As for insects (See Host Specificity Studies), preliminary information about the host specificity of a potential agent pathogen in its country of origin might be obtained by searching the literature, particularly crop protection literature, checking herbarium collections and records, and checking with workers in agriculture to determine whether the pathogen is a known pest of crops.

In the absence of other preliminary information, host range tests against a small selection of plants, usually closely related to the target weed, should be done in the country of origin. This will determine whether further, more detailed, testing is warranted.

Supply of test plants

Since testing for Australian weeds is carried out overseas, it is necessary to provide the laboratory carrying out the work with test plants. Where possible test plant seed is sent to the country where the testing is to be carried out. This is the easiest and cheapest method as there are usually fewer problems with quarantine requirements. However, in some cases it is necessary to grow the plants before despatch to the testing agency. In this case, the plants must be able to withstand reasonably rough handling, including being turned upside down, and exposure to extremes of temperature. A packaging method for sending plants overseas which has proven to be reliable involves sealing each plant in its own container which is lined with absorbent paper to prevent sweating and breakdown of the foliage in transit, and tightly packing the individual containers into a polystyrene box which is protected by an outer layer of strawboard.

Methodology

Batches of 4-5 test species were screened at any one time. Three to four plants were included for each species and only vigorously growing (flushing) plants were used. A range of leaf types (young to senescent) were inoculated by brushing their lower leaf surfaces with a dense urediniospore suspension ('saturation' inoculum, ca. 5 000 000 spores/mL) in sterile distilled water and 0.01% Tween 20, using a fine camel hair brush. In the case of aquatic plants (with stomata only on the upper leaf surface) and leafless, fleshy asclepiads, the sites of inoculation were modified accordingly.

Rubber vine plants (C. grandiflora, C. madagascariensis var. madagascariensis and var. glaberrima) inoculated with a spore concentration of 15 000 000 spores/mL  to avoid defoliation, were used as controls for each test. Plants were maintained for 24 hours at previously determined optimum conditions for rust infection (ca. 23ºC, 100% RH). Inoculated plants were then transferred to a greenhouse chamber at 25ºC and 50-60% RH during 12-hours light, and 20ºC and 60-80% RH during the night.

Plants were observed over a three-week period, i.e. more than double the latent period for sporulation of the rust on control rubber vine plants. Samples of all inoculated leaves were removed at 10 and 21 days for clearing and staining (Bruzzese and Hasan 1983). 

Samples from fleshy leaves and stems were cut longitudinally to facilitate subsequent microscopic examination of the inoculated area. In addition to the light microscope examination, all inoculated material was inspected for external symptoms of infection using a stereoscopic microscope.

Each test species was screened on at least two separate occasions, and any plant showing an unusual macro- or micro-reaction to the rust was investigated further. Symptomatology was assessed according to 15 categories (See Table 1) and the susceptibility to infection of each test species was rated based on systems devised by Kochman and Goulter (1983) and Bruzzese and Hasan (1983) (See Table 2).

Assessment categories within any one test species were, for the most part, consistent. However, variable reactions occurred occasionally between replications, although most frequently within the same treatment. Leaf age was the main factor governing variation in susceptibility ratings and the full range of reactions are shown in Table 2. Weak or non-dominant reactions are represented by parentheses.

Results of the host range tests for plants within the subfamilies Periplocoideae and Secamonidea are shown in Table 3 as examples.

Note: Parenthesis indicate variable but replicable symptoms between individual plants in a test.

All species and varieties within the genus Cryptostegia were highly susceptible to infection by Maravalia cryptostegiae. The only other species exhibiting susceptibility to M. cryptostegiae were the Madagascan Gonocrypta grevei and the Australian Cryptolepis grayi. Further testing of M. cryptostegiae types against G. grevei indicated that there are at least two distinct physiological races or pathotypes of M. cryptostegiae, one adapted to Cryptostegia, the other to Gonocrypta. Cryptolepis grayi showed varying levels of susceptibility to the rust, ranging from resistant to moderately susceptible, and the appearance of fertile pustules on one plant of the first test run warranted further investigation. Fertile pustules were observed in 3 of 14 plants in further tests, despite development of host cell plasmolysis and haustorial inhibition which restricted development of sporogenous tissues by the rust. Results of the screening suggest that Cryptolepis, like Gonocrypta, is a generic host of M. cryptostegiae. Decisions based on these tests are discussed by McFadyen and Heard (See Deciding on Release Candidates).

[ Back ]