Page Banner

United States Department of Agriculture

Agricultural Research Service

Related Topics

Blind Seed Disease

Methods for Detection and Assessment

Postharvest Disease Detection and Assessment

Early methods of blind seed detection involved the direct observation of seed. Gemmell (1940) detected infection by looking for small pinkish spots on dehusked seed under a binocular microscope illuminated by direct light on a white background. At Lincoln College in New Zealand, the usual procedure was to place 100 paled seeds under magnification and examine them for infection (Blair 1947), although removing the lemma or palea to examine the caryopsis can be tedious. Sproule and Wright (1966) developed a manually operated apparatus to facilitate the removal of lemma and palea.

Infected seeds generally appear more opaque than healthy seeds. A diaphanoscope was used to differentiate infected and healthy seeds based on opacity (Noble 1939, Glasscock 1940, Hyde 1945, Muskett 1948). However, opaque seeds can also occur if the seed is weathered before threshing, in which case opacity increases due to pigmentation (Gemmell 1940, Calvert and Muskett 1945, Muskett 1948).

For estimation of total infection, Hyde (1945), Blair (1947), and Matthews (1980) believed that direct observations were not as reliable as placing seed in water and looking for spores under the microscope. A magnification of 100× is suitable for examination for conidia of G. temulenta (Calvert and Muskett 1945). The lemma and palea may be removed (Calvert and Muskett 1945, Hyde 1945, Sproule and Wright 1966) or left intact (Kolk and Rennie 1978). Kolk and Rennie (1978) soaked seed for 4 hours; Matthews (1980) soaked seed for at least 2 hours.

The number of seeds considered to provide an accurate estimate of rate of infection was reported as 100 (Calvert and Muskett 1945, Blair 1947, de Tempe 1966), 200 (Hyde 1945, Matthews 1980), or 500 (Muskett 1948). Matthews (1980) referred to the soaking and examination of seed as the "soaking test." Matthews also performed a "droplet test," in which 100 seeds were individually soaked in drops of water on microscope slides for at least 4 hours. The drops were examined at 100× and classified subjectively as having light, moderate, or heavy spore concentration. However, Matthews did not find a significant correlation between the droplet test and ungerminated seed.

Rose (1945) correlated conidial numbers removed through soaking samples of 100 seeds with germination rate, but high variability in the number of conidia prevented accurate prediction of germination. Hardison (1957) mixed 18 ml of seeds and 18 ml of water in 250-ml flasks, soaked the seeds for 20 minutes, then counted conidia in a 0.0063-mm3 hemacytometer chamber. The number of conidia per 0.0063 mm3 corresponded to five infection classes ranging from trace to heavy. One to three conidia per 0.0063 mm3 corresponded to a trace infection level, and more than 30 conidia corresponded to a heavy infection. Alderman (1999) used a similar seed-washing procedure and established a linear relationship between the number of conidia washed from a standardized seed sample and the percentage of infected seed.

Matthews (1980) described a detection method based on production of apothecia. In this test, 200 seeds were scattered over moist perlite in 14-cm-diameter petri dishes. The dishes were placed in plastic bags and stored at 5 °C for 12 weeks. Normal germinated seeds were removed. Dishes were transferred to 20 °C under a 12 hour light/12 hour dark cycle for a further 4-5 weeks. Seeds with apothecia were recorded. This procedure estimates the potential inoculum from seed, but since many infected seeds do not produce apothecia, the total number of infected seeds is greatly underestimated.

The number of seeds infected with viable G. temulenta can be assessed by isolating the pathogen on nutrient media. In this test, the palea are removed from the seeds, the caryoposis is surface sterilized and bisected, and the halves are plated on malt-extract agar (Neill and Hyde 1942, Calvert and Muskett 1945, Muskett 1948).

Preharvest Testing

Preharvest testing of blind seed was common during the 1940s in New Zealand (Scott 1974) to determine if the ryegrass seed crop should be harvested. Greenall (1943) sampled seed heads 2 weeks before harvest and found good correlation between the percentage of seed not infected (healthy seed) and germination of machine-dressed seed. However, samples should be taken within 1 week of cutting (Hyde 1942, 1945; Lithgow and Cottier 1953; Munro 1978; Alderman 1988, 1991b). Infection can occur up to the time of cutting, so samples collected too early could underestimate postharvest infection levels.

The number of seed heads believed to be representative of the area was reported as 50 (Wade 1949), 300 (Hyde 1945, Osborn 1947), or 400-500 (Lithgow and Cottier 1953; Alderman 1988, 1991b).

previous pagenext page 

United States Department of Agriculture
Agricultural Research Service

The material on this page is in the public domain.

Original posting: October 2001.

Last Modified: 8/12/2016
Footer Content Back to Top of Page