A conceptual model for the development of Phytophthora disease in Quercus robur
Summary
Here, a conceptual model is presented for the development of Phytophthora disease in pedunculate oak. The model is presented using the causal loop diagram tool and gives an overview of how various abiotic and biotic factors, such as soil moisture, nutrient availability and mycorrhizal colonization, may affect the reproduction and the infective capacity of soil-borne Phytophthora species, the susceptibility of the host and subsequent disease development. It is suggested that the link between the root damage caused by Phytophthora species and overall tree vitality is in the assimilation and allocation of carbon within the plants. The potential impact of environmental factors on these processes is discussed. The model is presented with reference to scenarios related to variation in soil moisture and nutrient availability. The need for species-specific validation of the model and the implications of the model are discussed.
Introduction
During the past decade, several studies have demonstrated the involvement of soil-borne species of the well-known plant pathogenic genus Phytophthora in European oak decline. Brasier et al. (1993) suggested that Phytophthora cinnamomi contributed to oak decline of Quercus ilex and Quercus suber in Iberia. Blaschke (1994) observed progressive deterioration of fine roots and mycorrhizal systems in mature declining Quercus robur and suggested that the damage was caused by Phytophthora species. Since then, 12 other Phytophthora species have been recovered from oak stands growing in a wide variety of soil conditions across Europe (Jung & Blaschke, 1996; Jung et al., 1996, 2000, 2002; Robin et al., 1998; Gallego et al., 1999; Hansen & Delatour, 1999; Vettraino et al., 2002; Balci & Halmschlager, 2003). The most frequently recovered species is the oak-specific Phytophthora quercina (Jung et al., 1999, 2000; Vettraino et al., 2002; Balci & Halmschlager, 2003; Jönsson et al., 2003a). A number of studies have demonstrated that P. quercina, as well as many other Phytophthora species recovered from oak stands, can cause significant root rot in oak seedlings grown in glasshouses (Jung et al., 1996, 1999, 2003b; Robin & Desprez-Loustau, 1998; Robin et al., 1998; Luque et al., 2000; Sanchez et al., 2002; Jönsson et al., 2003b; Jönsson, 2004a) as well as in mature trees in the field (Jung et al., 2000; Jönsson, 2004b). Furthermore, significant correlations have been found between the presence of Phytophthora species in the rhizosphere soil and crown transparency of mature oaks in several European countries (Jung et al., 2000; Vettraino et al., 2002; Balci & Halmschlager, 2003; Jönsson et al., 2005).
Although Phytophthora species have been shown to be involved in the decline of oaks, the link between the root damage caused by the pathogens and the overall tree vitality is poorly understood. In addition, the development of the disease is unclear. The response of the trees to these pathogens varies widely, ranging from remaining healthy, despite close association with the pathogen, to succumbing to death shortly after infection by the pathogen. This indicates that other factors, biotic and/or abiotic, must be involved in the decline. From studies on other plant species, mainly agricultural crops, it is well-known that Phytophthora diseases are highly sensitive to the prevailing environmental conditions (Erwin & Ribeiro, 1996). By using the information available from other host–Phytophthora combinations, as well as knowledge of forest tree physiology and the responses of forest trees to stress, a more comprehensive picture of the Phytophthora disease on oak could be obtained.
The aim of this paper is thus to present a conceptual model for the development of Phytophthora disease in pedunculate oak. The model represents my understanding of the problem and is an attempt to make my understanding transparent to other scientists, thereby creating a basis for discussions on the development of Phytophthora disease in oaks and other forest tree species. The conceptual model gives an overview of how various abiotic and biotic factors may affect the reproduction and infective capacity of Phytophthora, the susceptibility of the host and the subsequent disease expression. It is suggested that the link between the root damage caused by the pathogens and the overall tree vitality is in the carbon (C) assimilation and allocation patterns within trees. A conceptual model for the C assimilation and allocation in oaks is thus included in the conceptual model for the development of Phytophthora disease in Q. robur. The method chosen to present the conceptual model is described and the literature on which the model is based is reviewed in the ‘Theoretical background’ section. The model for the development of Phytophthora disease in Q. robur is presented and its limitations and possible implications are discussed.
Method
To illustrate the interactions between the host, the pathogen and the numerous environmental factors involved, the Causal Loop Diagram (CLD) tool was used. This tool is widely used in systems analysis and systems dynamics studies, where complex multidimensional problems require a holistic approach and studies cannot be carried out solely on isolated parts of the system (Haraldsson, 2004). Similarly, the present work assumes a close interdependence between the host and the pathogen, and a strong response of their relationship to environmental factors. As indicated by their name, CLDs are diagrams that map the causal relationships between the parameters constituting a system, thus creating an integrated network of relationships that is presented in a readable and transparent way. The CLD developed in this study attempts to summarize the direct and indirect effects, as well as the feedback involved in the host-pathogen-environment system.
The CLDs presenting the conceptual model can be seen in 1, 2. The arrows in the CLD represent the causalities as they carry the effect of the variable at the tail of the arrow (origin) to the variable at the head of the arrow (target). As CLDs are conceptual, the only possible changes in the variables are either an increase or a decrease. All changes take place over time, but are driven by changes in another variable. A state of no change is also possible, and thus a variable may be constant over time. An arrow with a positive causality sign implies that a change in the origin variable drives a similar change in the target variable (i.e. an increase in the origin leads to an increase in the target, and a decrease leads to a decrease). Arrows with negative causality signs imply a change in the opposite direction (i.e. an increase in the origin leads to a decrease in the target, and a decrease leads to an increase). This means that the arrows themselves do not indicate whether a variable increases or decreases with time. Bearing in mind this simple rule regarding arrows, CLDs are able to reproduce dynamic behaviour that involves changes between increases and decreases in different variables as the entire system changes with time. In a balancing system, feedback loops move the system in the direction towards fluctuation around an equilibrium point. A feedback loop may also be reinforcing (i.e. move the system away from the equilibrium point).
A conceptual model of how different abiotic and biotic factors influence the aggressiveness of Phytophthora quercina, the susceptibility of Quercus robur and the subsequent disease development. The variable ‘Phytophthora activity’ includes all processes that affect the capacity of the pathogen to infect roots (such as sporangial production, zoospore production, dispersion, adhesion, cyst germination, mycelial growth and oospore germination); CH, carbohydrates. Note that arrows only indicate the direction of the causality. An arrow with a positive causality sign implies that a change in the affecting factor drives a change in the affected factor in the same direction. Arrows with negative causality signs imply a change in the opposite direction. Heavy arrows indicate major feedback loops in the system. For further information and a description of the causal loop diagram (CLD), see the section ‘A conceptual model for the Phytophthora disease in oak’.
A more detailed causal loop diagram describing how nutrient and water availability may influence the carbon (C) assimilation and allocation in plants. CH, carbohydrates; photosynthesis nutrients are potassium (K), magnesium (Mg) and manganese (Mn); growth nutrients are nitrogen (N), phosphorus (P) and sulphur (S). Note that arrows only indicate the direction of the causality. An arrow with a positive causality sign implies that a change in the affecting factor drives a change in the affected factor in the same direction. Arrows with negative causality signs imply a change in the opposite direction. For further information and a description of the causal loop diagram (CLD), see the section ‘A conceptual model for the Phytophthora disease in oak’.
Theoretical background
Linking root damage caused by Phytophthora to overall tree vitality
Plants appear to exhibit a number of rapid responses to infection by Phytophthora species. Cahill & Weste (1983) reported an increase in the rate of respiration of eucalypts as a response to infection by Phytophthora cinnamomi. Increased respiration rate is a general phenomenon occurring in plants subjected to stress (Smeedegard-Petersen, 1984; Agrios, 1997; Orcutt & Nielsen, 2000), and occurs in susceptible plants as well as resistant plants (Cahill & Weste, 1983). However, in resistant plants, the increase and the reversion to normal levels is usually much more rapid than in susceptible plants (Goodman et al., 1986). Furthermore, the increase in respiration in resistant plants is probably caused by the initiation of defence mechanisms that enable the plant to limit the area affected by the pathogen, while for susceptible plants, a great proportion of the increase in respiration is likely to be caused by the invasion and destruction of host tissue (Smeedegard-Petersen, 1984; Cahill et al., 1993; Orcutt & Nielsen, 2000).
In addition to an increase in respiration upon infection with Phytophthora, seedlings of Castanea sativa, Eucalyptus sieberi, Fagus sylvatica, Q. ilex and Q. suber have been found to exhibit a significant reduction in stomatal conductance and transpiration, as well as reduced leaf water potential and soil- to leaf-hydraulic conductance (Dawson & Weste, 1984; Robin et al., 1998, 2001; Maurel et al., 2001a,b, 2004; Fleischmann et al., 2004). These effects are usually explained as being the result of disease-induced water stress caused by impaired water uptake. However, in several studies, the decrease in stomatal conductance preceded the changes in the water relations in the plants (Dawson & Weste, 1982; Robin et al., 1998, 2001; Maurel et al., 2001a,b). Maurel et al. (2001b) demonstrated a linear decrease in stomatal conductance with the percentage of necrotic roots, while the soil- to leaf-hydraulic conductance and leaf water potential were not affected until the percentage of root rot reached 50–60%. This indicates that stomatal closure may be induced by some factor other than water stress, such as nutrient limitation or the presence of elicitin (a toxin produced by certain Phytophthora species; see, for example, Heiser et al., 1999). However, elicitins were found not to influence the activity of stomata in chestnut saplings (Maurel et al., 2004), leaving nutrient limitation as the more plausible factor.
Infection by Phytophthora species may also influence the allocation of C. Cahill & McComb (1992) and Cahill et al. (1993) found an increase in the activity of phenylalanine ammonia lyase (PAL) upon infection of resistant eucalypts with P. cinnamomi. This is an essential enzyme in the shikimic acid pathway, which, together with the pentose phosphate pathway, is used to produce phenolic compounds in plants (Taiz & Zeiger, 2002). Phenolic compounds have been shown to limit the infection of Phytophthora and other pathogens (Afek & Sztejnberg, 1988; Cahill & McComb, 1992; Cahill et al., 1993), as well as impairing the performance of phyllophagous insects (Joseph et al., 1993; Schafellner et al., 1994; Hättenschwiler & Schafellner, 1999). However, it has been suggested that the allocation of C to production of phenols diverts resources from the growth of plants (see the section ‘The C assimilation and allocation of forest trees’), and a large investment in defence against Phytophthora attack may thus result in retardation of growth.
Since infection by Phytophthora seems to have a substantial impact on the respiration, assimilation and allocation of C within the plant, I hypothesize that Phytophthora affect the plants primarily through their influence on the C budget of plants. This influence is mediated through reductions in water or nutrient uptake as a consequence of root damage, but depends on the prevailing environmental conditions, thus leading to variable expressions of the disease. In order to understand how environmental conditions may influence the effect that Phytophthora infection has on C assimilation and allocation in trees, it is necessary to look more thoroughly into the C physiology of forest trees.
The C assimilation and allocation of forest trees
Several different theories for the allocation of C between growth and defence have been put forward during recent decades, ranging from coevolution and sequential evolution to the role of the plant's resources in modifying defensive responses (reviewed in Hartley & Jones, 1997; Farrar & Jones, 2000). For long-lived plants growing on relatively infertile soils, such as acidified soils in temperate or boreal regions, nutrient availability is strongly limiting for growth. Bryant et al. (1983) thus suggested that the type and amount of chemical defence will vary with the environmental availability of nutrients and on the basis of the C/nutrient balance in the plant tissue (the carbon/nutrient balance hypothesis). Briefly, the carbon/nutrient balance hypothesis assumes that growth is generally more severely affected than photosynthesis under conditions of nutrient stress, resulting in the accumulation of carbohydrates. When carbohydrates are produced in excess of growth requirements, this excess is allocated to the production of C-based secondary metabolites such as phenols. Nitrogen-based secondary metabolites, on the other hand, decline. The plant phenotypic response to C stress is essentially the converse. Accordingly, severe drought, with reductions in stomatal conductance and consequently C assimilation (Horner, 1990), results in the preferential allocation of C to primary metabolites and a decrease in the formation of C-based secondary compounds (Bryant et al., 1983). Herms & Mattson (1992) later extended the hypothesis of Bryant et al. (1983) by adopting a more comprehensive approach, addressing the issue of growth and differentiation from the cell to the species level (growth/differentiation balance hypothesis).
A large body of evidence is consistent with the C/nutrient balance hypotheses and the growth/differentiation balance hypothesis (Bryant et al., 1987; Muzika, 1993; Hikosaka et al., 2005; Northrup et al., 1995a,b), suggesting that the C/nutrient balance in plant tissue may be suitable as a predictor of metabolic and biosynthetic changes in carbohydrate production and allocation in a number of different plant species. I suggest that these hypotheses are also applicable to oaks growing in temperate forests where growth is primarily limited by soil resources. The defence system of oaks largely constitutes phenolic metabolites, such as tannins (Feeny, 1970). Phenols originate from phenylalanine, which is a product of the shikimic acid pathway (Taiz & Zeiger, 2002). Certain amino acids share this pathway and phenylalanine can thus be used for either protein or phenolic synthesis (Taiz & Zeiger, 2002). The competition between phenols and other compounds for C is thus obvious (Jones & Hartley, 1999), and a trade-off between growth and defence, depending on the resource availability, seems reasonable. However, growth and defence may not be fundamentally alternative options for a tree in all cases. Some tree species use defence compounds such as terpenes. Terpenes originate from the mevalonic acid pathway (Taiz & Zeiger, 2002), in which a direct nitrogen-based competitor for the C is lacking. In these species, the levels of defence compounds have been shown to be insensitive to changes in nutrient availability (Muzika, 1993; Honkanen et al., 1999).
A generally held opinion for the mechanism behind the allocation of C between roots and shoots is that there is a functional balance between the size and activity of the shoot and the size and activity of the root system, with above-ground parts being favoured when factors perceived by the leaves (CO2 and light) are limiting, while growth of below-ground parts are favoured when factors perceived by the roots (water and nutrients) are limiting (Lambers et al., 1998; Marschner, 2003). A proportionally large investment in roots under moderately dry conditions is common in Quercus species (Osonubi & Davies, 1981; van Hees, 1997; Vivin & Guehl, 1997). This is probably caused by the stronger influence of drought on leaf expansion than photosynthesis, resulting in an increase in the availability of carbohydrates, and thus enhanced export to the roots (Ericsson et al., 1996; Lambers et al., 1998). Shading, on the other hand, usually results in a reduction of the root : shoot ratio of oak seedlings (van Hees, 1997; van Hees & Clerkx, 2003). However, Ericsson et al. (1996) suggested that the effect of nutrient availability on the C assimilation and allocation in trees is slightly more complex. According to their studies on birch, a shortage of nitrogen (N), phosphorus (P) or sulphur (S) usually results in an increased allocation of C to the roots, thereby favouring root growth over shoot growth, while a shortage of potassium (K), magnesium (Mg) or manganese (Mn) results in a decrease in the allocation of assimilates to the roots (Ericsson, 1995; Ericsson et al., 1996). The increased allocation of C to roots as a consequence of shortages of N, P and S is, according to Ericsson et al. (1996) not an example of a rational plant reaction. Rather, it appears to reflect a situation in which the availability of assimilates is not limiting for structural growth. Roots are the organ whose growth is least suppressed when the formation of new tissues rather than C fixation is the process most strongly inhibited by mineral shortage. This is likely to be caused by the close proximity between sites of mineral uptake and mineral utilization in the roots, and thus efficient capture of the limiting nutrient in the root meristems. Consequently, root growth is favoured over shoot growth. The decrease in assimilates as a consequence of a shortage of K, Mg or Mn is, according to Ericsson et al. (1996), a function of their influence on the photosynthetic apparatus in plants. Potassium is important for stomata regulation, Mg for capturing light energy and Mn for photosynthetic O2 evolution. In addition, K and Mg are components of enzymes essential to C fixation (Lambers et al., 1998; Marschner, 2003). Thus, limitation by K, Mg and Mn causes a reduction in C fixation and thus in assimilate availability. The suppression of root growth compared with shoot growth when plants are low in nonstructural carbohydrates is most probably a consequence of the roots being the organ most distant from the supply of the growth-limiting substrate (carbohydrates). It is important to note that Ericsson et al. (1996) describe the allocation in terms of fractions of carbohydrates and nutrients. Substantial cycling between roots and shoot of C and nutrients has been demonstrated, and a large fraction of the N taken up by roots is still exported to the shoot under N-limiting growth conditions. However, a larger fraction of the N will be used in the root under N-limited conditions than in more favourable N situations.
The literature reviewed above indicates that both roots and shoots exert some control over the net acquisition and allocation of C. Farrar & Jones (2003) suggested a shared control hypothesis in which each flux that contributes to the net acquisition of C exerts some degree of control over the process. However, the processes governing the distribution of C between different plant parts are not well understood.
The influence of abiotic and biotic factors on Phytophthora activity and disease development
Water availability Phytophthora diseases are multicyclic, which means that inoculum may multiply rapidly when the environmental conditions, the most important of which is the presence of free water, are favourable (Erwin & Ribeiro, 1996). Phytophthora diseases may thus develop epidemically when the soil remains excessively wet for prolonged periods and temperatures remain fairly low (Erwin & Ribeiro, 1996). However, regarding the development of disease in oak, drought or variations in water availability, especially fluctuations between drought and flooding, also seem to be detrimental to the fine-root system (Brasier et al., 1993; Robin et al., 2001; Sanchez et al., 2002). Jung et al. (2003a) demonstrated the ability of P. quercina to survive during periods of drought, and found higher amounts of root damage to seedlings under conditions where drought and flooding were alternated, than when moist soil conditions prevailed between flooding cycles. The high amount of damage following alternating flood and drought conditions may result from the simultaneous increase in the production of sporangia upon rewetting and the production of new, unsuberized fine roots, which are highly susceptible to infection. Extended periods of drought may also increase root exudation, thereby facilitating the initial establishment of soil-borne pathogens in the roots (Duniway, 1977), and may, in the event of an increase in soil moisture, favour Phytophthora over other soil microorganisms, since Phytophthora respond very rapidly to changes in soil moisture. Restricted water availability may also critically reduce the tolerance of the host to the pathogen through its negative influence on the defence mechanisms of the plants (see the section ‘The C assimilation and allocation of forest trees’). Predisposition of plants to infection by Phytophthora as a consequence of drought has been demonstrated by, among others, Duniway (1977), Blaker & MacDonald (1981) and Höper & Alabouvette (1996).
Chemical properties of soil In general, Phytophthora diseases are considered to be more severe at higher pH values (Schmitthenner & Canaday, 1983). This is supported by the increase in the activity of Phytophthora with increasing pH (Klisiewicz, 1970; Muchovej et al., 1980; Byrt et al., 1982). For the Phytophthora species found in oak stands, Jung et al. (2000) showed that sporangia cannot be formed at pH(H2O) values below 4.0. In addition, the production of sporangia was found to increase with increasing pH, at least up to pH 5. Concentrations of aluminium (Al) and calcium (Ca) in the soil are often inversely and intimately related to soil pH, and may also affect the aggressiveness of Phytophthora. High concentrations of Al have been shown to inhibit mycelial growth, as well as sporangial formation and germination (Maas, 1976; Muchovej et al., 1980; Benson, 1993; Andrivon, 1995). Because of the negative effects of pH and Al on the activity of Phytophthora, it has been suggested that these diseases do not occur on acidic, Al-rich soils. However, Jönsson et al. (2003b) and Jönsson (2004a,b) showed that P. quercina could cause substantial root damage to seedlings as well as mature trees under acidic conditions (pH < 4.3, Al concentrations > 100 µg g−1). This may possibly be caused by an increased susceptibility of trees in acidified soils (see the section ‘The C assimilation and allocation of forest trees’) or an adverse effect of Al on roots. A number of studies have demonstrated that elevated concentrations of Al in the rooting medium may lead to disruption of cell function and impaired root growth of trees (DeWald et al., 1990; Kochian, 1995; Matsumo, 2000). Conversely, Quercus species seem to be relatively tolerant to the concentrations of Al found in natural soils (Keltjens & van Loenen, 1989; Thornton et al., 1989).
Phytophthora diseases are usually regarded as more severe at high Ca levels (Schmitthenner & Canaday, 1983; Erwin & Ribeiro, 1996). However, there are also examples of high concentrations of Ca having no effect on or suppressing disease (Klotz et al., 1958; Muchovej et al., 1980; von Broembsen & Deacon, 1997) and a number of studies have demonstrated negative effects of high concentrations of Ca on the activity of Phytophthora in vitro (von Broembsen & Deacon, 1997; Hill et al., 1998; Messenger et al., 2000). Although the results with regard to Ca are contradictory, it is obvious that minimum levels of this element are necessary for zoospore production and cyst germination (Halsall & Forrester, 1977; von Broembsen & Deacon, 1996, 1997; Xu & Morris, 1998), for Phytophthora infection of roots through its effect on zoospore taxis and adhesion to solid surfaces (Gubler et al., 1989; Deacon & Donaldson, 1993), as well as for the stimulation of oospore germination (Ribeiro, 1983). With regard to the relatively low concentrations of Ca found in the soil solution in temperate oak forest soils (Berger & Glatzel, 1994; Freer-Smith & Read, 1995; Bakker et al., 1999), and the relatively high concentrations that are usually necessary to suppress disease, it seems likely that increasing concentrations of Ca in this type of forest soil would accelerate rather than suppress the development of disease in oaks.
The results are also inconsistent for K, Mg and P, both with regard to in vitro activity and disease development in vivo (Kincaid et al., 1970; Newhook & Podger, 1972; Maas, 1976; Halsall & Forrester, 1977; Grau et al., 1989; Phukan, 1993; Hill et al., 1998; Appiah et al., 2005). It is possible that the impact of these elements on disease development is more closely related to host susceptibility than to pathogen aggressiveness (see the section ‘The C assimilation and allocation of forest trees’), and are thus strongly dependent on their relations to N. This is supported by the decrease in disease severity when complete fertilizer is applied (Schmitthenner & Canaday, 1983; Pacumbaba et al., 1997).
The effect of N on disease development seems to depend on both the host–pathogen combination and the type of soil N (Klotz et al., 1958; Newhook & Podger, 1972; Lee & Zentmeyer, 1982; Halsall et al., 1983; Utkhede & Smith, 1995; Das et al., 2003). In general, increasing concentrations of NO3− stimulate the activity of Phytophthora (von Broembsen & Deacon, 1997; Jung et al., 2003a), while the effect of NH4+ and organic N seem to be smaller (Elliott, 1989; Jung et al., 2003a). For P. quercina and Q. robur, Jung et al. (2003a) found that the production of sporangia in vitro increased with increasing concentrations of nitrate in the soil leachate, and that the difference in the fine-root length and the number of fine-root tips between uninfected and infected seedlings also increased with increasing nitrate concentration in the soil. Studies of ‘little leaf disease’ caused by P. cinnamomi in shortleaf pine (Pinus echinata), growing in degraded soils in the south-eastern USA, have shown that fertilization with N may prevent symptom development in healthy trees and improve the condition of little leaf trees in early stages of disease (Tainter & Baker, 1996). However, in situations where N is in excess of plant demand, Phytophthora are likely to have an advantage because of the nitrate-induced stimulation of sporangia, the lower availability of C-based defence compounds in roots (see the section ‘The C assimilation and allocation of forest trees’), and possibly also a greater susceptibility of tissues as a consequence of nutrient imbalances (Marschner, 2003).
Biotic factors A number of studies have shown that various biotic populations in the soil influence the severity of disease caused by Phytophthora species, most likely by exerting competitive pressure on the Phytophthora species, thereby excluding them from the root surface (Marx, 1972; Weste & Vithanage, 1977; Keast & Tonkin, 1983; Malajczuk, 1983). In particular, mycorrhizal colonization of roots has been suggested to be an efficient barrier to Phytophthora infection (Zak, 1964; Marx, 1972; Barham et al., 1974; Branzanti et al., 1999). Using agar plates and liquid cultures, Marx (1969) demonstrated an inhibitory effect of several ectomycorrhizal species on the growth of mycelia, as well as on the motility and germination of zoospores of P. cinnamomi. Furthermore, the presence of fully developed mycorrhiza with a fungal mantle and a Hartig net on individual roots of shortleaf pine seedlings was seen to provide resistance to zoospore and mycelial infection, and to have a protective influence on adjacent nonmycorrhizal roots (Marx & Davey, 1969a,b; Marx, 1970). The protective effect of the mycorrhizal symbiosis was evident throughout at the whole root system of the seedlings (Marx, 1973). By contrast, nonmycorrhizal roots and roots with incomplete fungal mantles were infected by the pathogen (Marx & Davey, 1969a,b; Marx, 1970; Barham et al., 1974). Similar results were found by Branzanti et al. (1999) when investigating the importance of ectomycorrhizal fungi in the reduction of chestnut ink disease.
These results suggest that mycorrhizal colonization and the presence of other microorganisms in the rhizosphere soil may render oak trees less susceptible to infection by Phytophthora species, although root infection is probably not completely inhibited (Jönsson, 2004a). However, low vitality of oak trees (Causin et al., 1996; Kovacs et al., 2000) and a high input of N into ecosystems (Brunner, 2001; Nilsson, 2004) have been found to influence the ectomycorrhizal colonization of roots and mycelial growth of ectomycorrhiza negatively, and may thus counteract these positive effects.
A conceptual model for the Phytophthora disease in oak
Based on the literature presented in the Theoretical background section, a conceptual model for the development of disease in oak as a consequence of Phytophthora infection was developed (Fig. 1). The aim of the model is to give an overview of how different environmental conditions, including abiotic as well as biotic factors, act simultaneously on the host–pathogen system, sometimes accelerating disease development, and sometimes retarding it. The model is applicable to the interaction between soil-borne Phytophthora species and Q. robur growing in acid forest soils (pH < 5) in a temperate climate. Only the effects of the pathogen on the fine-root system are considered, since damage to the stem of oak as a consequence of Phytophthora infections is limited. The processes that influence the capacity of Phytophthora to infect roots, such as sporangial and zoospore production, dispersion, taxis, adhesion and cyst germination, are collected in one variable called ‘Phytophthora activity’. The variable ‘microbial activity’ does not include Phytophthora or other pathogens, since these are considered separately. For simplicity, some of the information regarding the C assimilation and allocation has not been included in Fig. 1. For a more detailed illustration of the C submodel, see Fig. 2. The purpose of the C submodel is to elucidate the balance between root and shoot growth and allocation to secondary metabolites in the trees. The model thus only describes these three pools of carbohydrates in the tree and the flow of carbohydrates between them. Carbohydrates required for maintenance and construction of new tissue are included in the parameter ‘metabolized CH for growth’. It is assumed that nutrient uptake is strongly correlated to fine-root length (Atkinson, 2000), and that an increase in the allocation of carbohydrates to roots results in an increase in the live root length rather than an increase in the uptake efficiency per unit length of root. In addition, the model does not consider various groups of C-based compounds or individual compounds, and it does not explain the physiological processes. Despite differing results with regard to the influence of Ca on the reproduction and dispersion of Phytophthora, and on disease development, it was assumed that an increase in the concentration of Ca enhances the activity of Phytophthora at the concentrations found in acidic oak forest soils (see the section ‘The influence of abiotic and biotic factors on Phytophthora activity and disease development’). It was also assumed that K, Mg and P, which have been demonstrated to have variable effects on Phytophthora diseases, influence the development of disease primarily through their effects on C assimilation and allocation patterns in the trees. Temperature is likely to have a certain effect on Phytophthora activity and may also influence the C allocation in trees. However, the effects of temperature are often overridden by concomitant changes in water and nutrient availability and growth patterns of plants (Pregitzer et al., 2000; Pregitzer, 2003), and thus temperature was not included in the model. Note that increased soil moisture does not necessarily imply increased uptake of water. However, the link has been included since low availability of water will lead to a reduced water uptake.
To illustrate how the CLD in Fig. 1 can be used to predict tree response to Phytophthora infection under different conditions, two different climatic scenarios are considered. The subscenarios have been chosen to illustrate a situation that has been common in European forests until now – tree growth is limited by N – and two situations that may become more common in the future as a consequence of soil acidification and an extensive deposition of N – an excess of N in the ecosystem and limitation of tree growth by a base cation.
Scenario 1: Heavy rain or prolonged periods of rain
Assuming that soil moisture increases, as a result of heavy rain or prolonged periods of rain, the production and dispersion of Phytophthora zoospores will increase, thereby increasing the probability of root infection and, consequently, root damage. Root damage may lead to an increased leakage of carbohydrates from the roots, thereby facilitating zoospore taxis of Phytophthora (Carlile, 1983). When the zoospores can find the roots easily, this is likely to result in a higher degree of infection and thus an increase in the amount of damage. Microbial competition may also be reduced under anaerobic conditions in the soil, thus enhancing the effect of flooding on root infection by Phytophthora. Lack of O2 may also cause hypoxic damage and thus facilitate Phytophthora infection. However, Q. robur is relatively tolerant to both hypoxia and the oxidative stress imposed on roots when the waterlogged conditions cease (Colin-Belgrand et al., 1991; Wagner & Dreyer, 1997), and anaerobic conditions are thus not likely to cause direct damage to roots. Infection by Phytophthora may initiate defence reactions in the plants, and more C will then be allocated to the production of secondary metabolites. The fraction of carbohydrates available for metabolization for growth will then decrease. Continuous infection by Phytophthora may empty the stores of carbohydrates in the tree, and subsequently cause retardation of growth. If growth is reduced, this may, together with the damage caused by pathogen infection, result in a reduction in the length of live roots, and thus in impaired uptake of water and nutrients.
Tree growth is limited by N If damage or growth reduction as a consequence of Phytophthora infection results in shortage of N, an element which is important for metabolizing carbohydrates for growth (see the section ‘The C assimilation and allocation of forest trees’), the available mineral is likely to be metabolized in the root first because of the close proximity of the sites of uptake and mineral assimilation. Consequently, only a small proportion of the N will be transported to the shoot. Less carbohydrates will thus be metabolized in the shoot, resulting in a reduction in shoot growth. A larger fraction of carbohydrates will then instead be available for transport to the roots, and may thus be used for metabolization in the root, and subsequently root growth, provided that there is sufficient N in the root to metabolize it. Roots lost owing to pathogen infection may then be replaced and ectomycorrhizal colonization may be promoted, thus limiting root infection by Phytophthora. Furthermore, the relatively high availability of nonstructural carbohydrates is likely to lead to an increased flow through pathways that result in the synthesis of C-rich secondary metabolites, some of which may be used to protect the trees towards various stress factors. Nitrogen-limited trees may therefore have a good protection and may also be able to maintain a high root production, replacing roots that are lost as a result of, for example, Phytophthora infection. However, severe N-deficiency may eventually lead to a reduction in the leaf area, and subsequently in the rate of photosynthesis, thus resulting in a lower availability of nonstructural carbohydrates.
High N deposition has resulted in excess N in the ecosystem If there is a high availability of N, a large fraction of the available carbohydrates is likely to be invested in growth, particularly shoot growth (see the section ‘The C assimilation and allocation of forest trees’), while the amount allocated to root growth and the production of secondary metabolites will be low. Tree tissue may thus become more susceptible to root infection and damage, and the ability of the tree to replace lost roots will be low. Moreover, an accumulation of NO3− in the soil may influence the development of the disease through the positive effect of NO3− on zoospore production.
Tree growth is limited by K or Mg If the reduction in nutrient uptake instead results in a shortage of K or Mg, elements that are important for photosynthesis, the rate of photosynthesis will be reduced and, subsequently, so will C fixation. Since the carbohydrates are formed in the photosynthetically active part of the seedlings, the available carbohydrates will probably be metabolized for growth in the above-ground part first, leading to a low translocation of C to the roots and, consequently, low root growth compared with shoot growth. In addition, little carbohydrate will be available for the production of C-based secondary metabolites. If root growth, and thus live root length, is decreased, the uptake of base cations is likely to be further impaired, and the ability of the tree to replace roots lost as a result of pathogen infection and to protect its tissue will be low. Low C allocation to roots may also lead to reduced ectomycorrhization and growth of external mycelia (Brunner, 2001), thereby further increasing the potential for Phytophthora to infect roots. However, soils in which K or Mg limit tree growth are often acid, implying a low pH, low concentrations of Ca and high concentrations of Al. All of these factors may influence the Phytophthora activity negatively and may thus decrease the probability of root infection. However, if infection takes place, the tree is bound to have a low resistance towards the pathogen.
Scenario 2: Tree growth is limited by water
If we instead assume there is a drought, this will have the reverse effect on sporangial production and the dispersion of zoospores. Drought may thus decrease the degree of root infection. However, severe drought may also decrease the rate of photosynthesis and subsequently impede the replenishment of the pools of carbohydrates in the plant. Carbon will then be preferentially allocated to primary metabolites in the shoot and the fraction of C allocated to roots and the production of secondary metabolites will decrease. A long period of drought may thus completely empty the stores of carbohydrates in the plant. Plant tissue will then become more susceptible to damage and in the event of an increase in soil moisture, the ability of Phytophthora to infect the tree and cause damage to the roots may increase. In addition, drought influences the soil microbial community negatively, thereby increasing the potential for Phytophthora to infect roots in the event of a subsequent increase in soil moisture, since these pathogens respond extremely quickly to an increase in moisture. Secondary pathogens often follow the infection of roots by Phytophthora, and may exacerbate root damage, thus increasing the carbohydrate leakage from the roots, and, possibly thereby influencing the activity in the rhizosphere of Phytophthora as well as other microbes.
To summarize Fig. 1, soil-borne Phytophthora species have the ability to reduce the live root length of oak substantially. However, as long as the tree can sustain new root production to replace the lost roots, nutrient and water uptake will not be affected. The continuous production of new fine roots, to replace those lost because of pathogen infection, is likely to gradually deplete the carbohydrate stores. The susceptibility of the tree to pathogen infection as well as to other types of stress may thus increase. If environmental conditions are favourable for the pathogen, the live root length of the host is likely to eventually decrease, leading to a lower capacity for water and nutrient uptake. Depending on the nutrient limiting the tree, growth or photosynthetic rate may be more strongly affected. If the rate of photosynthesis is reduced, smaller amounts of carbohydrates will be available for compensatory growth and for the production of secondary metabolites involved in resistance. Tree vitality will subsequently decrease, as disease development is accelerated through negative feedback loops, as presented in Fig. 1. The range of field responses to Phytophthora, from slow decline to sudden death of trees, is likely to reflect differences in the interaction between the pathogen and the host, as a consequence of their response to different abiotic and biotic factors.
Discussion
This paper presents a conceptual model for the development of Phytophthora disease in oak. By using a CLD to describe the interaction between the host, the pathogen and the environment, and by emphasizing changes in the susceptibility of the host as well as the aggressiveness of the pathogen, it was my intention to visualize the connections and the feedback in the host–Phytophthora system in a more comprehensive way than has previously been done. When reviewing the literature and creating the conceptual model, it became apparent that many environmental factors that are regarded as influencing the development of disease, as a result of their effect on aggressiveness of Phytophthora might also affect the susceptibility of the tree. These effects are sometimes additive, increasing both the aggressiveness of the pathogen and the susceptibility of the host. At other times, the effects counteract each other, and the balance between pathogen growth and host resistance may thus remain unchanged. The CLD enables detection of such causalities and may, in addition, facilitate the interpretation of fictive disease development when creating scenarios.
The C assimilation and allocation in trees is a disputable topic. The studies we found for Quercus species regarding the influence of N on the production of defence compounds (Thomas & Schafellner, 1999) and the influence of drought on C assimilation and root : shoot ratios (Osonubi & Davies, 1981; van Hees, 1997; Vivin & Guehl, 1997) support the theories suggested by Bryant et al. (1983), Herms and Mattson (1992) and Ericsson et al. (1996). In addition, a previous pathogenicity test in the glasshouse suggests that soil nutrient availability and the nutrient status in oak seedlings influence the production and the allocation of biomass to roots differently, depending on whether N or a base cation is limiting (Jönsson, 2004b). However, none of these experiments were designed to test the theories and more research is needed to confirm whether they are valid for oak. It also seems likely that these theories are applicable primarily to the constitutive defences of plants. In the case of induced defences, it seems probable that plant signalling may overshadow the C/nutrient balance and redirect the flow of carbohydrates to the wounded area. This suggests that the C/nutrient balance hypothesis is of importance primarily for the continuous maintenance of defences, such as during repeated infection or when the plant is subjected to multiple forms of stress. Sheen et al. (1999) suggested that carbohydrate concentrations may provide a path to control responses to light, nutrient concentrations and stresses, while plant hormones may govern intrinsic developmental responses.
The information used to determine the influence of environmental variables on the development of disease was gathered from various host–pathogen combinations, ranging from Phytophthora diseases on agricultural crops to Phytophthora species on Q. robur. The effects of Phytophthora species have often been described as being host specific, and generalizations regarding the effects of various environmental factors, as in this paper, are therefore questionable. In addition, much of the data for trees have been gathered from experiments on seedlings. The response of seedlings and mature trees to pathogens and environmental conditions will probably differ. This must be taken into account when interpreting the model. However, the aim of this model was to create a basis for further discussion. To validate the model, substantial research on the susceptibility of oak, the aggressiveness of the pathogen and the final disease outcome will be required. Furthermore, the interaction between Phytophthora and other factors that have been suggested to contribute to the complex course of European oak decline, such as defoliators, frost and secondary pathogens (Thomas et al., 2002), must be investigated.
If the allocation patterns in oak change as a consequence of a change in the nutrient that is limiting the growth of the tree, this may have serious implications for oak decline in Europe in the future. Nitrogen has hitherto been the limiting nutrient for tree growth in the major part of the temperate as well as the boreal zone. However, the extensive deposition of N during the past decades, together with substantial acidification-induced leaching of nutrients from the soil, has led to a situation in which N is no longer bound to be limiting. Instead, base cations, such as Mg and K, have been shown to limit the growth of trees in certain regions (Hüttl, 1990; Thelin, 2000). According to the C/nutrient balance hypothesis, this implies a reduced capacity for trees to assimilate C and a reduction in the allocation of C to roots and the production of defence compounds (Bryant et al., 1983; Herms & Mattson, 1992; Ericsson et al., 1996). Several studies have implied a reduction in the ability of forest trees to protect themselves against parasites and frost damage as a consequence of N fertilization (Flückiger & Braun, 1999; Hättenschwiler & Schafellner, 1999; Jönsson et al., 2004). Acidification-induced reductions in soil-available K and Mg, and a high N availability, may thus accelerate the damage caused by Phytophthora in oak in Europe. No evidence of excess N or nutrient imbalances in European oak stands has yet been reported (Berger & Glatzel, 1994; Thomas et al., 2002), but considering the high acidity of soils and the small pools of base cations, the continued input of acidifying compounds is likely to eventually lead to nutrient deficiencies and decreased ecosystem stability.
Conclusions
The conceptual model presented in this paper is an attempt to bring together the current knowledge on Phytophthora-induced diseases in oaks. By visualizing the available information and shifting the focus from single causal relationships to a dynamic network of relationships incorporating simultaneous multiple effects on the host–pathogen interaction, gaps in our knowledge can be identified, new questions raised and hypotheses proposed. In addition, the development of the disease caused by Phytophthora species in pedunculate oak may be investigated under different conditions by using fictive scenarios. However, since the model is based on literature from a wide range of host–pathogen combinations, and since the effects of environmental factors on pathogen aggressiveness and host susceptibility vary substantially, species-specific studies are necessary to validate the model. In vitro studies of the influence of nutrient availability on the growth and reproduction of P. quercina are currently being undertaken and will be followed by investigations on C assimilation and allocation patterns in oaks growing under differing conditions and infected with Phytophthora.
Acknowledgements
This project was funded by The Swedish Research Council for the Environment, Agricultural Sciences and Spatial Planning. U. Rosengren is acknowledged for assistance in the initiation of this work and for commenting on previous versions of the manuscript. S. Belyazid is gratefully acknowledged for interesting and fruitful discussions, useful comments on the manuscript and for technical assistance. H. Sheppard corrected the language.
