INRA Bordeaux, Pathologie Végétale, BP 81, 33 883 Villenave D'Ornon, France

Background and objectives
The involvement of Phytophthora cinnamomi Rands in the decline of cork oak and holm oak (Quercus ilex and Q. ;suber) observed in Mediterranean Europe, has been proposed by Brasier et al. [1]. Primary symptoms caused by P. ;cinnamomi are necrosis of unsuberized root, which may extend into larger roots and trunks. Secondary symptoms, similar to those induced to water stress, may lead to tree death. Several factors may be involved in oak decline and interact with P.  ;cinnamomi infection. In southern Europe, the most severe stress is undoubtably drought. However, repetitive floodings, after heavy rainfall, may induce host stress by root anoxia. These stresses may increase the susceptibility of trees. But development of P.  ;cinnamomi in tissue and plant is favoured by increasing water contents.

The aim of this study was to investigate how P.  ;cinnamomi infection and different watering treatments may affect tree-water relations.

Materials and methods
The factor ‘inoculation by P. cinnamomi’, by soil infestation, was crossed with the factor ‘watering treatment’. In a first experiment, seedlings of cork, holm and red oaks <Q.  ;rubra) and of chestnut (Castanea sativa) were subjected to water stress by withholding water supply for 3 ;weeks. In a second experiment, seedlings of cork and holm oaks were subjected, every 4 ;weeks, to 1-day flooding episodes. In both experiments, the physiological status of seedlings was monitored by predawn leaf water potential (PLWP) and stomatal conductance (gs) measurements. At 10 and 20 ;weeks after inoculation (wai), respectively, plants were harvested, scored for mortality, root loss and for percentage of infected taproot (PIT).

Results and conclusions
In both experiments: mortality rates, taproot infection, root loss caused by P.  ;cinnamomi, were higher for chestnut and Q.  ;ilex than Q.  ;rubra and Q.  ;suber. These results confirm the first reports of interspecific variability of susceptibility in oaks [2].

The water withholding had no significant effect on root loss and taproot infection in oaks. The repetitive flooding episodes induced a decrease in the root weight, in controls as well as in inoculated plants. Whatever the flooding treatment and the species, the PIT was greater than 50%. Fungal infection induced an early significant decrease in PLWP of inoculated plants of chestnut and holm oak. A simultaneous significant effect was always observed on gs. In cork oak, the decrease induced by infection was significant on PLWP only after 10 wai, but significant on gs as soon as 5 wai. No effect was observed in red oak. No significant interaction between watering treatments and infection was observed on gs. However, in holm oak, there was a synergistic effect of floodings and fungal infection on PLWP.

In conclusion, we showed that root infections results in a decrease in PLWP, which is explained by a decrease in water absorption capacity and appears to be quantitatively related to the root loss. Root infection also induces a stomatal regulation, which leads to a stomatal closure. As in the case of droughts this regulation occurs before the reduction in PLPW. Although the water stress did not increase the root necrosis, it is likely that infected plants are more vulnerable to drought or flooding and these effects are additive or synergistic to P. ;cinnamomi root infection effects on water relations.

1. Brasier CM, 1996. 53, 347-358.
2. Robin C, Desprez-Loustau ML, Capron G, Delatour C, 1997. Annals of the Science of Forestry, in press.