Relationships between health of Quercus robur, occurrence of Phytophthora species and site conditions in southern Sweden
*To whom correspondence should be addressed.
The effect of Phytophthora species, soil chemistry, precipitation and temperature on the vitality of oak was evaluated in 32 oak stands in southern Sweden. In addition, the relationship between the occurrence of Phytophthora species and soil conditions was determined. The results showed that there was a weak association between the presence of P. quercina, the most frequently recovered Phytophthora species in southern Sweden, and the vitality of the oak stands (determined from estimates of crown defoliation of individual trees). The pathogens occurred more frequently in clayey and loamy soils that were less acidic and which had higher base saturation. However, they were found in all but the most acidic soils (pH < 3·5). In stands where Phytophthora species were not present, positive correlations between the average crown defoliation and proportion of damaged trees with average summer precipitation and average annual precipitation were found. There were no significant differences in soil chemistry between healthy and declining stands included in this study, and no significant correlations were found between any soil parameter and crown vitality. Based on the results from these 32 oak stands, it is likely that the decline of oaks in southern Sweden can be attributed to several different site-specific factors, such as infection by P. quercina or unusual weather events, which interact with a number of biotic and abiotic factors, leading to oak decline.
Several studies performed during the past two decades have shown an extensive decline of Quercus species throughout Europe and North America (Tainter et al., 1983; Hartmann et al., 1989; Luisi et al., 1993; Anonymous, 2000; Sonesson & Anderson, 2001; Rizzo et al., 2002). The decline is characterized by a variety of symptoms, including thinning of the crown, discoloration or yellowing of the leaves, formation of leaf clusters, dieback of branches, formation of epicormic shoots and bark lesions. The low specificity of these symptoms, together with the many single causal factors proposed to be responsible for the decline in various parts of Europe (Hartmann et al., 1989; Oosterbaan & Nabuurs, 1991; Hartmann & Blank, 1992; Brasier et al., 1993; Siwecki & Ulfnarski, 1998; Thomas & Hartmann, 1998; Jung et al., 2000), suggest that oak decline can be attributed to several different site-specific factors and/or that the decline may involve a succession of several abiotic and biotic factors (Manion, 1991; Thomas et al., 2002).
Of the biotic factors suggested to be involved in oak decline, the aggressive primary plant pathogens residing in Phytophthora have received most attention. Phytophthora species have been found to be widespread in European, as well as in Turkish oak stands (Brasier et al., 1993; Robin et al., 1998; Hansen & Delatour, 1999; Jung et al., 2000; Vettraino et al., 2002; Balcì & Halmschlager, 2003a,b; Jönsson et al., 2003a) and significant negative effects of these pathogens on the fine-root systems of oak seedlings have been shown in many glasshouse studies performed under controlled conditions (Jung et al., 1996, 1999, 2003; Robin et al., 1998; Gallego et al., 1999; Jönsson et al., 2003b; Jönsson, 2004a). Furthermore, significant correlations have been found between the presence of Phytophthora species in the rhizosphere, the condition of the fine roots (investigated only in Germany) and crown defoliation of mature oaks in Germany, Italy, Austria and Turkey (Jung et al., 2000; Vettraino et al., 2002; Balcì & Halmschlager, 2003a,b). However, Phytophthora may be present in oak stands without being associated with the decline of the trees. In France, no evidence was found of an association between the presence of Phytophthora species and tree health (Robin et al., 1998; Hansen & Delatour, 1999; Camy et al., 2003).
Of the abiotic factors discussed in connection with European oak decline, site quality has been emphasized as being important for tree vigour and growth (Standovar & Somogyi, 1998). Air pollutants, on the other hand, are generally not believed to contribute to oak decline (Thomas et al., 2002). A number of studies on leaf and soil nutrient concentrations in oak stands support the suggestion that changes in the nutritional status of soils and trees, as a consequence of an increased input of N and acidifying compounds, is not a causal factor behind oak decline, at least not in central Europe (Berger & Glatzel, 1994; Thomas & Kiehne, 1995; Simon & Wild, 1998). However, chemical changes in the soil and in the trees may eventually destabilize the oak ecosystems, thereby reducing the stress tolerance of the trees (Thomas et al., 2002). It has also been suggested that extreme climatic conditions are involved in oak decline. Changes in precipitation patterns and droughts seem to be frequently associated with aggravated decline (Oosterbaan & Nabuurs, 1991; Siwecki & Ulfnarski, 1998; Thomas & Hartmann, 1998), and exceptionally cold winters have also been found to have major damaging effects on oak trees, with large areas of bark being killed (Hartmann et al., 1989; Hartmann & Blank, 1992). Droughts, and in particular fluctuations between dry and wet periods, may interact with biotic agents such as Phytophthora spp., thus increasing disease severity (Jung et al., 2003).
In southern Sweden, oak decline has been widespread since the late 1980s, and in 1999 the proportion of damaged trees (crown defoliation > 25%) reached a level of 59% (Sonesson & Anderson, 2001). Few attempts have been made to clarify the reasons behind the decline, but recently, three different Phytophthora species were recovered from 11 out of 32 investigated stands in the southernmost part of the country (Jönsson et al., 2003a). The most frequently recovered species, P. quercina, was found to have a negative impact on the root systems of seedlings grown in acid forest soils under controlled conditions as well as on root systems of mature trees (Jönsson, 2004b). The objectives of this study were therefore to investigate whether there is an association between the presence of Phytophthora species in oak stands in southern Sweden and the vitality of the stands, and to determine in what kind of soils these pathogens occur. The involvement of other factors that have been discussed in connection with European oak decline were also evaluated, namely soil nutrient status and climate.
The following hypotheses were tested: (i) Phytophthora species are more likely to occur in declining than in healthy oak stands; (ii) soils in which Phytophthora species occur are less acidic and more nutrient-rich than soils without the pathogen; (iii) soils associated with declining stands have lower base saturation and higher acidity than soils associated with healthy stands; and (iv) lower amounts of summer precipitation and lower winter temperatures are associated with higher average crown defoliation and a higher proportion of damaged trees in the stands.
Materials and methods
The 32 oak (Quercus robur) stands included in this study have been previously investigated with regard to the occurrence of Phytophthora species (Jönsson et al., 2003a). They are located at 55·3–59·4° latitude, on the border between the northern nemoral vegetation zone and the southern boreal zone. A more detailed description of the location of these sites can be found in Jönsson et al. (2003a). Most of the stands are monitoring sites for forest condition of oak, administered by the Regional Forestry Board Södra Götaland and the Regional Forestry Board Östra Götaland. The size of the survey plots varies between 30 × 30 m and 80 × 80 m (Sonesson & Anderson, 2001). The mean annual temperature and mean annual precipitation in the area studied ranged from 5·4 to 8·8°C and from 545 to 837 mm, respectively, between 1984 and 1999 (SMHI, 1983–2000). Information on site and stand characteristics, and a summary of the occurrence of Phytophthora species are given in Table 1. The sites used in this study have predominantly silty to clayey soil texture, as these soil types are where Phytophthora species are expected to be found (Jung et al., 2000).
Table 1. Site and stand characteristics of investigated oak stands in southern Sweden and occurrence of Phytophthora species. For definition of regions and stand vitality, see ‘Materials and Methods’
|1||A||D|| 95||F||M||Loam||Mesic||3·74||x|| || |
|2||A||D||105||Q||M||Silt||Mesic||3·52|| || || |
|3||A||D|| 90||F||M||Silt||Mesic||3·55|| || || |
|4||A||D|| 75||Q||M||Loam||Mesic||3·96||x|| || |
|5||A||D|| 80||Q||M||Sand||Mesic||3·76|| || || |
|6||A||D|| 60||F, D||M||Sand||Moist||3·19|| || || |
|7||A||D|| 65||F, D, C||M||Silt||Mesic||3·33|| || || |
|8||A||D|| 90||F||M||Silt||Mesic||3·90||x|| || |
|9||A||D|| 65||D||S||Clay||Mesic||3·53||x|| || |
|10||A||D|| 90||F, D||M||Silt||Moist||3·64|| || || |
|11||A||H|| 80||Q||S||Clayey loam||Mesic||3·50|| || || |
|12||A||H||110||F||M||Loam||Mesic||3·75|| || || |
|13||A||D|| 80||F||M||Silt||Mesic||3·42|| || || |
|14||A||H|| 80||F, D||M||Silt||Mesic||3·29|| || || |
|15||A||D||100||Q||S||Clay||Mesic||3·50|| || || |
|16||B||D|| 70||F||M||Silt||Mesic||3·53|| || || |
|17||B||D|| 70||F, D||M||Silt||Moist||3·52|| || || |
|18||B||D||100||F||S||Loam||Mesic||3·87||x|| || |
|19||B||D|| 75||F||M||Loam||Mesic||3·35|| || || |
|20||A||H||110||F, D||M||Silt||Mesic||3·57|| || || |
|21||A||D|| 65||F, D||M||Silt||Mesic||3·70|| || || |
|22||B||D||160||Q||M||Sandy silt||Mesic||3·70|| || || |
|23||B||H||170||Q||M||Silt||Mesic||3·75|| || ||x|
|24||B||D|| 90||F||M||Silt||Mesic||3·55||x||x|| |
|25||B||D|| 60||D||M||Silt||Mesic||4·97||x|| || |
|26||B||D|| 90||D||S||Loam||Mesic||3·85||x|| || |
|27||B||H|| 60||Q||M||Silt||Mesic||3·80||x|| || |
|28||B||H|| 60||Q||M||Silt||Mesic||3·68|| || || |
|29||B||H|| 65||D||S||Loam||Mesic||5·18|| || || |
|30||B||D||110||Q||S||Loam||Mesic||4·67||x|| || |
|31||B||D|| 70||Q||M||Silt||Mesic||5·90|| || || |
|32||B||H||110||Q||M||Silt||Mesic||3·68|| || || |
Soil sampling for detection of Phytophthora species and method of isolation
Soil from the rhizosphere of four to eight trees at each site was sampled to investigate the occurrence of Phytophthora species (Jönsson et al., 2003a). Three to five soil monoliths with a size of 20 × 30 cm were taken around each tree, at a distance of 30–150 cm from the stem base and at a soil depth of 10–30 cm. Aliquots of rhizosphere soil together with oak roots (diameter 5 mm) from all monoliths were bulked, and subsamples were used for isolation tests. Isolations were carried out using the soil baiting method described by Jung et al. (1996, 2000). Two- to 7-day-old leaflets of Q. robur seedlings were floated over flooded soil and fine roots at 17°C. Leaflets with necrotic spots, which usually appeared after 3–12 days, were blotted dry, cut into pieces, and transferred to selective PARPNH medium. All PARPNH plates were incubated at 20°C in the dark, and examined daily under a stereomicroscope. Developing Phytophthora and Pythium colonies were transferred onto V8 agar (V8A) and malt-extract agar (MEA), incubated for 1–2 weeks at 20°C, and then kept at 5°C.
Following description of the occurrence of Phytophthora species in Jönsson et al. (2003a), additional isolation attempts on new soil samples were carried out at the sites which had only been sampled once and from which Phytophthora species had not been recovered (stands 28, 29, 31 and 32). Including all sites and sampling occasions, sampling was thus performed from October 1999 to November 2002. Sampling was performed during different seasons: early spring, summer and/or late autumn. The pH(BaCl2) in the rhizosphere soil was determined (Jönsson et al., 2003a).
Soil sampling for chemical analysis
In addition to soil sampled for the isolation of Phytophthora species, soil from each of the investigated oak stands was also sampled for chemical analysis. At most sites, sampling was performed by the Regional Forestry Board Södra Götaland and the Regional Forestry Board Östra Götaland. The results of the chemical analyses have kindly been provided for use in this study. One composite sample per stand (= survey plot) was analysed. The composite sample consisted of eight to nine subsamples, which were collected 0·5–1·0 m from the trunks of one to eight trees per cardinal point of the survey plot (depending on how many cardinal points with sample trees were available). The soil was sampled at a depth of 20–30 cm in the mineral soil using a 32-mm-diameter auger. For further information about the soil sampling, see Sonesson & Anderson (2001). At sites 21–25, soil was sampled for chemical analysis only from the five trees from which soil was sampled for the determination of Phytophthora species. At sites 1–20, sampling was performed during July and August 1999, and at sites 21–32 during November and December 2000. At each site, soil sampling was performed within 1 year of sampling the soil for Phytophthora species. Before chemical analysis, the soil was sieved through a 2-mm mesh to exclude roots and large particles. All samples were dried at 40°C to constant weight.
Chemical analysis of soil
Twenty grams of soil were extracted in 100 mL 0·1 m BaCl2 for 2 h (Anonymous, 1998). Extraction took place at room temperature (20°C) and the pH was then measured in the BaCl2 filtrate. Aluminium (Al) concentrations as well as concentrations of base cations (Ca, Mg, K, Na), manganese (Mn), iron (Fe) and boron (B) were determined with an inductively coupled plasma analyser (Perkin Elmer, USA).
For soils collected at sites 1–25, carbon (C) concentrations were determined using an automatic carbon elementar analyser (CR12, LECO Corporation, Michigan, USA), while total nitrogen (N) was analysed using the Kjeldahl technique (Anonymous, 1998). For soil from the other sites, C and N were determined using an element analyser (CHN600, LECO Corporation). The results obtained from the chemical analyses were normalized to the dry matter content at 85°C. The total exchangeable acidity, the cation exchange capacity and the base saturation were calculated for each site.
The crown defoliation of the oak trees was determined at each site, except for sites 24 and 25, by staff from the Regional Forestry Board Södra Götaland and the Regional Forestry Board Östra Götaland. At sites 1–23, the inventory was performed during July and August 1999, and at sites 26–32 during September and October 2000. At sites 24 and 25, crown defoliation was estimated by the authors. Crown defoliation was assessed on the uppermost two-thirds of the crown and was expressed as a percentage of a normal ample crown. The estimated defoliation included leaf loss, branch loss and dying branches. Normal self-pruning by self-shading was not included. Defoliation was expressed in 1% classes from 0 to 100. For further information about these crown assessments, see Sonesson & Anderson (2001).
The impact of Phytophthora species and site conditions on the decline of oaks was evaluated using three different crown parameters: the average crown defoliation of each stand, the standard error of the average crown defoliation and the proportion of damaged trees in each stand. A tree was considered damaged if the crown defoliation exceeded 25% (Anonymous, 2000). The number of trees with crown defoliation between 25 and 60% and greater than 60% was also determined in each stand. In addition, each stand was classified as healthy or in decline. This classification was based on the average crown defoliation and the proportion of damaged trees in each stand; if the average crown defoliation exceeded 25% and at least one-third of the trees were damaged (crown defoliation > 25%), the stand was considered to be declining.
Average summer precipitation (April to October) and average annual precipitation for the period 1984–99 were calculated together with the average winter temperature (November to March) and the average annual temperature, using data from the meteorological station closest to each site (SMHI, 1983–2000). The oak stands were situated between 5 and 55 km (although few more than 30 km) from their respective meteorological station. Measurements are therefore only rough estimates of the conditions at the sites. The close geographical location of some of the stands means that meteorological data from the same station were used for several of the sites. In total, data from 14 meteorological stations were included.
Contingency tables were used to evaluate the importance of geological substrate and soil texture for the occurrence of Phytophthora spp. and vitality of the stands, as well as to evaluate the association between presence of the pathogens and stand vitality. Based on the results from the soil investigations in this study, and those in Germany and Austria (Jung et al., 2000; Balcì & Halmschlager, 2003a), two criteria for the occurrence of Phytophthora species in southern Sweden were set: Phytophthora species occur only on sites with clayey, sandy-loamy to loamy and silty soil textures where the pH(BaCl2) of the rhizosphere soil exceeds 3·5, and on sandy sites with pH(BaCl2) > 3·9. Sites with other soil conditions usually do not host any Phytophthora species, probably because these conditions are not suitable for the survival and reproduction of the pathogen. In the contingency analysis of the relation between stand vitality and presence of Phytophthora species, only sites that fulfilled the above criteria were used. This means that the number of stands was reduced by seven, giving a total of 25 stands (Table 1). To test for significant differences in the average crown defoliation, the standard error of the average crown defoliation and the proportion of damaged trees at sites with and without Phytophthora species, a t-test was used.
Chemical soil data, precipitation and temperature were not normally distributed and differences between healthy and declining stands were therefore analysed using the nonparametric Mann–Whitney test. Each stand was also classified into one of two groups with regard to its pH, Al and base saturation. These groups were formed based on minimum and maximum values of chemical parameters in soil, proposed for the exchangeable phase at 20–30 cm depth in a productive, sustainably managed southern Swedish boreal coniferous or mixed forest soil (Stjernquist et al., 2002, Table 5). The difference in crown defoliation, standard error of crown defoliation and the percentage of damaged trees between the groups was evaluated statistically with a t-test. The association between the average crown defoliation of each stand and the proportion of damaged trees and different soil chemical variables (Ca, K, Mg, H, Al, B, N, C/N, total exchangeable acidity, base saturation), precipitation and temperature were tested using Spearman's rank correlation. The total number of correlations performed with regard to chemical soil parameters and climatic factors was 120 and 16, respectively.
Table 5. Average crown defoliation, standard error of the average crown defoliation and the proportion of damaged trees (crown defoliation > 25%) in oak stands for different soil chemical classes (see ‘Materials and Methods’ for further description). Values given are means ± SE. There were no significant differences in any of the crown parameters between the classes (Mann–Whitney test)
| ≤ 200 µg g−1||21||34·3 ± 3·0||5·0 ± 0·5||59·1 ± 6·8|
| > 200 µg g−1||11||25·8 ± 3·1||3·5 ± 0·4||43·3 ± 8·6|
| ≤ 20%||14||29·7 ± 3·3||4·4 ± 0·6||51·9 ± 8·0|
| > 20%||18||32·8 ± 3·2||4·5 ± 0·5||55·1 ± 7·5|
| ≤ 4·2||22||32·6 ± 2·9||4·7 ± 0·4||59·5 ± 6·3|
| > 4·2||10||28·9 ± 3·9||4·0 ± 0·8||40·9 ± 9·6|
In addition to evaluating all 32 sites together, some of the tests were also performed after excluding sites infested with Phytophthora and after separating the sites into two geographic regions. Region A included sites 1–15, 20 and 21, and consisted of sites in the southwestern part of Sweden (the provinces of Skåne and Halland), while region B included sites 16–19 and 22–32, which are situated in the southeastern part of the country (the provinces of Blekinge and Småland). The division into two regions was performed to evaluate possible causal factors depending on the region. In region B, two sites had base cation and one site had N concentrations which differed substantially from the concentrations at all the other sites. These sites were excluded when performing the correlation analysis, and are defined as outliers in the text. All statistical calculations were performed using the software SPSS 10 for Macintosh (SPSS Inc., USA).
Soil chemistry and site conditions in relation to the occurrence of Phytophthora species
The additional sampling did not give any positive isolations of Phytophthora species, and the number of stands from which the pathogens have been recovered remain unchanged from Jönsson et al. (2003a). Oak stands with Phytophthora spp. had a significantly higher pH(BaCl2) in the rhizosphere soil than stands without Phytophthora (Table 2). However, there was no significant difference in the overall pH in the soil. The occurrence of all Phytophthora species was independent of geological substrate (Fisher's exact test, P = 0·197) and texture (Fisher's exact test, P = 0·123), but a weak association was found between the occurrence of P. quercina and soil texture (Fisher's exact test, P = 0·056). The pathogen seemed to be more strongly associated with clayey and loamy soils than with silty and sandy soils (odds of the soil texture being clay and loam where P. quercina is present = 2·864; odds of the texture being clay and loam where P. quercina is absent = 0·561). Accordingly, stands infested with Phytophthora were generally more nutrient-rich than those without the pathogen. Concentrations of Ca and B and the base saturation were significantly higher in infested stands and the concentration of K showed a tendency to be somewhat higher, although the difference was not significant (P = 0·067; Table 2). For the acid ions, results were contradictory, with lower concentration of Fe but higher concentration of Mn in infested stands. Concentration of Al and total exchangeable acidity showed no difference between infested and noninfested stands (Table 2). There were no differences in average summer precipitation or average annual precipitation between Phytophthora-infested and noninfested stands (data not shown).
Table 2. Concentration of chemical elements at 20–30 cm depth in the mineral soil of oak stands infested with Phytophthora spp. (n = 11) and noninfested stands (n = 21). Values given are mean ± SE, except for pH where values are medians, and numbers within parentheses are minimum and maximum values. Statistics are for the Mann–Whitney test: * P ≤ 0·05, ** P ≤ 0·01, *** P ≤ 0·001 and ns, not significant
|Stands infested with Phytophthora spp.||4·1||3·9||1024·9||50·7||59·2||170·1||6·8||44·9||0·29||21·4||50·5||1·9|
|(3·5–6·5)||(3·5–5·0)||(± 481·5)||(± 13·3)||(± 19·8)||(± 37·3)||(± 2·3)||(± 14·7)||(± 0·08)||(± 4·6)||(± 9·9)||(± 0·5)|
|Stands without Phytophthora spp.||4·0||3·6||522·42||27·1||14·7||198·5||15·3||11·9||0·05||23·8||26·0||1·5|
|(3·7–6·8)||(3·2–5·9)||(± 333·5)||(± 2·8)||(± 2·2)||(± 25·5)||(± 3·4)2||(± 2·7)2||(± 0·01)2||(± 3·0)||(± 5·8)2||(± 0·1)|
Crown condition in relation to the occurrence of Phytophthora species
The contingency analysis showed no relationship between the vitality of the stands and the occurrence of Phytophthora species (Table 3). However, when only sites with P. quercina, the most frequently occurring Phytophthora species in southern Sweden, were included in the analysis, a weak association, although not significant (P = 0·088), was found between decline of the stand and presence of the pathogen. The odds ratio indicated that a stand is more likely to be declining if P. quercina is present in the stand (Table 3). There were no significant differences in average crown defoliation, standard error of crown defoliation or the proportion of damaged trees between stands infested with Phytophthora and stands without the pathogen (data not shown). Considering only P. quercina gave similar results.
Table 3. Occurrence of Phytophthora species and P. quercina in relation to oak stand vitality (n = 25). Statistics given are for contingency analysis. Only stands that were found suitable for Phytophthora species based on soil texture and pH(BaCl2) are included. For definitions of healthy and declining stands, see ‘Material and Methods’
| Declining stands|| 9|| 8||17|
| Healthy stands|| 2|| 6|| 8|
| Declining stands|| 9|| 8||17|
| Healthy stands|| 1|| 7|| 8|
Crown condition in relation to soil chemistry and site conditions
There were no significant differences in any chemical soil parameter between healthy and declining stands (Table 4). Separating the stands geographically or excluding sites infested with Phytophthora species did not change the results (data not shown). Division of sites into different chemical soil classes (low and high pH, Al concentration and base saturation) gave no significant differences in average crown defoliation, standard error of crown defoliation or the percentage of damaged trees (Table 5), although the average crown defoliation and standard error of crown defoliation tended to be higher in sites where the concentration of Al was below 200 µg g−1 (P = 0·079 for the average crown defoliation and P = 0·061 for standard error of crown defoliation).
Table 4. Concentration of chemical elements at 20–30 cm depth in the mineral soil for healthy (n = 9) and declining (n = 23) oak stands. Values given are mean ± SE, except for pH where values are medians, and numbers within parentheses are minimum and maximum values. There were no significant differences between healthy and declining stands (Mann–Whitney test)
|(3·7–5·5)||(3·4–5·9)||(± 297·2)||(± 4·3)||(± 18·4)||(± 55·2)||(± 7·1)||(± 13·9)||(± 0·04)||(± 6·6)||(± 11·2)||(± 0·1)|
|(3·5–6·8)||(3·2–5·0)||(± 365·0)||(± 7·0)||(± 8·3)||(± 19·0)||(± 1·9)||(± 6·4)||(± 0·04)||(± 2·3)||(± 6·3)||(± 0·2)|
Furthermore, no significant correlations were found between any of the chemical soil parameters (Ca, K, Mg, Al, H, B, N, C/N, total exchangeable acidity and base saturation) and average crown defoliation, standard error of crown defoliation or the proportion of damaged trees (values of ρ < 0·37). Excluding the sites from which Phytophthora species were recovered gave similar results (values of ρ < 0·33). When separating the sites geographically, some weak correlations were found in both regions. In region A, a weak negative correlation was found between Mg and the proportion of damaged trees, when excluding the outliers (ρ = −0·45, P = 0·093, n = 15). In region B, significant positive correlations were found between the three indicators of crown condition and the concentrations of K, Mg and N, but after removing the outliers, the values of ρ decreased considerably (ρ < 0·43). When considering only sites without Phytophthora species, weak correlations were found between the average crown defoliation and the concentrations of K and H in region A (K: ρ = −0·50, P = 0·083, n = 13; H: ρ = −0·54, P = 0·055, n = 13). In region B, concentrations of H and N seemed to be positively correlated with average crown defoliation (H: ρ = 0·48, P = 0·233, n = 8; N: ρ = 0·61, P = 0·148, n = 7) and, for H, also with the proportion of damaged trees (ρ = 0·54, P = 0·168, n = 8).
No association was seen between the vitality of the stands and geological substrate and soil texture (data not shown). Furthermore, there was no significant difference in average crown defoliation, standard error of crown defoliation or the percentage of damaged trees between stands growing on moraine or sediment (data not shown). With regard to soil texture, sites with silt or sand had significantly higher average crown defoliation and standard error than sites with loamy or clayey soils, although the average difference was less than 10% (t-test, P = 0·025; average crown defoliation for clay and loam, 25·4 ± 2·3%; and for silt and sand, 34·6 ± 3·1%).
There were no differences between healthy and declining stands with regard to average summer precipitation, average annual precipitation, average winter temperature or average annual temperature for the period of 1984–99 (data not shown). Nor were there any significant correlations between the average crown defoliation and the proportion of damaged trees and these climatic parameters when all sites were included (ρ < 0·37; data not shown). However, when the Phytophthora-infested sites were excluded, weak positive correlations were found between average summer precipitation and average crown defoliation (ρ = 0·49, P = 0·026, n = 21), average summer precipitation and the proportion of damaged trees (ρ = 0·51, P = 0·018, n = 21), average annual precipitation and average crown defoliation (ρ = 0·50, P = 0·022, n = 21), and average annual precipitation and the proportion of damaged trees (ρ = 0·52, P = 0·016, n = 21). There were no significant correlations between temperature and the crown parameters (ρ < 0·03).
The results presented in this study show that there is a weak association between the occurrence of the root pathogen P. quercina in the soil and the vitality of oak stands in southern Sweden, and that oak stands with P. quercina in the soil are more likely to be declining than stands without the pathogen (Table 3). It is known that the southern Swedish isolates of this pathogen can cause damage to root systems of oak seedlings grown in acid forest soils under controlled conditions (Jönsson et al., 2003b; Jönsson, 2004a), and this pathogen has also been demonstrated to be very aggressive to oak seedlings grown in artificial soil mixtures (Jung et al., 1996, 1999, 2003; Jönsson et al., 2003b). Furthermore, studies of mature oaks growing in Phytophthora-infested stands and in noninfested stands in Germany (Jung et al., 2000) and southern Sweden (Jönsson, 2004b) indicate that P. quercina can also cause substantial damage to root systems of mature trees. It is therefore reasonable to believe that the symptoms of decline in trees in Phytophthora-infested oak stands in southern Sweden are a consequence of reduced root vitality due to pathogen attack by P. quercina, with the disease process probably being promoted by interactions between the pathogen and other biotic and abiotic factors. Strong associations between the presence of P. quercina in the rhizosphere soil and decline of individual trees have previously been demonstrated in Germany and Italy (Jung et al., 2000; Vettraino et al., 2002).
The soil conditions in the oak stands from which Phytophthora species were recovered ranged from mesic sediments to moraines, with clayey, loamy and silty textures and soil pH(BaCl2) in the rhizosphere between 3·5 and 5·0 (Jönsson et al., 2003a; Table 1). Phytophthora quercina was more frequently associated with clayey and loamy soils than with silty and sandy soils. This is consistent with the survey in Germany by Jung et al. (2000). However, Jung et al. (2000) suggested that Phytophthora species could be isolated from sandy and silty-sandy soils if the chemical soil conditions are favourable, mainly with regard to pH. In southern Sweden, few oak stands have pH(BaCl2) values that exceed 4·2 (Sonesson & Anderson, 2001). Consistent with previous investigations in Germany and Austria (Jung et al., 2000; Balcì & Halmschlager, 2003a), the Phytophthora species in southern Sweden were found at more nutrient-rich sites, primarily with regard to Ca, and where the proportion of base cations in relation to acid ions was high. However, it was shown by Jung et al. (2000) and has been shown in the present study that the pathogens can survive at all but the most acidic sites, i.e. sites with pH(BaCl2/CaCl2) < 3·5. Phytophthora diseases are generally considered to be less severe at these low pH values (Schmitthenner & Canaday, 1983). In contrast, a pathogenicity test with oak seedlings showed that P. quercina caused more damage to root systems of seedlings grown in an acid forest soil than to seedlings grown in a more nutrient-rich artificial soil with a higher pH (Jönsson et al., 2003b). The root growth of seedlings in the acid forest soil was apparently negatively influenced by the chemical and physical conditions in the soil and this probably predisposed them to the damage caused by P. quercina, possibly through the influence of soil chemistry on the carbon allocation patterns within the seedlings (Jönsson, 2004b). Soil moisture seems to be less important for the occurrence of Phytophthora species, at least for P. quercina. This is due to its thick-walled oospores, making it well adapted to survival in dry soil conditions (Hansen & Delatour, 1999; Jung et al., 2000; Balcì & Halmschlager, 2003a,b).
Infestation by Phytophthora species is obviously not the only cause associated with oak decline in southern Sweden. Many of the stands where Phytophthora species were not found also show severe symptoms of crown damage. The situation in Germany and France is similar (Hansen & Delatour, 1999; Jung et al., 2000; Camy et al., 2003). In the present study, soil chemistry and soil texture, as well as precipitation and temperature, were evaluated in addition to Phytophthora species, to elucidate their possible involvement in the decline of oaks in southern Sweden. Although several different approaches were used to evaluate the significance of the chemical status of the soil for the vitality of the oak stands, no clear associations could be detected, except for a few weak correlations when separating the sites geographically. With regard to the large number of correlation analyses performed, the low values of ρ and the inconsistent results, it is concluded that soil nutrient status and soil acidity do not seem to be related to stand vitality in these 32 stands with predominantly silty to clayey texture. However, results from a large-scale survey on 109 oak stands in southern Sweden, some of which are included in this study, showed that stands with very low base saturation values (< 10%) had significantly higher crown defoliation than stands with normal base saturation values (Sonesson & Anderson, 2001). Oak vitality therefore seems to be related in some way to nutrient availability in the soil, at least in certain regions of southern Sweden.
The increase in average crown defoliation and the proportion of damaged trees with increasing precipitation on sites without Phytophthora suggest that climate is a third factor that may be of importance in oak decline in southern Sweden. Fluctuating water levels on hydromorphic soils have been suggested as the cause for oak decline in several European countries (Thomas et al., 2002), and may possibly be the cause of the association with precipitation found here. The interpretation of the correlation between crown condition and precipitation is complicated, since precipitation also includes deposition of certain elements, such as N and S. These elements may have direct effects on the oaks, or influence them indirectly through their effects on the microbial activity in the soil. In addition, the method used for estimating climatic conditions at the sites is very coarse since meteorological stations were situated at varying distances from the sites and a relatively small number of sites have been included in the study. Further investigations involving a larger number of sites are therefore needed before any conclusions about the role of climate in the decline of oak in southern Sweden can be drawn. However, the results support a previous investigation by Barklund & Wahlström (1998), which suggested that the initial cause of oak decline in southern Sweden was severe bark necrosis due to frost damage caused by the unusually cold winters at the end of the 1980s. It seems likely that extreme weather events, with fluctuations between years with exceptionally high precipitation, years with hot, dry summers and years with low winter temperatures (which occurred in the studied region in the late 1980s as well as in the 1990s; Table 6), rather than a single climatic factor, have affected the trees negatively and predisposed them to the action of other abiotic and biotic stress factors.
Table 6. Average temperature from November to March and average precipitation from April to October for the years 1984–99, and for the period 1961–90, derived from 14 meteorological stations in the area studied (SMHI, 1983–2000)
The results obtained in this study identify some of the factors that may be involved in oak decline in southern Sweden. In some cases, one specific factor is probably the most important for the decline of an oak stand. However, the indistinct results for the 32 oak stands suggest that there are multifactorial causes behind the decline, where several abiotic and biotic factors interact to cause deterioration of the trees (Manion, 1991). The importance of single factors is thus difficult to discern from field studies. Increasing the number of stands sampled would consolidate the results with regard to the impact of Phytophthora, soil chemistry and climate on oaks in southern Sweden, and would simplify the interpretation of possible interactions between the variables investigated. However, based on the results obtained in this study, the following conclusions can be drawn:
- 1Phytophthora quercina seemed to be more likely to occur in declining than in healthy stands. More detailed sampling on the within-stand level is required to determine whether the damage observed is caused by Phytophthora alone, or by the interaction between several abiotic and biotic factors.
- 2Phytophthora species in southern Sweden were found more frequently in clayey and loamy soils, where the rhizosphere was less acidic and where the base saturation was relatively high.
- 3There were no clear relationships between any chemical soil parameter and stand vitality in these predominantly silty to clayey soils. Base saturation does not seem to be lower, or soil acidity higher, in declining stands than in healthy stands. The lack of significant relationships between chemical soil variables and stand vitality may be due to the small number of stands sampled or the limited geographical area from which the stands were selected.
- 4Crown defoliation and the proportion of damaged trees increased with increasing average summer precipitation and average annual precipitation for the period 1984–99, when sites infested with Phytophthora species were excluded. These results suggest that precipitation, or a factor associated with precipitation, is involved in oak decline in southern Sweden.
This project was generously financed by The Swedish Research Council for the Environment, Agricultural Sciences and Spatial Planning, AB Gustaf Kähr, Gustafsborgs Säteri, the National Board of Forestry, Region Skånes Miljövårdsfond, the Regional Forestry Board Södra Götaland, the Regional Forestry Board Östra Götaland, Sparbanksstiftelsen Kronan and Tarkett Sommer. The National Board of Forestry, the Regional Forestry Board Södra Götaland and the Regional Forestry Board Östra Götaland are acknowledged for allowing us to use their database on oak. Special thanks to S. Anderson and M. Lundgren for their help in providing us with the data. We are grateful to A. Jonshagen, J. Jönsson and L. Lundberg for their help with the fieldwork. R. Ohlin performed the chemical analyses. H. Sheppard corrected the language.