytophthora-related disease (Fernandez-Escobar et al., 1999).
In Central and Western Europe, deterioration of oak fine roots was observed in several studies (Näveke & Meyer, 1990; Vincent, 1991; Eichhorn, 1992; Blaschke, 1994; Jung, 1998; Thomas & Hartmann, 1998). Assessment of the root systems of declining and healthy oaks in 33 stands involving five European oak species all over Central Europe showed a progressive destruction of the fine root system, dieback of long roots and necrotic lesions on suberized and nonsuberized roots. These symptoms were observed in both healthy and declining oaks, but the damage was generally more severe in decline situations (Blaschke, 1994; Jung, 1998). Attempted isolations from fine roots and rhizosphere soil samples revealed the widespread occurrence of several Phytophthora species including P. cactorum, P. citricola, P. cambivora, P. gonapodyides, and the new species P. quercina in 29 out of the 33 stands (Jung, 1998; Jung et al., 1996). In Northern France, Galoux & Dutrecq (1990) isolated a Phytophthora species from fine roots of declining oaks, and Hansen & Delatour, 1999) demonstrated a diverse Phytophthora population including P. quercina in oak forest soils.
In several soil infestation tests, P. quercina and P. cambivora were the most aggressive species to root systems of young Q. robur plants (Jung, 1998; Jung et al., 1996, 1999a). In another test with various phytophthoras and a range of broadleaf tree species, P. quercina proved to be host-specific to the genus Quercus (Jung et al., 1999b).
In biotests, several of the isolated Phytophthora species released toxic substances into their culture medium that induced wilting and intercostal chlorosis and leaf necrosis in Q. robur cuttings (Jung, 1998): symptoms often observed in declining oaks in the field. Later, Heiser et al. (1999) showed for P. quercina, P. citricola and P. gonapodyides that these substances belong to the family of Phytophthora leaf necrotic proteins called elicitins.
In this study, the relationships between the root condition and crown status of healthy and declining oaks, the population of soilborne Phytophthora species and site factors, especially geological substrate, soil pH and soil texture, were investigated, with the aim of improving understanding of the role of Phytophthora pathogens in oak decline in Central Europe.
Materials and methods
Study sites and sample trees
In 1997 and 1998, a total of 106 healthy (crown transparency ≤ 25%; no dieback symptoms) and 111 declining oaks (crown transparency ≥ 35%; dieback symptoms if present) were sampled in 35 oak stands (three to four healthy and three to four declining oaks per stand) in seven different growth regions in Bavaria on a broad range of typical oak forest sites (Table 1, Fig. 1). Inside each stand the sample trees had a scattered distribution in order to get results representative for the whole stand and to try to avoid the possibility of healthy sample trees sharing parts of their rhizosphere with declining trees. In August 1997, crown transparency was assessed according to Anonymous (1994) and Anonymous (1996). The mean crown transparency of the declining oaks was 63·2%, compared with 21·0% in the healthy oaks.
Table 1. Occurrence of Phytophthora species in oak stands on different geological substrates and soil types in Bavaria
namest="col7" nameend="col17" rowsep="1">No. of stands with occurrence of Phytophthora/Pythiumd
a stand ID numbers according to Fig. 1. b g = gravelly, s = sandy, u = silty, l = loamy, c = clayey, S = sand, L = loam, C = clay.
m = moist; mm = moderately moist; d = dry; md = moderately dry; pfw = periodically fluctuating watertable.
CAC =Phytophthora cactorum; CAM = P. cambivora; CIT = P. citricola; GON = P. gonapodyides; MEG = P. megasperma; QUE = P. quercina; SYR = P. syringae; P.sp.2 =Phytophthora species 2; P.sp.3 =Phytophthora species 3; P.sp.5 =Phytophthora species 5; UND =Pythium undulatum.
The relationship between defoliation and presence or absence of Phytophthora spp. in the soil was investigated in a 45- to 50-year-old stand (stand 14, Table 1, Fig. 1) that was completely defoliated by caterpillars of Tortrix viridana, Lymantria dispar and Thaumatopoea processionea in 1994 and 1995. Four healthy oaks at the totally recovered site on the upper slope, and four healthy and four declining oaks at the declining site at the bottom of the slope and in the depression, were sampled.
Field sampling, isolation method and root analysis
In each stand, root samples were taken only once at the first sampling date, while soil used for isolation tests was sampled 2–3 times in different seasons.
Three to five soil-root-monoliths (size about 20 × 30 × 30 cm) were taken around each tree at a distance of 50–150 cm from the stem base. Oak roots with diameter ≤ 5 mm extracted from the monoliths were forwarded for root analysis. Aliquots of rhizosphere soil from all monoliths were bulked (about 1 litre) and subsamples were used for isolation tests and soil analysis.
Isolations were carried out using 2- to 7-day-old leaflets of Q. robur seedlings as baits floated over flooded soil (Jung, 1998; Jung et al., 1996). Infected brownish leaflets, which normally appeared after 3–7 days, were cut into pieces, plated onto selective PARPNH-agar (V8-agar amended with 10 µg mL−1 pimaricin, 200 µg mL−1 ampicillin, 10 µg mL−1 rifampicin, 25 µg mL−1 pentachloronitrobenzene (PCNB), 50 µg mL−1 nystatin and 50 µg mL−1 hymexazol) and incubated at 20°C in the dark. After 24–48 h, Phytophthora hyphae were transferred under the stereomicroscope onto V8-agar.
For soil analyses, which were performed by the Institute of Soil Sciences of the Technische Universität München, water extracts were prepared (10 g soil/40 mL H20−1). Nitrate and sulphate were analysed using ion chromatography (DX−120, Dionex, Idstein, Germany), ammonium using a segmented flow analysis system (Skalar, Erkelenz, Germany), and aluminium, calcium and manganese using ion coupled plasma ICP (Optima 3000, Perkin-Elmer, Überlingen, Germany). The pH was measured both in water (= pH (H2O)) and in calcium chloride (= pH (CaCl2)).
Root samples of each oak were thoroughly washed and analysed as follows. The percentage of root damage (referred to here as root damage) was estimated visually after spreading the roots uniformly and randomly on 50 × 30-cm trays etched with 6 × 10-cm rectangles. Afterwards, living fine roots (diameter < 2 mm) per m mother root (= small woody roots with diameter = 2–5 mm) were counted. After drying at 65°C, fine roots were scanned and weighed. With the computer software Delta-T Scan 2·04 (Delta-T Devices Ltd, Cambridge, UK) the total fine root length and the number of root tips were determined. Subsequently the specific fine root length (SFRL; = fine root length/dry weight fine roots [cm g−1]) and the specific root tip density (SRTD; = fine root tips/dry weight fine roots [n g−1]) were calculated. After drying at 65°C, length and dry weight of the mother roots were measured. With these data, the parameters fine root length/mother root length (FRL/MRL), fine root tips per m mother roots (FRT/MRL [n m−1]) and FRL/dry weight mother roots [cm g−1] were calculated.
The root samples of different stands were taken in different seasons and during different weather conditions. Since the fine root system is subject to a strong seasonal and weather-induced turnover, the absolute values of the root parameters of healthy and declining trees can certainly be compared within one stand, but comparisons between stands should be interpreted with caution. Therefore, the absolute values of all root parameters of each tree were transformed into season- and weather-corrected relative values. For each root parameter, the mean value of the healthy sample trees of a stand was considered as stand internal standard (= 100%), and the single values of all healthy and declining sample trees of the stand were compared with this standard. The statistical analysis was made with the absolute and the relative values. The data for all Phytophthora-infested stands and the data for all stands not infested with Phytophthora (referred to here as Phytophthora-free stands) were combined, and the two datasets were analysed separately in order to calculate differences between both stand categories. For each root parameter, the significance of the difference between the mean values of the healthy and the declining oaks was tested with the nonparametric Mann–Whitney test. The nonparametric Spearman Correlation was used to analyse whether the root parameters assessed were significantly correlated with the crown transparency. The relationship between the presence of P. quercina and other Phytophthora spp. in the rhizosphere and the risk of oaks developing severe above-ground symptoms (crown transparency ≥ 35%) were tested with contingency tables. Furthermore, the relationship between crown status and presence/absence of P. quercina was analysed using a logistic regression. Finally, the significance of the difference in crown transparency and each root parameter between the mean values of oaks with and without Phytophthora spp. in their rhizosphere (in the following referred to as oaks infested and not infested with Phytophthora spp.) was tested using the Mann–Whitney test. All statistical analyses except for the logistic regression (programme SPSS for Windows 6·1) were made using the programme Prism 3 for Windows 95 (GraphPad, San Diego, USA).
Distribution of Phytophthora species and correlation with crown status and site factors
Ten different Phytophthora species were isolated from rhizosphere soil of 73 (out of 124) oaks in 19 (out of 35) stands. In addition, several Pythium species, including Py. undulatum and Py. anandrum, and a Saprolegnia sp. were frequently recovered from soil samples of Phytophthora-infested stands. P. quercina was the species with the widest geographical distribution (18 stands in five growth regions) (Fig. 1, Table 1) and the highest isolation frequency, especially from declining oaks. P. quercina was recovered from 41 out of 65 declining (63·1%) and from 14 out of 59 healthy oaks (23·7%). P. cambivora, P. citricola and Py. undulatum were isolated from seven stands, while the other species occurred only sporadically (Table 1). Descriptions of the new Phytophthora species 2, 3 and 5 as P. psychrophila sp. nov., P. europaea sp. nov. and P. uliginosa sp. nov. have been proposed (T Jung and EM Hansen, unpublished).
Oaks with P. quercina in their rhizosphere were 2·1 times more likely (P < 0·0001) to exhibit severe crown symptoms (transparency ≥ 35%) than oaks without P. quercina. Considering all 10 Phytophthora species, the relative risk increased to 2·8 (Table 2).
Table 2. Contingency table: occurrence of P. quercina and Phytophthora spp./crown status in 19 Phytophthora-infested oak stands
A significant correlation existed between crown status and the occurrence of P. quercina in the rhizosphere (logistic regression analysis; r = 0·2870; P < 0·0001).
The distribution of Phytophthora species showed a clear relationship to the geological substrate and the texture and pH of the soil (Table 1). Phytophthora spp. were isolated from soil samples of stands on a broad range of geological substrates characterized by sandy-loamy to loamy, silty or clayey soil texture (in the following referred to as heavy soils) and a mean soil pH (CaCl2) between 3·5 and 6·6. On the other hand, all isolation attempts failed in oak stands on triassic and jurassic sandstones, pleistocene gravels and chalk with mean soil pH values ≤ 3·9 and mainly sandy to sandy-loamy soil texture (in the following referred to as sandy soils). Even within single stands, the distribution of Phytophthora spp. in the soil followed this pattern. In oak stands 20 and 22 (Table 1 and Fig. 1), all sample trees without Phytophthora spp. in their rhizosphere were growing on microsites with sandy dystric or podzolic cambisols and soil pH values between 3·3 and 4·1, whereas all oaks infested with P. quercina and various other Phytophthora species were growing on microsites with clayey vertic cambisols and soil pH values between 3·7 and 6·0.
The results of the soil analyses are summarized in Table 3. The soils of Phytophthora-infested and Phytophthora-free stands differed significantly (P < 0·0001) in their mean pH (4·16 ± 0·85 vs. 3·56 ± 0·28) and their mean calcium content (28·35 ± 37·36 p.p.m. vs. 5·54 ± 4·61 p.p.m.). Nitrate concentrations were quite high in both stand categories (35·60 ± 40·49 and 31·0 ± 25·7 p.p.m., respectively). Differences in pH and mineral contents between healthy and declining oaks in both stand categories were not significant.
Table 3. Mean concentrations (p.p.m.) of selected minerals and mean pH values of soil samples from oak stands examined
Among all Phytophthora species isolated, P. quercina was the least demanding with regard to site conditions. The pathogen was recovered from various nonhydromorphic sites as well as from sites with periodically fluctuating watertables. P. quercina displayed the highest plasticity concerning the soil pH (3·5–7·0), followed by P. citricola (3·5–6·6), while P. syringae (3·4–3·9, except for one isolate at 6·0) and Py. undulatum (3·3–3·8) seemed to be more acidophilic species (Table 4).
Table 4. pH range of oak forest soils with occurrence of different Phytophthora and Pythium species
No. of isolates
pH range (CaCl2)
3·65–4·39 (1 isolate 6·95)
3·39–3·85 (1 isolate 6·02)
In oak stand 14 no Phytophthora species were recovered from the rhizosphere of four healthy oaks growing on the upper slope on a moderately fresh orthic luvisol. In contrast, P. quercina and various other Phytophthora spp. were isolated from rhizosphere soil of all declining and almost all healthy sample trees on a stagno-gleyic luvisol with periodically fluctuating watertables at the bottom of the slope and in the depression.
Health status of the root systems and correlation with the crown status
The mean crown transparency of the healthy and declining sample trees was 22% and 62%, respectively, in the Phytophthora-infested stands, and 20% and 66%, respectively, in the Phytophthora-free stands.
The results of the root analyses, including statistical analysis for five out of seven root parameters examined, are summarized in Tables 5 and 6. Regarding the absolute values of all root parameters except root damage, both healthy and declining oaks in the 16 Phytophthora-free stands on sandy to sandy-loamy sites were superior to those in the 19 Phytophthora-infested stands on sandy-loamy to clayey sites (Table 5). This is in accordance with the results of Thomas & Hartmann, 1998) for oak stands on sandy and clayey soils in Northern Germany, and might mainly be due to differences in site conditions but also to fine root damage caused by Phytophthora spp.
Table 5. Root parameters of healthy and declining oaks and significance of differences (Mann–Whitney test)
In the Phytophthora-infested stands, the differences between healthy and declining oaks for the absolute values of the parameters root damage, number of fine roots per m mother root, fine root length/mother root length, fine root tips/mother root length and fine root length/dry weight mother roots were significant (P < 0·0001). The nonsignificance of the parameters SFRL and SRTD is not surprising because their absolute values are strongly affected by season and weather. In the Phytophthora-free stands, differences between healthy and declining trees were markedly smaller and mostly not significant (Table 5).
Considering the season- and weather-corrected relative values, the differences between the two stand categories are even more evident. The differences between healthy and declining oaks were highly significant (P < 0·0001) for all root parameters in the Phytophthora-infested stands, but either less significant (P < 0·05) or even not significant for almost all root parameters in the Phytophthora-free stands. In addition, the level of the differences was much higher in the Phytophthora-infested than in the Phytophthora-free stands. Regarding the fine root specific parameters SRTD and SFRL, declining oaks were even superior to healthy oaks in the Phytophthora-free stands (Table 5).
The nonparametric Spearman-correlation analysis showed significant (P < 0·0001) correlations between crown transparency and the absolute values of five and the relative values of all root parameters in the Phytophthora-infested stands. For almost all root parameters the correlation coefficients were markedly higher if the relative instead of the absolute values were used for the calculation. The strongest correlations were found for crown transparency and the relative values of the parameters root damage (r = 0·55), and fine root length/mother root length, fine root tips/mother root length and fine root length/dry weight mother roots (r = − 0·45).
On the other hand, almost no significant correlations between crown transparency and the absolute values of all root parameters were found in the Phytophthora-free stands. Even for the relative values, significant correlations (P < 0·001) existed only for the parameters root damage, fine root length/mother root length and fine root tips/mother root length.
Not only the significance level of the correlations between crown transparency and all root parameters, but also all correlation coefficients were markedly higher in the Phytophthora-infested (r = 0·3–0·55) than in the Phytophthora-free stands (r = 0–0·33). In some heavily declining Phytophthora-infested stands, correlation coefficients up to 0·7 were found (data not shown).
Oaks with P. quercina or other Phytophthora spp. in their rhizosphere displayed an average crown transparency of about 50%, compared with 30% in oaks without Phytophthora spp. Also the level of fine root damage was significantly higher in oaks infested with P. quercina alone or Phytophthora spp. (Table 6).
Table 6. Crown transparency and root parameters (relative values in percentage) of oaks with and without presence of Phytophthora spp. in 19 Phytophthora-infested stands, and significance of differences (Mann–Whitney test)
It was shown that Phytophthora species are widespread on a range of different geological substrates with sandy-loamy, loamy, silty or clayey soils with a mean pH (CaCl2) between 3·5 and 6·6. P. quercina was by far the most frequently isolated species and had the highest plasticity concerning geological substrate and soil type, texture and pH. In addition, P. quercina was isolated more frequently from declining (63·1%) than from healthy oaks (23·7%) on these sites. This latter finding is in contrast to the results of Hansen & Delatour (1999) and Delatour et al. (2000) who found, in four stands in north-east France, no evident association between the presence of P. quercina and other Phytophthora species and the decline status of oaks. It seems most likely that differences in sampling methodology are responsible for this discrepancy between the Bavarian and the French results. In both French studies, the number of trees and stands examined, as well as the size of the plots, was relatively small. On small plots, healthy and declining trees probably share more or less the same rhizosphere soil. Therefore, a similar Phytophthora population, regardless of the crown status, can be expected. In the present study, comprising a larger data base with 217 oaks in 35 oak stands, such neighbourhood effects were avoided by sampling trees with a scattered distribution in the stand instead of sampling all trees on relatively small defined plots. Also Hansen & Delatour (1999) and Delatour et al. (2000) stated that their sampling, which was aimed at evaluating the population of soilborne Phytophthora species, was inadequate for supporting firm conclusions about correlations between crown status and the Phytophthora population in the soil.
In general, the epidemiology of Phytophthora-induced fine root diseases is considered to be multicyclic (Erwin & Ribeiro, 1996). Phytophthoras can increase and disseminate their inoculum from low, nearly undetectable levels during a relatively short time of favourable environmental conditions. On the other hand, it can take decades of inoculum build-up and progressive fine root destruction before a mature tree begins to show noticeable above-ground symptoms (Tsao, 1990). Newhook (1988) pointed out the importance of latent infections and excess soil moisture, and mentioned that ‘the margin between outward health and disease can depend on the state of balance between rootlet death and rootlet replacement’. According to these statements it can be concluded that the presence of P. quercina in the rhizosphere of healthy oaks does not argue against its possible involvement in oak decline nor does it necessarily mean that these trees will fall into decline. However, it could be shown that oaks with P. quercina or other Phytophthora spp. in their rhizosphere displayed on average 20% higher crown transparency and significantly higher levels of fine root damage than oaks without Phytophthora. The statistical analysis of the isolation results revealed that oaks with P. quercina or other Phytophthora spp. in their rhizosphere have a significantly higher probability of exhibiting severe above-ground disease symptoms than oaks without Phytophthora spp.
These findings indicate that the high in vitro aggressiveness of at least P. quercina and P. cambivora as shown by soil infestation tests (Jung et al., 1996, 1999a, 1999b; Jung, 1998) might also exist under favourable environmental conditions in the field. In addition, significantly higher levels of fine root damage in declining than in healthy oaks were found in the Phytophthora-infested stands, resulting in significant correlations between crown transparency and all root parameters assessed. All root parameters with the exception of root rot were negatively correlated with crown transparency, confirming that the fine root status is of great importance for the overall crown status in Phytophthora-infested stands.
No Phytophthora species were recovered from soil samples of stands on sandy to sandy-loamy sites with a mean soil-pH ≤ 3·9. On these sites the correlation coefficients between crown transparency and all root parameters, as well as the differences between declining and healthy oaks for all root parameters assessed, were markedly smaller and less significant than in the Phytophthora-infested sites. In some stands, the health of the fine root system was even better in declining than in healthy oaks (data not shown), indicating that crown damage may have occurred before root damage.
The present findings are in accordance with main results of the study of Thomas & Hartmann (1998) who, in northern Germany, found clear differences in fine root density and biomass between healthy and damaged oaks on a clayey and hydromorphic site, but not on a sandy site. However, on another clayey site damaged oaks were not inferior to the healthy ones with respect to root density and biomass. In a later investigation on the occurrence of soilborne Phytophthora species in northern German oak stands, a strong association of P. quercina with declining oaks growing in disease pockets was found in the first two stands (T Jung & G Hartmann, unpublished data).
In a pH (CaCl2) range from 3·5 to 3·9, Phytophthora spp. were isolated from heavy but not from sandy soils. A possible explanation could be that, in sandy soils, even after heavy rain, the period with free water is too short to allow sufficient production of sporangia and release of viable zoospores at these suboptimal pH values. The isolation of P. quercina from sandy to silty-sandy soils in northern Germany with mean pH values > 4·3 (T Jung, unpublished data) indicates that the importance of soil texture for the occurrence of Phytophthora spp. decreases with increasing soil pH.
The transformation of the absolute values of the root parameters into season- and weather-corrected relative values resulted in markedly higher correlation coefficients between crown transparency and almost all fine root parameters, thus proving it to be a suitable method for comparing the fine root status of oaks from different stands. For the correlation analysis of crown transparency and fine root condition, and for the calculation of differences in fine root condition between healthy and declining oaks as well as between Phytophthora-infested and Phytophthora-free stands, the parameters root damage estimated followed by fine root length/mother root length, fine root length/dry weight mother roots, and fine root tips/mother root length were the most suitable.
Isolation of P. quercina from various nonhydromorphic sites and from sites with periodically fluctuating watertables shows that this pathogen is well adapted to temporary dry conditions, possibly due to its particularly thick oospore walls (Jung et al., 1999a). The observed level of fine root damage indicates that in temperate climates, even on nonhydromorphic sites, the period of water saturation in the soil after heavy rain or during the cool season is long enough to allow production of sporangia, discharge and dissemination of zoospores and subsequent infection of fine roots.
In general, phytophthora diseases are more severe at high soil pH values (Schmitthenner & Canaday, 1983). In the present study, soil pH was a limiting factor for the occurrence of Phytophthora spp.. No Phytophthora species could be isolated from soil samples with a pH (CaCl2) < 3·4 (< 4·2 measured in H2O). This finding is in accordance with the results of an in vitro test on sporangial formation in nonsterile soil filtrate with pH (H2O) values ranging from 3·5 to 6·0. Most Phytophthora species tested, including P. quercina, P. cambivora and P. citricola, failed to produce sporangia at pH values < 4·0, and sporangial formation increased rapidly with increasing pH. Exceptions were P. cinnamomi, which produced sporangia even at pH 3·5, and P. syringae and P. gonapodyides, which had optimum sporangial production at pH 5·0 (T. Jung, unpublished data). These results confirm the observations of Ribeiro (1983) who noted that sporangia may not form at pH values < 4·0. According to Ribeiro, this is probably due to the inability of Phytophthora oospores to germinate at low pH. In addition, T Jung, (unpublished data) observed an increasing proportion of hyphal lysis and sporangial abortion with decreasing pH. It can be derived from the liming experiments of Muchovej et al. (1980), working on pepper blight caused by P. capsici, that an increase in soluble aluminium with decreasing soil pH and calcium concentrations can also contribute to an inhibition of sporangial formation of Phytophthora, thus decreasing disease severity.
The concentration of exchangeable calcium in the soil was also an important factor for the occurrence of Phytophthora. Soils of Phytophthora-infested oak stands had significantly higher calcium concentrations than soils of Phytophthora-free stands (28·4 ± 37·4 vs. 5·5 ± 4·6 p.p.m.). This might be due to both a linkage of the calcium content with other soil properties such as pH and clay content, and the essential role of calcium for the asexual reproduction and infection process of Phytophthora. External calcium has a stimulatory effect on sporangial production (Ribeiro, 1983), and minimum levels are required for zoospore taxis (Carlile, 1983), adhesion on solid surfaces (Donaldson & Deacon, 1992) and cyst germination (Xu & Morris, 1998). Furthermore, it is known that cations such as calcium enhance oospore germination (Ribeiro, 1983). Although there are examples of high calcium suppressing root rot caused by Phytophthora spp., Phytophthora diseases are generally considered as being more severe at higher calcium concentrations (Schmitthenner & Canaday, 1983).
In Bavaria, as well as in other European countries, crown damage and mortality of oaks is markedly higher in older than in younger stands (Anonymous, 1999; Hansen & Delatour, 1999). This is possibly related to the observed decrease in plasticity and regeneration capacity of oak root systems with increasing age (Thomas & Hartmann, 1998; Becker & Lévy, 1982). It is proposed that, as a consequence, the balance is changing gradually from one of relative host tolerance to one of susceptibility to Phytophthora. Thus, under favourable environmental conditions, the infection rate of new fine roots by Phytophthora zoospores in mature oaks could exceed the rate of production of such roots, leading to a progressive destruction of the whole fine root system from year to year, and predisposing the weakened trees to secondary pathogens or droughts. The hypothesis that the susceptibility of the fine root system of oaks to Phytophthora is increasing with the age of the trees, resulting in a gradual inoculum build-up of Phytophthora, is supported by the isolation results in stand 29. In this stand, decline of oaks was restricted to mature trees, all of which were affected by P. quercina. In contrast, 25-year-old oaks, growing under the same site conditions in a part of the stand where the mature oaks had been cut 20 years previously, were healthy, and no Phytophthora was recovered (data not shown).
In some recent publications (Hartmann & Blank, 1998; Lobinger, 1999; Delb & Block, 1999) leaf defoliators have been considered as causal agents of rapid mortality in German oak stands. In the present study, the relationship between defoliation and presence or absence of Phytophthora spp. in the soil was investigated in stand 14 near Würzburg. In 1994 and 1995, the 45-to 50-year-old stand was completely defoliated by caterpillars of Tortrix viridana, Lymantria dispar and Thaumatopoea processionea. Subsequently, the oaks growing on the upper slope on a moderately fresh orthic luvisol recovered totally, whereas in July 1996 a dramatic mortality of about 75% of the oaks started at the bottom of the slope and in the depression on a stagno-gleyic luvisol with periodically fluctuating watertables. The absence of any Phytophthora species in the rhizosphere of oaks on the upper slope and a Phytophthora infestation at the bottom of the slope and in the depression indicate a synergistic detrimental effect of site conditions, fine root destruction by Phytophthora spp. and defoliation events. However, additional investigations are needed before general conclusions can be drawn.
Schütt (1993) and Schlag (1994) mentioned that oak decline is a complex of different diseases. This hypothesis was confirmed by the results of the present study. From the data it can be concluded that there are at least two different complex diseases being referred to under the name ‘oak decline’ in Central Europe: (i) Phytophthora-mediated oak decline and (ii) decline mediated by droughts, defoliations or miscellaneous causes.
Phytophthora-mediated oak decline occurs on sites with mean soil pH (CaCl2) values ≥ 3·5 and sandy-loamy to loamy, silty or clayey soil texture where soilborne Phytophthora species are strongly involved in the aetiology of oak decline by causing a fine root disease with highly significant correlations between crown transparency and most fine root parameters. Among all Phytophthora species isolated, P. quercina plays the major role because of its widespread occurrence, its high plasticity concerning site conditions and soil pH, and its host specificity and high aggressiveness to oak species (Jung et al., 1996, 1999a, 1999b; Jung, 1998). Because of the multicyclic nature of Phytophthora fine root diseases (Erwin & Ribeiro, 1996) and the strong dependence of the infection process on the presence of free water in the soil, it may be some decades from the point where rootlet death begins to exceed rootlet replacement to the appearance of first visible crown symptoms in mature oaks. From this point onwards the progressive destruction of the fine root system can cause a chronic dieback of the crown, weakening the oaks and predisposing them to a series of abiotic and biotic stress factors.
Oak decline mediated by droughts, defoliations or miscellaneous causes occurs on sites with a mean soil pH (CaCl2) ≤ 3·9 and sandy to sandy-loamy soil texture and is generally a chronic disease without consistent correlations between crown and root status. On these sites, Phytophthora species have not been isolated and therefore cannot be considered as being involved in the decline process. The hypothesis is put forward that, on these sites, defoliations (defoliation-mediated oak decline), droughts (drought-mediated oak decline) or miscellaneous agents such as Collybia fusipes (Marçais & Delatour, 1996; Marçais et al., 1998) or extreme winter frost and late winter frost (Hartmann et al., 1989) are weakening the trees and predisposing them to attacks by secondary pathogens (e.g. Armillaria mellea) or parasites (e.g. Agrilus biguttatus). On these sites, as on Phytophthora-infested sites, repeated defoliations or prolonged droughts can lead to rapid mortality, mainly because of the low water retention capacity of the sandy soils. In these cases mortality often starts on, or is even restricted to, the shallowest and sandy parts of the stand or the upper slope.
All these biotic and abiotic stress factors often act synergistically, and can also accelerate Phytophthora-mediated decline of oaks, as described above.
The distribution of the disease in the field is often a good indication of the type of aetiology, but can also be misleading. For instance, decline and mortality in groups of trees can be indicative of both a Phytophthora-related disease beginning with inoculum build-up around the most susceptible trees, resulting in a heavier attack on adjacent roots of neighbouring trees and build-up of disease pockets, or of drought stress on shallow and sandy microsites. Another example is the slow scattered decline of oaks that is characteristic of damage by Collybia fusipes (B Marçais, personal communication), but also of Phytophthora-mediated decline in stands with small-scale changes between sandy and clayey soils (e.g. stands 20 and 22 in this study).
According to Manion (1981) and Manion & Lachance, 1992) a decline is a gradual, general loss of vigour caused by the interaction of a number of interchangeable, specifically ordered biotic and abiotic factors, which often ends in the death of trees. Following this definition, Central European oak decline is an excellent example of a decline-type disease of varied aetiology
It should be emphasized that there may certainly be stands or even sites where dieback and mortality of oaks is not complex, but can be explained simply by the action of only one biotic (defoliation, Phytophthora on permanently or periodically wet, moderately acid to neutral soils, Collybia on sandy acid soils) or abiotic factor (drought on sandy shallow soils, extreme winter frost, late winter frost). By definition these single-cause diseases should not be called oak decline. On the other hand, even in these situations, transitions to more complex interactions of different factors exist, which are most often due to changing microsite conditions within single stands.
The isolation of Phytophthora species from rhizosphere soil samples of 88 out of 129 oak stands (68·2%) in 11 European countries (Germany, France, England, Scotland, Sweden, Luxembourg, Poland, Switzerland, Italy, Slovenia and Hungary), and similar relationships between the occurrence of Phytophthora spp. and soil pH and texture (T. Jung, unpublished data), indicate that the conclusions from the present study may have general validity for Central and Western Europe as well as for parts of Southern and Northern Europe. Furthermore, Iberian oak decline is also considered to be caused by an interaction of fine root damage by Phytophthora (here P. cinnamomi), climatic perturbations such as prolonged droughts and episodes of unseasonal heavy rain and flooding, site factors, and secondary pathogens or parasites (Brasier, 1993, 1996).
Considering the epidemic extent and the long duration of oak decline in Central and Western Europe in its current phase, the question arises of whether any environmental changes have occurred during recent decades that have unbalanced the host–parasite relationship between the oaks and the naturally occurring soilborne Phytophthora species. Excess anthropogenic nitrogen input into forest soils (Nihlgard, 1985; Kreutzer, 1991; Mohr, 1994; Thomas & Kiehne, 1995), the rise in mean winter temperature of 0·03°C per year, and a seasonal shift of precipitation from summer to winter in Central Europe and parts of Western and Northern Europe (Rapp & Schönwiese, 1995; Schönwiese et al., 1994) have already been discussed as potential triggering factors by Jung et al. (1996, 2000) and Jung (1998). Results of two recent studies further support these hypotheses: sporangial production is enhanced with increasing concentrations of nitrate in soil filtrates for P. quercina and five other Phytophthora species (T. Jung, unpublished data), and in Bavaria after analysing all forest monitoring data between 1984 and 1997, Mayer (1999) found that the probability of oaks showing severe crown symptoms (crown transparency > 25%) is 1·6 times higher following wet winters with average temperatures more than 2°C above the long-term average.
We are grateful to the Bavarian State Ministry for Food, Agriculture and Forestry and the Allianz Stiftung zum Schutz der Umwelt for their generous financial support, and to the Bavarian Foresters involved in this study for their kind help. We also want to thank Professor Karl-Eugen Refuess and the technicians Ariane Schaub and Michael Englschall from the Institute of Soil Sciences of the Technische Universität München for the soil analysis, and Dr Jan Nechwatal and the technicians Johanna Lebherz and Renate Bernhard for their assistance in the laboratory routines. Furthermore, Professor Clive Brasier and both reviewers are acknowledged for their valuable comments and suggestions, which greatly improved the paper.