Ecological aspects of seed health testing

The purpose of this study was to investigate the factors affecting the outcome of incubation methods used to test seeds for the presence of fungal pathogens. The reason for starting such an investigation was the poor reproducibility of the results of those methods.Chapter 3 starts with the observation that for a number of infections lower percentages are found on blotters with a relatively high moisture content than on blotters with a lower moisture content. This so-called 'wet blotter effect' (WBE) was found to be related to the speed of water uptake by the seeds; if the water uptake is relatively slow the effect does not show, whereas in the case of more rapid water absorption the effect can be noted. If the rate of water uptake of seeds without WBE was increased by presoaking them in water, a similar decrease was noted ('soak effect' or SE). WBE and SE could both be nullified by the addition of antibacterial antibiotics, and so were attributed to antagonism by saprophytic seed-borne bacteria.Factors stimulating bacterial development (higher temperature and pH, addition of nutrients or bacteria) were found to increase WBE and SE; factors stimulating fungi more than bacteria decreased or eliminated it. If one of both groups (bacteria or fungi) obtains the initial lead it will dominate and reduce the expression of the other group. Elimination of bacterial antagonism results in higher percentages of both pathogenic and saprohytic fungi found. The use of antibiotics eliminates bacterial antagonism as a source of variation.Bacterial antagonism in the agar method is often less important due to low pH of the medium. This offers an explanation for the higher agar than blotter percentages found for such an infection as Botrytis cinerea in flax seed. The hypochlorite pretreatment customary for the testing of most seeds by the agar method also restricts bacterial antagonism. Bacterial antagonism in agar tests may not only lower infection percentages found but also change the colony appearance. The addition of a small quantity of an antibiotic to the agar medium eliminates it altogether.The importance of the WBE and SE may vary with the sample, due to differences in the bacterial flora of the seeds. Plating soaked seeds on agar media yields hardly any pathogen, but agar media with antibiotics which were routinely used in all further work, more or less nullify the previous SE. Soaking in terramycin solution often gave increased percentages of the pathogen.The same factors stimulating bacterial antagonism in blotter tests also favour it in sand and soil.Chapter 4 starts with the observation that in case of Phoma valerianellae infection of corn salad seeds blotter tests with 2,4-D gave remarkably higher percentages than normal blotter tests. This led to the realization of the importance of the factor 'host vigour' in incubation methods. Various other treatments such as water soak, exhaustion, mutilation and freezing gave similar results. The reduction (2,4-D), elimination (freezing) or circumvention (agar method) of host vigour often lead to higher percentages being found for seed-borne infections. It also offers an explanation for the difference between blotter and agar test results for some infections. Reduction of the host vigour causes a shift from symptoms of the disease to signs of the pathogen to be looked for at inspection.The freezing of pre-incubated seed followed by a post-freezing incubation period (freezing method) was found to be promising for a number of infections. For some infections (e.g. Helminthosporium spp.) it simplifies inspection, no seedling structures being produced. Provided bacterial development is checked by means of antibiotics, fungi grow easier and more abundantly, probably due to increased leachage of nutrients from seeds. Fungal infections can therefore be more easily recognized and identified, and higher infection percentages may be found. Some seeds which are rich in easily available nutrients (e.g. wheat) must be protected from air-borne contamination. In the freezing method the trays with paper can be replaced by plastic sheets with depressions filled with liquid (containing e.g. antibiotics) for the seeds.In chapter 5 it is observed that interfungal antagonism becomes more prominent when bacterial antagonism and host vigour are either not important or have been eliminated. It is consequently especially a problem in the agar method. Interfungal antagonism is influenced by incubation conditions, e.g. temperature. Its reduction by the use of hypochlorite pretreatment is rather non-selective, eliminating also part of the pathogen, although it may be all right for such deep-seated infections as Ascochyta spp. in pea and Phoma betae in beet seeds.More selective is the 'dry heat' pretreatment advocated by MALONE (1962) for H.avenae in oat seeds tested by the agar method. Its use was extended to other Helminthosporium spp. in oat and barley seeds but has the disadvantage of increased Mucor development in subsequent agar tests. In blotter tests with dry heat pretreatment Mucor presented no problem, but the results were not superior to those of the freezing method.A highly selective method is the use of peptone-PCNB agar for Fusarium spp. In this method neither bacterial and fungal antagonism nor host vigour play a role, so that very high infection percentages may be obtained. Tests carried out by this method are easy to prepare and to inspect.Another way to combat interfungal antagonism is the use of growth-restricting agents, such as oxgall, in agar media. This offered good prospects for the slow growing Septoria nodorum in wheat seed, which is otherwise easily suppressed by other fungi. The use of oxgall eliminates the need of surface disinfection which is an advantage as light infections of S.nodorum can be important.Chapter 6 deals with measurement of the severity of infection. Very light Fusarium infections of single seeds may be considered to be of little consequence. Their elimination by hypochlorite pretreatment can therefore be accepted and may be used as a correction of the very sensitive peptone-PCNB agar method. This led to considerations on the importance of inoculum potential and threshold level of infection. Seed health data are usually given in terms of percentage of infected seeds. Possible ways to correct them for inoculum potential were indicated, such as the use of pretreatments of standardized strength. This could perhaps compensate for the elimination of such factors as antagonism and host vigour.Chapter 7 describes experiments in which a large number of wheat seed samples containing Fusarium spp. and Septoria nodorum were tested in various ways. The results were then compared and correlation coefficients between them were computed. Although the latter were low, they were generally significant. The correlations of germination capacity with soil emergence decreased with temperature due to the presence of F. nivale. This pathogen only reduces emergence at lower temperatures, being of no consequence at higher ones. S.nodorum did not significantly reduce emergence in the samples used, although at lower temperatures it caused more abnormal seedlings to be produced. S. nodorum was always highly correlated with coleoptyle lesions.The blotter test for Fusarium spp. was shown to be very unreliable. The symptoms are hard to judge so that errors of judgment are easily made the presence of the different Fusarium spp. is not equally well indicated and other fungi e.g. S. nodorum may interfere by producing the same symptoms. The blotter test for S.nodorum is only good for a number of cultivars that produce the characteristic protuberance symptoms. For other cultivars the symptoms to be used are highly unreliable, especially if Fusarium spp. are present, which usually is the case. The agar method is therefore to be preferred, preferably without surface disinfection, but with e.g, oxgall added to the medium.It is suggested that due to the complex nature of the factors involved, which are moreover unpredictable the usefulness of correlation coefficients for seed-borne infections is very limited. Their main use may be in the comparison of various methods with seed ling performance in conditioned soil tests.In chapter 8 it is concluded from the preceding that the incubation conditions applied in seed health testsdo not only affect the expression of the pathogen in a direct way. They also affect the pathogen indirectly via the seed biotope (seed microflora and host seed or seedling). The seed biotope varies with the sample so that the most important source of variation is presented by the ecology of the germinating seed itself. Elimination and circumvention of these biotic factors, which also decreases the chances of errors of judgment, seem to offer the best prospects for the development of simple, cheap, fast and reproducible methods.Application of correction factors for inoculum potential may be important for a number of infections (e.g. Fusarium spp. in cereal seeds). Tolerances if used may be regarded as a correction for local agricultural conditions and should be based on the methods used for testing.