By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Forest declines are usually complex multifactorial phenomena that involve interactions between different factors. The possible interaction between different types of mycelial pathogens was investigated through artificial inoculation of oak seedlings, involving two root rot basidiomycetes, Collybia fusipes and Armillaria mellea, and two Phytophthora species, P. cinnamomi and P. cambivora. These pathogens were inoculated onto young Quercus robur saplings in greenhouse conditions, either alone or combining a root rot basidiomycete with a Phytophthora species. Three out of the four Phytophthora spp.*root rot basidiomycete combinations tested resulted in significantly greater damage to the oak host than the sum of the damages induced by the individual pathogens. This positive interaction could be significant in oak decline syndrome.
Forest declines are usually described as complex multifactorial phenomena that require the combination of several factors to occur. Manion (1991) developed a conceptual model of forest decline that postulates a conjunction of three different types of factor that must occur for the onset of a decline: predisposing factors which act over the long term to weaken the trees; inciting factors, short-term stresses that trigger the decline; and contributing factors, mostly opportunistic organisms, which act on the weakened trees to increase or speed up the level of decline and mortality. The importance of interactions between fungal pathogens and insects is often of prominent importance. Indeed, the interactions between bark beetles and their associated ascomycetous fungi have been well documented in pine declines (Aukema et al., 2010), whilst interactions between lepidopterous defoliators and either root rot basidiomycetes or leaf pathogens is important in oak decline (Wargo & Harrington, 1991; Thomas et al., 2002; Marçais & Bréda, 2006). By contrast, although declining trees are often faced with several pathogenic fungi (Guillaumin et al., 1985; Camy et al., 2003b), the potential interaction between those natural enemies has not been investigated.
Oak decline has been a chronic problem in Europe in the last 30 years and typically involves the interaction between several factors, biotic and abiotic (Landmann et al., 1993; Thomas et al., 2002). Such declines have been a growing concern, in particular because of the possible relation with climatic change. In oak decline, fungal pathogens colonizing either the root system (Armillaria sp.) or the bole bark (Biscogniauxia mediterranea) and bark insects such as Agrilus spp. have been reported as important opportunistic parasites, contributing to tree mortality. As in most cases of forest decline, the importance of opportunistic pathogens is controversial (Thomas et al., 2002). Other root pathogens such as Collybia fusipes have been shown to be important in some oak declines and show significant pathogenicity in inoculation trials (Marçais & Delatour, 1996). However, this pathogen destroys the root system of affected oak trees very slowly and its role in the decline is difficult to assess (Marçais & Caël, 2001; Camy & Marçais, 2003). In the last 10 years, the frequent presence of several Phytophthora species on declining oaks has been reported. In particular, P. cinnamomi and P. quercina have been significantly associated with oak decline (Robin et al., 1998; Jung et al., 2000; Jönsson et al., 2003; Balci et al., 2008). However, although these Phytophthora species show pathogenicity toward oak, inducing significant fine-root mortality in inoculation trials on greenhouse-grown oaks, they usually do not induce any seedling mortality or significant reduction in growth (Marçais et al., 1996; Jung et al., 1999, 2000; Jönsson et al., 2003). Phytophthora spp. and root rot basidiomycetes often occur together on the same trees (Camy et al., 2003b). Both are likely to be significantly influenced by host weakening by other natural enemies. Indeed, pathogens such as Armillaria sp. or P. cinnamomi have been shown to be strongly influenced by host physiological condition (Blaker & MacDonald, 1981; Wargo & Harrington, 1991; Marçais et al., 1993; Maurel et al., 2001; Marçais & Bréda, 2006).
The aim of this work was to determine, through inoculation trials on young saplings, whether significant interactions between Phytophthora spp. and root rot basidiomycetes might have some importance in oak decline.
Materials and methods
Plant material, fungal isolates and inoculum preparation
There were two experiments, both conducted in glasshouses. The plants used were 2-year-old Quercus robur seedlings grown from seed originating from the forest of Amance, nearby Nancy (NE France, Meurthe-et-Moselle). They were potted in 9-L pots filled with substrate composed of a 50/50 peat-sand mixture.
Two Phytophthora species were used. The first one, P. cinnamomi isolate P382, an aggressive isolate obtained by G. Strouts in 1980 from soil under Nothofagus procera in Surrey, UK, was used in both experiments. The second was P. cambivora isolate P463, obtained by C. Robin from Castanea sativa (Lot et Garonne, France). C41, the Collybia fusipes isolate used in both experiments was isolated by B. Marçais in 1994 from a root infection on Q. robur at Siarrouy (Hautes-Pyrénées, France). A69, the Armillaria mellea isolate used in the second experiment was obtained by J. J. Guillaumin in 1969 from a root infection on a Prunus species (Pont-St-Esprit, France). The two root rot basidiomycete isolates have been inoculated and reisolated from their hosts several times since the first isolation.
Pieces of wood were colonized by C. fusipes or A. mellea (Marçais & Delatour, 1996). Briefly, stems of hazel (Corylus avellana), 1·5–2·5 cm in diameter, were collected and cut into segments 3 cm long. These were placed in glass jars filled with tap water and sterilized twice at 120°C for 30 min, 24 h apart. The water was drained at the end of each sterilization. A liquid malt medium was added to cover half the height of the wood segments and a third sterilization was performed for 20 min at 120°C. To improve aeration, a hole was drilled in the jar top and plugged with cotton wool. Each of the jars was aseptically seeded with blocks of inoculum (0·5 × 0·5 cm) from a C. fusipes or an A. mellea culture on malt agar [20 g malt (Difco), 15 g agar L−1]. For C. fusipes, the jars were incubated for 30–45 days at 23°C, then all the liquid was drained from the jars with a syringe and the jars were further incubated at 23°C for 8–9 months. For A. mellea, jars were incubated at 23°C for 4 months.
Inoculum of Phytophthora species for soil infestation was produced on a mixture containing 700 mL whole millet grains, 50 mL clarified multivitamin juice broth (14·1 g CaCO3 and 1000 mL vegetable Joker juice) and 200 mL water which was autoclaved twice for 20 min at 120°C. The inoculum was incubated at 20°C in the dark for 3 weeks before use.
Inoculation with the pathogens
Two inoculation experiments were conducted. In the first one, only two pathogens, C. fusipes and P. cinnamomi, were used. Two inoculation times were included for C. fusipes, one before the P. cinnamomi inoculation and one after. Collybia fusipes was inoculated by placing a wood segment colonized by the pathogen onto the collar of each 2-year-old seedling. The collar of the seedling was carefully exposed to avoid wounding, a colonized hazel stem segment was tied to the taproot with an elastic band at about 4–8 cm below the soil line and the soil was replaced over the inoculum. Phytophthora cinnamomi was inoculated in the summer by adding colonized millet seeds to the potting medium. The inoculum (40 mL per seedling) was mixed with the superficial substrate at four locations in the pots. Seedlings were then flooded just after the P. cinnamomi inoculation. Pots were placed in plastic bags and water was added until the level reached the soil surface. The water was allowed to drain after a period of 48 h. The flooding was repeated 2 months later. Pots were separated by treatment during the flooding to avoid cross contamination by Phytophthora spp. Altogether, six treatments were compared: three C. fusipes treatments (no inoculation, inoculation in December 1998 or inoculation in October 1999) × two P. cinnamomi treatments (no inoculation or inoculation in August 1999). Each treatment contained 20 seedlings (altogether 120 seedlings). The experiment was organized in a completely randomized design. The status of the root system was assessed in December 2000, i.e. 2 years after the first C. fusipes inoculation and 1·5 years after the P. cinnamomi inoculation. Seedlings were 4 years old at the end of the experiment.
In the second experiment, two Phytophthora spp. (P. cinnamomi and P. cambivora) and two root rot basidiomycetes (C. fusipes and A. mellea) were included. The two root rot basidiomycetes were inoculated in November 2002 and the two Phytophthora spp. in May 2003, as previously explained. In particular, two periods of flooding 15 days apart were applied after inoculation with Phytophthora spp. Altogether, nine treatments were compared: three root rot basidiomycete treatments (no inoculation, inoculation by C. fusipes or inoculation by A. mellea) × three Phytophthora spp. treatments (no inoculation, inoculation by P. cinnamomi or inoculation by P. cambivora). Each treatment contained 25 seedlings (altogether 225 seedlings). The experiment was organized in a split-plot design, with the root rot basidiomycete treatments randomized within the Phytophthora spp. treatments, to avoid cross contamination between the different Phytophthora spp. treatments. The status of the root system was assessed in March 2004, i.e. 17 months after the root rot basidiomycete inoculation and 5 months after inoculation with Phytophthora spp. Seedlings were 3·5 years old at the end of the experiment.
In both experiments, seedling diameter at 10 cm above the substrate was measured in winter each year, starting at the time of the root rot basidiomycete inoculation. In experiment 2, the height of seedlings was also recorded each year.
Assessment of root damage
Seedlings were monitored for above-ground symptom development during the incubation time. For assessment of root status, seedlings were removed from the pots and all fine roots < 1 mm in diameter were cut. Fine roots were gently washed to remove all traces of substrate and oven-dried at 65°C for 48 h. The dry weight of fine roots was recorded for each seedling. The collar area and inoculum were then examined. Width and height of dead areas of bark on the taproot and main lateral roots were recorded and the surface of the lesions was estimated as the geometric mean of those two diameters [π × (height × width)/4]. Isolations of the parasites were attempted both from the inoculum and the host tissues. They were washed under water, immersed for 1 min in sodium hypochlorite (3·75% active chlorine) and rinsed three times in sterile water. The outer bark was removed and chips of dead bark or decayed wood were plated on either MAT medium [10 g malt (Difco), 100 mg penicillin, 100 mg streptomycin, 250 mg thiabendazole and 15 g agar L−1] or BARPBHy medium [corn meal agar (Difco) with 200 mg ampicillin, 10 mg rifampicin, 10 mg pimaricin, 15 mg benomyl and 50 mg hymexasol L−1]. Inoculations by C. fusipes or A. mellea were recorded as successful when cambial death had occurred.
When no Phytophthora sp. had been isolated from bark lesions on a seedling, a baiting was attempted from the substrate to determine whether the pathogen had established on the seedling root system. Baiting was performed under standard laboratory conditions (about 20°C, diffuse light). About 200 mL of substrate was flooded with 500 mL deionized water and baited by floating two or three rhododendron leaves on the surface. Baits were removed after 3 days and blotted dry; necrotic parts of the leaves were transferred to BARPBHy selective agar medium. Plates were examined daily and possible colonies of Phytophthora spp. were transferred to three different media for identification: corn meal agar amended with ß-sitosterol, potato dextrose agar, or 200 mL Joker (V-8-like clarified multivitamin juice) amended with 7 g CaCO3 500-mL−1, 20 g agar and 800 mL deionized water.
The effect of the inoculation treatments on seedling stem diameter or height growth was analysed by variance analysis (glm, sas/stat 8·1).
Only data from inoculated seedlings where root rot basidiomycetes and/or the Phytophthora sp. had successfully established were used for the analysis (isolation of the pathogen from bark lesions, woody inoculum or successful baiting from the substrate). Frequency of successful establishment on the seedlings was analysed by logistic regression, assuming a binomial distribution and a logit link function (genmod, sas/stat 8.1; SAS Institute Inc.).
The effect of the inoculation treatments on seedling fine-root biomass was analysed by variance analysis. The impact of the various species and their interactions were tested. When the interaction effect was significant, it was tested whether the combined impact of each individual root rot basidiomycete*Phytophthora sp. association was only additive using linear functions of the fitted parameters. More specifically, for experiment 1 for example, the fitted model was:
with DWn the seedling fine root dry weight and CiPj variables taking the value 1 for seedlings not inoculated with P. cinnamomi or C. fusipes (i = 0, j = 0), seedlings inoculated with P. cinnamomi (j = 1) or seedlings inoculated with C. fusipes in December 1998 (i = 1) or in October 1999 (i = 2). To test for an additive effect of the October 1999 C. fusipes inoculation and the P. cinnamomi inoculation, it was tested whether (β0 − β1) − (β2 − β3) was equal to 0.
Lesion sizes on the seedling taproot were analysed using generalized linear modelling assuming a gamma distribution of the lesion areas (genmod, sas/stat 8·1). The gamma distribution is a function taking only positive values and a retrospective analysis of many inoculation experiments with C. fusipes (Marçais & Delatour, 1996; Camy et al., 2003a) shows that it very well suits this type of lesion size data. The identity function was used as a link function. Model validity was checked using the deviance/degree of freedom ratio, by plotting deviance residuals against the linear predictor and with a half-normal plot (Collett, 1991). The impact of the various species and their interactions were tested. Whether the combined impact of each individual root rot basidiomycete*Phytophthora sp. association was only additive was tested as explained previously using linear functions of the fitted parameters.
In the first experiment, with inoculation only with P. cinnamomi and C. fusipes, no foliar symptoms or visible lesions at the stem base developed during the experiment, and the seedlings grew vigorously with no significant difference in stem diameter increment between treatments (F = 1·49, P = 0·2034).
Survival of C. fusipes in woody inocula was 88%, shown by successful isolation. Survival was similar for the two inoculation dates and whether the seedlings were inoculated by P. cinnamomi or not (likelihood χ2 = 1·74, P = 0·6282). Bark lesions were observed on all seedlings for which C. fusipes had survived on the inoculum. Phytophthora cinnamomi was recovered from either bark lesions or from soil by baiting for all the seedlings inoculated with Phytophthora spp.
There was a very significant effect of the C. fusipes inoculation treatments on both fine-root biomass and lesion area on the taproot (Tables 1 and 2, Fig. 1a,b). Seedlings inoculated by C. fusipes alone had fine-root biomass either similar to or higher than that of controls (Table 1, Fig. 1a). Lesions on the taproot were significantly larger on seedlings inoculated with either P. cinnamomi or C. fusipes than on non-inoculated control seedlings. Collybia fusipes induced significantly larger lesions on seedling taproots than P. cinnamomi, whilst the area of the C. fusipes-induced lesions was not significantly different for the two inoculation dates (Table 2, Fig. 1b). The interaction between the impacts of C. fusipes and P. cinnamomi was very significant for taproot lesion area and was close to significance for fine-root biomass (Tables 1 and 2) The two pathogens induced more damage in combination than would have been expected from the addition of their individual impacts when C. fusipes was inoculated in December 1998, for both reduction in fine-root biomass (t = −9·71, P = 0·017, Fig. 1a) and lesion area on the taproot (likelihood χ2 = 6·3, P = 0·012, Fig. 1b). However, for fine-root biomass the significant C. fusipes × Phytophthora sp. interaction was linked to an increase in the biomass of the seedlings inoculated only with C. fusipes. The average size of the taproot lesion induced in the combined inoculation treatment was 2·6-fold higher than what would have been expected from the addition of the two pathogens’ individual impacts. By contrast, when C. fusipes was inoculated in October 1999, after the P. cinnamomi inoculation, their combined impact was not significantly different from what would have been expected from the addition of their individual impacts, whether for the fine-root biomass (t = −5·39, P = 0·189) or the taproot lesion area (likelihood χ2 = −0·90, P = 0·083).
Table 1. Impact of inoculation with a Phytophthora sp. (P. cinnamomi/P. cambivora) and/or a root rot basidiomycete (Collybia fusipes/Armillaria mellea) on Quercus robur seedling fine-root biomass (variance analysis)
C. fusipes inoculation (control/December 98/October 99)
Table 2. Impact of inoculation with a Phytophthora sp. (P. cinnamomi/P. cambivora) and/or a root rot basidiomycete (Collybia fusipes/Armillaria mellea) on lesion area on the taproot of Quercus robur seedlings (generalized linear model assuming a gamma distribution of lesion areas)
C. fusipes inoculation (control/December 98/October 99)
In the second experiment, no foliar symptoms or visible lesions on seedling stems were observed. Again, no differences in diameter growth between the different inoculation treatments were significant (F = 1·75, P = 0·160). However a significant impact of inoculation by Phytophthora sp. was observed on height growth (F = 6·62, P = 0·002), whilst C. fusipes and A. mellea had no impact (F = 0·20, P = 0·821). Phytophthora cinnamomi and P. cambivora had a similar impact on height growth (height increment in 2 years of 150, 145 and 188 cm for seedlings inoculated with P. cinnamomi, P. cambivora or neither, respectively).
A significant difference in survival in the woody inocula was observed between A. mellea and C. fusipes (43% vs. 96%, respectively; likelihood χ2 = 1·74, P <0·001). The survival of the root rot basidiomycetes was similar whether the seedlings were inoculated with P. cambivora or P. cinnamomi, or not inoculated with a Phytophthora sp. (likelihood χ2 = 3·57, P = 0·168). Bark lesions were observed on all seedlings for which C. fusipes or A. mellea had survived on the inoculum. Phytophthora cambivora and P. cinnamomi were recovered from either bark lesions or from the soil by baiting for 90% of the seedlings inoculated with Phytophthora spp. The recovery was the same for the two Phytophthora species (likelihood χ2 = 1·08, P = 0·298) and whether the seedling had been inoculated with a root rot basidiomycete or not (likelihood χ2 = 0·71, P = 0·700).
There was a very significant effect of treatment with Phytophthora spp. on fine-root biomass and of all inoculation treatments on lesion area on the taproot (Tables 1 and 2, Fig. 2a,b). Only P. cinnamomi significantly reduced fine-root biomass compared to the control (Fig. 2a, 60% biomass reduction). Inoculation by P. cinnamomi, C. fusipes and A. mellea induced significantly larger lesions on the taproot than did P. cambivora (Fig. 2b). The interaction between the impact of root rot basidiomycetes, A. mellea and C. fusipes, and the Phytophthora species P. cinnamomi and P. cambivora was very significant for the lesion area on the taproot, but far less for the reduction in fine-root biomass (Tables 1 and 2).
The reduction in fine-root biomass induced by the combination of the root rot basidiomycetes, A. mellea and C. fusipes, and the Phytophthora spp., P. cinnamomi and P. cambivora, was not significantly different from what would have been expected from the addition of their separate impacts (t = 0·34, P = 0·734; t = 0·77, P = 0·441; t = 1·83, P = 0·071; and t = −0·75, P = 0·457; respectively). The only combination close to being significant was that of C. fusipes × P. cinnamomi.
The impact of C. fusipes and Phytophthora spp. and of A. mellea and P. cinnamomi in combination on taproot lesion area was much more than the addition of their separate impacts (likelihood χ2 = 13·35, P <0·001; likelihood χ2 = 14·86, P <0·001; and likelihood χ2 = 6·79, P = 0·001; respectively, Fig. 2b). Depending on the root rot basidiomycete *Phytophthora spp. interaction, the average lesion size induced in combination was 3·6- to 4·9-fold higher than the addition of the average lesion sizes induced alone. By contrast the area of taproot lesion induced by A. mellea and P. cambivora together was not significantly different from what would have been expected from the addition of their separate impacts (likelihood χ2 = 1·90, P = 0·168).
When inoculated alone on the oak seedlings, the root rot basidiomycetes produced typical lesions with a central area with mycelial fans and outer margins with dry, dark-brown necrotic areas from which the pathogen could not be isolated (Fig. 3a,b). Phytophthora spp. produced numerous lesions of limited area located at the base of necrotic fine roots (Fig. 3c). When root rot basidiomycetes and Phytophthora spp. were inoculated in combination, large areas of water-soaked olive-brown dead bark tissues typical of infection by Phytophthora spp. were present around necrotic tissues containing basidiomycete mycelial fans (Fig. 3d–f). Phytophthora cinnamomi was successfully isolated from water-soaked olive-brown dead bark tissues in 85% of cases (n = 54), whilst P. cambivora was isolated in only 13% of the cases (n = 15). Almost all attempted isolations of C. fusipes and A. mellea from central necrotic areas with mycelial fans were successful (65 out of 66).
This study showed that, when inoculated on oak seedlings, Phytophthora spp. and root rot basidiomycetes such as A. mellea or C. fusipes are able to interact, inducing together much more damage than would be expected from the addition of their impacts when inoculated separately. This was particularly true for lesions induced on the taproots.
Interactions between the studied rot root basidiomycetes and Phytophthora spp. appeared to be most significant at the coarse-root level. By contrast, the impact on fine-root biomass was marginally different from what would have been expected from the addition of the separate impacts. The occurrence of the two types of pathogens associated within the same lesions suggests that they directly interact at the infection court to induce larger lesions. During the first experiment, this happened only when P. cinnamomi was inoculated after C. fusipes, which might suggest that the lesions produced by Phytophthora spp. do not provide entry points into host tissues for the root rot basidiomycetes. This is in accordance with the very high inoculation efficiency, close to 100%, observed when C. fusipes was inoculated alone in this work and previous ones (Marçais & Delatour, 1996; Camy et al., 2003a). It also does not appear likely that the root rot basidiomycetes significantly provided entry points into host tissues for Phytophthora spp. Indeed, Phytophthora spp. infected mainly fine-root apexes, and most lesions on oak taproot were induced at the base of infected lateral rootlets when a Phytophthora sp. was inoculated alone. Phytophthora cinnamomi and P. cambivora, are moderately aggressive on European oaks such Q. robur and Q. petraea. Indeed, although they greatly significantly reduced the fine-root biomass of those oak species, as reported in the literature (Jung et al., 1999; Maurel et al., 2001; Jönsson, 2004), usually they do not have a strong impact on the seedling growth or survival. No impact on seedling growth of Phytophthora species has yet been observed on Q. robur (Maurel et al., 2001; Jönsson et al., 2003; Jönsson, 2004). The present study showed an impact on height growth but not on stem diameter growth. During this work, Phytophthora spp. alone were unable to induce extensive lesions on the seedling coarse rots, but when in the presence of A. mellea or C. fusipes, they produced large lesions on the taproot.
A more likely explanation of the interaction between the two types of pathogens would be that bark stressed at the periphery of root rot basidiomycete lesions was especially suitable for colonization by Phytophthora spp., enabling P. cinnamomi and P. cambivora to develop large lesions. On the other hand, root rot basidiomycetes are able to efficiently colonize unstressed root tissues. In particular, C. fusipes was shown to be little influenced by host stress or vigour (Marçais & Delatour, 1996; Camy et al., 2003a). However, root rot basidiomycetes are slow growing and induced limited lesions over a 2-year period (Marçais & Delatour, 1996; Marçais et al., 1996). It was also shown that enlargement of C. fusipes-induced root lesions on mature oaks was very slow over a 6-year period (Camy & Marçais, 2003). By contrast, Phytophthora spp. are able to quickly induce large lesions provided that suitable conditions are met (Robin et al., 2001; Desprez-Loustau et al., 2006).
Whatever the mechanism of the interaction between the two types of pathogen was, the phenomenon was generic. Indeed, the interaction was significant in three of the four root rot combinations that were tested on oak seedlings. This interaction might prove to be very significant under natural forest conditions. Indeed, the impact of pathogens in tree declines, in particular in oak decline, have been underestimated (Thomas et al., 2002), in part because of the low impact of tested pathogens in inoculation trials. However, in natural situations, trees are never faced with one pathogen alone, but more typically an array of several pathogens either specific to the host or generalist. Root rots caused by Armillaria and Phytophthora species are widespread in oak forests (Twery et al., 1990; Jönsson, 2004; Balci et al., 2008) and often occur on the same trees (Camy et al., 2003b). Thus, inoculation trials testing the impact of a pathogen alone on a tree species are probably inadequate to assess the impact of the pathogen when interactions with other frequently occurring pathogenic species are as important as shown during this work. This might explain why some Phytophthora species such as P. quercina have been shown to be significantly associated with oak decline in natural conditions, although they show only moderate pathogenicity in inoculation trials (Jung et al., 1999, 2000; Jönsson et al., 2003).