Mycosphaerella graminicola (Septoria tritici blotch)

Infected leaves show necrotic, usually grey-green to off-white patches studded with brown or black dots about 0.5 mm across, arranged in lines. On mature leaves, these patches are initially more or less limited by veins, so that they appear rectangular, and vary in size from 1 x 2 mm to 3 x 10 mm; however, many such patches may become confluent. On juvenile leaves and glumes, necrotic patches have irregular edges. As populations build up on tissues, entire leaves may become necrotic. When this happens, the pigmentation, size and density of pycnidia is reduced, and a hand-lens may be needed to see them.

M. graminicola causes sporadic but serious losses, on both durum and bread wheats, in many wheat growing regions of the world: Europe, North and South America, Africa, Asia, Australia and New Zealand (Shipton et al., 1971). The disease is especially prevalent in Western Europe and around the Mediterranean basin (Eyal and Ziv, 1974), and is currently considered the major threat to European wheat crops (Kema and Verstappen, 1999). In the last 25 years M. graminicola has increased in importance in the UK; from a random survey of 300-400 crops, the pathogen was found at a frequency of 2% in 1976 and by 1988 this had risen to 86% (Polley and Thomas, 1991). This rise was due, in part, to the introduction of more susceptible varieties (Bayles, 1991). Particularly severe attacks of M. graminicola on crops of winter wheat are correlated with higher than normal rainfall in May and June (Polley and Thomas, 1991). Hence, the pathogen is especially prevalent in the wetter areas of the UK: from most to least, Wales - southwest - southeast - east - west midlands - east midlands - north - Yorkshire/Lancashire (Polley and Thomas, 1991). Total yield losses to species of Septoria (including Stagonospora nodorum) are estimated at over 9 million tons worldwide (Kolomiets, 1999). A worldwide survey indicated that the disease decreases linearly with increasing distance from the equator and with increasing non-growing season precipitation, but increases linearly with increasing phosphate applications and length of growing season (Leath et al., 1994). Adequate rainfall is generally needed for epidemics to develop, and probably accounts for the sporadic nature of disease losses (Shaw and Royle, 1993). Under drought, moderate amounts of disease (1-5% severity) may increase yields by increasing water use efficiency (Shtienberg, 1991). Early sowing and cool temperatures, slowing plant growth, increase yield losses by allowing more time for infection and disease development (Shaw and Royle, 1993; Lovell et al., 1997). Nitrogen (N) applications have varying effects but generally increase disease severity (Lovell et al., 1997). Under UK conditions, the severity of M. graminicola post-anthesis increases with increasing rate of N fertiliser applied at an average of 11% per 100 kg/ha N (translating to a potential yield loss of 5% for each 100 kg/ha N given as fertiliser) (Leitch and Jenkins, 1995). Yield losses are mainly due to poor grain filling (Cornish et al., 1990), and are determined by the severity of attack on the flag leaf and sub-flag leaf (Shaw and Royle, 1993). Greenhouse trials in Europe have indicated that early season disease may reduce dry matter production more than late infections, even though the flag leaf has not emerged at this time (Adolf et al., 1993). Early infections mainly reduce the grain number per ear, whereas late infections mainly give a lower 1000-grain weight (Adolf et al., 1993). However the evidence from fungicide trials on time of application, suggests that infection of the top two leaves is needed for economically significant yield losses. In poorly established crops severe early attacks may affect tillering and prevent full recovery of the plant population (Mielke, 1981). As well as reducing quantity, severe attacks can result in shrivelled grains (Ziv and Eyal, 1978), reducing milling and baking quality (McKendry et al., 1995) and thus market value (Mehta et al., 1979). Models have been constructed to predict yield losses from disease severity. However, it is difficult to generalise as the cultivars used and husbandry employed (e.g. sowing date, fertiliser and fungicide regimes etc.) and local environmental conditions can all strongly influence disease severity and subsequently disease losses (Ziv and Eyal, 1978). With this proviso in mind, a few models from different areas in the world are given below. For northwestern European conditions, Shaw and Royle (1989) constructed a model based on 2 years of field experiments with artificial inoculations of cv Longbow. This took place in the south west of England, and the model was verified in a third year with naturally infected wheat. L = 0.012 (0.003) g1 <sup>0.5 + 0.004 (0.002) g2 0.5

Where L = yield loss rate (t per ha per °C-day); g1 and g2= square root severity on the flag leaf and leaf 2, respectively, over the normal lifetime of each leaf, measured in thermal time.

In field experiments at 2 locations and over 3 years in Arkansas, USA, regression was used to determine the relationship of leaf blotch severities to yield and test weight losses (Milus, 1994). Average yield losses caused by Septoria leaf blotch were 0.43, 0.47 and 0.32% for each 1% increase in leaf blotch severity on cultivars Florida 302, Rosen and Caldwell, respectively (Milus, 1994).

From data from 2 seasons (1973-75) with 4 spring wheat cultivars in Israel (Ziv and Eyal, 1978):

Percentage loss in yield = 0.7111x, where x = percentage area affected at GS 71 of top 3 leaves.

Yield losses caused by M. graminicola are summarized as follows:

Africa

Morocco, Schlutter and Javati (1976) report an 18% yield loss based on results from trials.

Asia

Israel, Ziv and Eyal (1978) report 30-50% yield loss in trials with 3 cultivars and 3 isolates of the pathogen during the period 1973-75.

Former USSR, Derecha and Kostymets (1977) report a 6-16% yield loss.

Australasia

Australia, Brennan and Murray (1988) report a yield loss of $152 million without fungicides, $59 million with fungicides in 1988, estimated from national losses. In New South Wales (1988) a yield loss of 8% was estimated from national losses (Brennan and Murray, 1988). In Victoria (1970-1990s), Loughman and Thomas (1992) and Brown and Paddick (1980) reported up to 31% yield losses in trials.

New Zealand, Thomson and Gaunt (1986) reported yield losses of 18% in winter wheat trials with 1 cultivar in 1979.

Europe

Denmark, Jorgensen et al. (1996) reported yield losses of up to 25% in trials.

England and Wales. Estimated winter wheat losses for 1970-75 and 1985-89 are reported by King (1977) and Cooke et al. (1991), respectively. King reports losses of 0.1-7.4% while Cooke et al. report a yield loss of 2% (or 0.329 Mt per annum or £34.5 million based on a price of £105 per t). In estimates from 51 organic farms in England and Wales in 1991, Yarham and Turner (1992) report a yield loss of 4.4%, 88% of samples being infected.

Germany, Mielks (1981) reports yield losses of 17-21% in trials with 17 cultivars.

Netherlands, Forrer and Zadoks (1983) report a yield loss of 878 kg per ha in winter wheat trials in 1980.

Southern Romania, Ionescu-Cojocaru and Rugina (1986) reported a yield loss of up to 25% in farm surveys during 1981-85. Up to 98% of samples were infected.

Spain (Andalusia), Aguirre-Berruezo et al. (1988) report yield losses of 39-46% in trials in 1983-84.

Ukraine, Kolomiets (1999) report yield losses of up to 50% in winter wheat.

North America

Canada, Gilbert and Tekauz (1992) report a yield loss of up to 20%.

USA, Ziv and Eyal (1978) report yield losses of up to 50%. In Illinois, Jacobsen (1977) report yield losses of 15-20% in winter wheat trials in 1974-75. These losses include those for Leptosphaeria nodorum and Gibberella zeae.

South America

Brazil, Mehta et al. (1979) report yield losses of up to 40% in 1973-76 in trials at 5 locations with 3 cultivars.

Chile, Gilchrist and Madariage (1980) estimated yield losses of 2-10% in 1980 based on national losses.

Uruguay, Zamuz et al. (1970) report yield losses of 16-60% in 6 trials, each with 25 cultivars.</sup>