The soil-borne fungus Fusarium oxysporum f.sp . lilii Imle causes bulb and scale rot of lilies ( Lilium L.) , annually resulting in a considerable economical damage in bulb and flower cultivation. Presently, the prevention of Fusarium damage depends on the application of a combination of fungicides and cultivation practices, such as crop rotation and the use of healthy starting material, however, this is not satisfactory. In the Netherlands the application of soil-disinfestation chemicals must be significantly reduced in the future. An enviromnental friendly and durable solution is the cultivation of (partial) resistant cultivars. These are, however, not widely available in the current assortment and, therefore, must be developed. The objectives of the experiments described in this thesis were: (1) the development of screening methods, (2) the evaluation of genetic variation in lily and Fusarium, and (3) the development of selection methods, by which the breeding companies will be able to develop Fusarium resistant lily cultivars (see General Introduction).
A standardized screening method for an accurate determination of partial resistance to Fusarium in lily clones is described. The level of resistance can be measured by weight changes of the plant caused by Fusarium affection, as well as by a classification of the affected bulbs in ordinal disease ratings (on a scale of 1 to 6, with 1 = not affected and 6 = completely decayed). The data of the weight measurements were analyzed statistically using an analysis of variance. The disease ratings were analyzed using a threshold model. This model calculates a disease severity score (DSS) using an underlying continuous and linear scale. Both measurements were repeatable, accurate, and significantly correlated. The use of disease ratings is, despite a more complicated analysis, less laborious and can suffice with a smaller control group (see Chapter 1).
Vegetative propagation of lily has several developmental stages. It was possible to test reproducibly lily plant material in four different stages (separate scales, scale bulblets, yearling bulbs, and commercial bulbs). The sensitivity for Fusarium decreased with an increasing bulb age. A significant correlation was found between the resistance in the four stages mentioned, which allows testing in one stage only. In the scale test some deviating results were obtained, possibly because of the presence of a large wound. Since the scale test is more efficient than the bulb tests, this test is preferable for determination of Fusarium resistance in large numbers of genotypes. Selected material has to be retested, preferably using the scale bulblet test which is second in efficiency (see Chapter 2).
Standardisation of the tests is not always possible. A high temperature, a high inoculum concentration, and a longer duration of the experiment raised the level of affection. The ranking of the cultivars was not changed; only the accuracy was reduced using more extreme testing conditions (see Chapter 3). Since no genotype x environmental interactions were found, it is expected that cultivars which are most resistant in the standardized tests, will also be most resistant under field conditions (see General Discussion).
Sources of resistance are necessary for breeding purposes. Within the Asiatic hybrid lilies, which belong to the Sinomartagon section, a high level of Fusarium resistance has been found. Also, in species within this section which are related to the Asiatic hybrid lilies, such as L.dauricum , L.davidii , and L.tigrinum , high levels of resistance were detected. Absolute resistance was not found. In the Oriental hybrid lilies ( Archelirion section), a low level of resistance was found. In the Oriental hybrid lily related species of the Archelirion section no resistance was obtained. In L.henryi, however, resistance was observed. This species can be crossed with the Oriental hybrid lilies. Since only a few Oriental hybrid lilies were tested, it is recommended that a large collection will be screened for Fusarium resistance. In cultivars of L. longiflorum a moderate level of resistance was found. The level of resistance of cultivars within this species can probably be raised by specific cross combinations and selection. More variation may be detected in wild accessions from Japan. Interspecific hybridisation research showed that crossings between genotypes from different sections is possible. This can result in the introduction of Fusarium resistance in the overall cultivar assortment (see Chapter 4).
Research on the variation of the pathogen, testing different Fusarium -isolates against the most resistant lily genotypes, showed that a sustainable durability of the resistance can be expected. A deviating pattern of virulence was only found in some low aggressive isolates. These isolates appeared to affect L. longiflorum less than expected on the basis of their aggressiveness. The affection of these deviating isolates was, however, lower than using aggressive isolates. Some Fusarium -isolates of other formae speciales (f.sp. gladioli and f.sp. tulipae ) could also affect lily at a low level. This could have negative consequences for the crop rotation schedules used for bulb crops (see Chapter 5).
Selection at seedling level can substantially reduce costs and time. A seedling test for selection for Fusarium resistance in lily is described. Variation within and between populations was demonstrated. The seedling test resulted in a clear selection response, and this makes testing at a seedling level practical. The efficiency of the seedling test was determined on the basis of the percentage of escapes (selected susceptible plants) and missings (rejected resistant plants). The percentage of escapes was low compared to the number of seedlings tested, the percentage of missings on the other hand was rather high. This could be due to a high disease pressure. The percentage of escapes compared to the number of selected seedlings was such that retesting at clonal level is necessary. Genetic analysis of the results of the seedling test showed that the general combining ability (GCA) was the most important genetic component. The GCA-value of a cultivar was in agreement with the level of resistance at clonal level (see Chapter 6).
Indirect selection using molecular markers is independent of environmental conditions and can be carried out with a small amount of plant tissue. Using the Polymerase Chain Reaction (PCR) and Random Amplified Polymorphic DNA (RAPD), in descendants of the cross 'Connecticut King' x 'Orlito' three markers were traced, which were highly linked to Fusarium resistance. These markers, however, only explained a part (24 %) of the resistance. Linkage of markers to each other to obtain a genetic map of lily was hampered due to the marker type and the low number of the same plants used per marker (see Chapter 7).
The methods described in this thesis can be of a great importance for advanced breeding programmes of lilies. The importance of partial Fusarium-resistance under field conditions in an integrated disease control programme, in combination with a reduction of chemical disinfection and the influence of other pathogens, requires further attention (see General Discussion).