Ann. appl. Biol. (1996), 129:343-354 Printed in Great Britain 343

Control of Chenopodium album by soil application of Ascochyta caulina under greenhouse conditions

By C KEMPENAAR*, R WANNINGEN and P C SCHEEPENS DLO-Research Institute for Agrobiology and Soil Fertility (AB-DLO), P 0 Box 14,

6700 AA Wageningen, The Netherlands

(Accepted 5 September 1996)

Summary Effects of soil application of Ascochyta caulina spores on seedlings of

Chenopodium album and five cultivated plant species were investigated under greenhouse conditions as a part of a study on biological control of C. album.

Application of A. caulina spores to soils resulted in disease development on C. album and to a lesser degree on Spinacia oleracea seedlings, but not on Beta vulgaris subspecies vulgaris, Zea mays, Triticum aestivum and Pisum sativum seedlings. Affected C. album seedlings had an abnormal olive-green colour or necrotic spots on cotyledons and hypocotyls, and were stunted or died. Affected S. oleracea seedlings were pale in colour or had necrotic spots on the cotyledons, but did not die. Time courses of disease incidence and of mortality of C. album could be described by a monomolecular model. Effects of spore density, sowing depth, soil water content, soil type and time of sowing on disease development were examined. Disease incidence and mortality were influenced by spore density, soil water content and soil type, but not by sowing depth. S ores in a moist soil maintained infectivity at least 2 wk. Spore densities of 10 to lo1' spores mP2 were required for 50% mortality of emerged C. album plants.

Aspects of the development of A. caulina into a soil-applied mycoherbicide are discussed.

Key words: Biological control, mycoherbicide, soil treatment, weed control, integrated pest management

Introduction The perthotrofic fungus Ascochyta caulina (P. Karst.) v.d. Aa & v. Kest. is the cause of leaf

spots and stem necroses on various species of Chenopodium L. and Atriplex L. (Van der Aa & Van Kesteren, 1979). An asexual, pycnidial life cycle of the fungus is observed. A. caulina is native to Europe and Siberia. It is related to Ascochyta hyalospora (Cooke & Ellis) Boerema et al. which occurs on Chenopodium species in North America.

A. caulina is under investigation for use as a mycoherbicide against Chenopodium album L. (e.g. Kempenaar, 1995). Mycoherbicides are biological control agents developed from indigenous pathogenic fungi that normally remain at endemic levels (Templeton, 1992). Natural levels are increased by applying the respective fungi to the target weeds in inundative strategies. C. album is a weed species with an annual life cycle (Holm, Pluckett, Pancho &

*Present address: Department of Theoretical Production Ecology, Wageningen Agricultural University, Bornsesteeg 53, 6708 PD Wageningen, The Netherlands 0 1996 Association of Applied Biologists

344 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

Herberger, 1977; Van den Brand, 1985). It is a common weed on arable fields all over the world. Seeds of C. album can survive in soil for many years. Plants of C. album, which are erect herbaceous plants, can be strong competitors to crops. These traits, and the development of resistance to some chemical control agents, rank C. album in the top of lists of noxious weeds (Holm et al., 1977; Schroeder, Mueller-Schaerer & Stinson, 1993). Experiments with foliar application of A. caulina spores to young C. album plants showed promising control of the weed when free water for spore germination and infection was available for at least 12 h (Kempenaar, Horsten & Scheepens, 1996a,b). In the search for a possible method of application that is less dependent on the availability of water, effects of soil application of A. caulina on the control of C. album were studied. We assumed that the soil can be a buffer for the availability of water. A potential activity of soil applications ofA. caulina can be expected by looking at the life cycles of related Ascochyta species.

The objectives of this study were twofold: first, to investigate infection of C. album and five cultivated plant species by application of A. caulina spores to the soil, and second, to investigate the influence of spore density, sowing depth, soil water content, soil type, and time of sowing on disease development, factors that may have a constraint on disease development (Holocomb, 1982).

Materials and Methods

Plant material and plant growth conditions Seeds of C. album were collected from plants on arable fields at Wageningen, The

Netherlands, in 1992. They were stored in the dark at 5°C. Seeds of Beta vulgaris subspecies vulgaris cv. Carla (sugarbeet), Spinacia oleracea cv. Martine (spinach), Zea mays cv. Mandigo (maize), Triticum aestivum cv. Arminda (wheat) and Pisum sativum cv. Eminent (pea) were from commercial sources.

Prior to sowing, seeds were placed on coagulated water agar (1 g agar per litre of water) in 9 cm diameter Petri dishes, and incubated in a chamber with a day-night rCgime of 14 h light (20 pmol mpz s-'; Philips TL 8W/33) at 25<C, and 10 h darkness at 15"C, for 2 days. Seed coatings, if present, were removed by several washings in water. Several pre-germinated seeds were sown in a grid pattern into soils in plastic pots, 25 seeds pot-' for C. album or 16 seeds pot-' for the other plant species tested. Each pot had a soil volume of 600 cm3; the soil surface was 72 cm'. Depth of sowing was 5 mm for C. album seeds or 15 mm for the other plant species tested. The pots were placed in a greenhouse at 12-22"C, 65-90% relative humidity, 15-17 h day light period with an incoming photosynthetically active radiation of 20-30 mol m-2 day-'.

A sterilised sandy soil (sand 1) containing a mixture of coarse sand and quartz sand (1:l w/ w) was used in all experiments. The mixture contained very small amounts of nutrients and organic material. In one experiment, three other soils were also used; a sandy soil from an arable field at Wageningen (sand 2), a loamy clay soil from an arable field at Lelystad, The Netherlands, and a peat potting soil (Triomf no. 17; Trio b.v., The Netherlands). The water content of the soils was maintained at determined levels. Every 2-3 days, pot weights were determined, and additional water was added if needed.

Inoculation A. caulina was isolated from a leaf of a C. album plant on an arable field at Wageningen,

The Netherlands, in 1990. The isolate was maintained on coagulated oatmeal agar (60g

Control of C. album by soil application of A. caulina 345

oatmeal plus 20 g agar per litre of water) in test tubes in the dark at 5°C. Every 6 months, the isolate was inoculated on C. album plants, and re-isolated to prevent loss of virulence.

Inocula to be used in experiments were produced following a standardised three-steps procedure (Kempenaar, 1995). Step 1: fungal biomass from the culture in cold storage was inoculated onto plates of oatmeal agar in 9cm diameter Petri dishes, and incubated in a chamber with continuous light (75 pmol m-2 s-'; Philips TL 13W/83) at 20°C for 3 wk. Step 2: cultures on the inoculation 1 plates were flooded with sterile demineralised water (10 ml plate-'). After 3 h, spore suspensions were inoculated onto new oatmeal agar plates (lo6 spores plate-'). The plates were incubated for 2 wk. Step 3: cultures on inoculation 2 plates were flooded with demineralised water (10 ml plate-'). After 3 h, pycnidiospore suspensions were collected, filtered through cheese cloth, and adjusted to desired densities. Step 3 was carried out 1-2 h before spores were applied to the soil. Two application methods were investigated. With method 1, spores were applied and mixed into the soil before seeds were sown. The pots were partly filled with 500 cm3 soil, and then topped with a 1-cm layer of 100cm3 soil supplemented with 5 ml of spore suspension. With method 2, spores were applied to the surface of the soil after seeds were sown. Five ml of spore suspension was sprayed onto the soil surface of a pot using a DeVilbiss atomiser.

Disease assessment and data analysis Numbers of emerged, diseased and dead plants per pot were regularly assessed. Numbers of

diseased and dead plants per pot and per observation date were added up to give the number of affected plants per pot. Then, numbers were converted into proportions. Presented proportions of emerged plants per pot are relative to the number of seeds sown per pot. Presented proportions of affected plants per pot and dead plants per pot are relative to the number of emerged plants per pot.

A monomolecular regression model (Campbell & Madden, 1990) was used to describe disease incidence over time and plant mortality over time. In equation 1, y is a proportion of plants, r is a rate parameter, t is time after sowing and a is the intercept of the regression line with the x-axis.

(1) y = 1 - et-r(t-a)l

A logit-log regression model (Zadoks & Schein, 1979) was used to estimate spore densities that caused disease symptoms on 50% of the plants (ED50) and mortality of 50% of the plants (LDso). In equation 2, y is a proportion of plants, a is the intercept of the regression line with the y-axis, x is the spore density and b is the slope of the regression line. ED50 and LD50 were calculated with equations 3 and 4.

Logit(y) = a + b Log@) (2)

(4) L D ~ ~ = ~ O ~ - ~ L D / ~ L D }

Treatment effects on relative proportions were examined using a generalised linear regression model with a logistic link function (McCullagh & Nelder, 1989). Treatment effects on bLD, bm, and on log-transformed ED50, LD50 and plant dry matter were examined using analysis of variance. Regression analyses and analyses of variance were undertaken using Genstat 5 (Payne et al., 1987).

346 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

Host-specificity (Expt 1) Seeds of six plant taxons (see Plant material and plant rowth conditions) were sown into

sand 1 infested with a spore suspension (lo6 spores cm- ) or into uninfested sand 1. There were two pots per treatment. The experiment had a complete randomised design. The soil water content was 15% at sowing, and 12-18% afterwards. Disease development was monitored until the plants had four to six leaves. Seventeen days after sowing, some diseased plants were harvested for pathogen isolation. Small pieces of diseased tissue were surface- sterilised with a sodium hypochlorite solution ( 3 ml per litre of water) for 30 s, placed on oatmeal agar plates, and incubated as previously described. Cultures were compared with those of the original A. caulina isolate.

B

Effect of spore density (Expts 2a and 2b) Seeds of C. album were sown into sand 1 infested with spore suspensions (Expt 2a). Spore

density ranged from 0 to 2 x lo6 spores ern-'. There were five spore density levels. Each treatment was applied to five pots. The pots were arranged in a one-factor (spore density) randomised block design. In a separate experiment, spore suspensions were sprayed over sand 1 containing just seeds of C. album (Expt 2b). The experimental design was similar to that of Expt 2a. In both experiments the soil water content was 15% at sowing, and 12-18% afterwards. Plant emergence, disease incidence and plant mortality were determined every 3- 5 days until approximately 30 days after sowing. The proportions of emerged, affected and dead plants on the final days of observation were analysed statistically. The monomolecular model was fitted to data of each treatment. The logit-log model was fitted per block to data of the final days of observation.

Effect of sowing depth (Expt 3) Seeds of C. album were sown into sand 1 at a depth of 1, 5 or 10 mm. S ore suspensions

were sprayed onto the soil surface at the rate of 0 or 2x106 spores cm-: The soil water content was 15% at sowing, and 12-18% afterwards. The experiment had a two-factor (sowing depth and spore density) randomised block design with five replicates. The proportions of emerged, affected and dead plants 30 days after sowing were analysed statistically.

Effect of soil water content (Expt 4) Seeds of C. album were sown into sand 1 infested with a spore suspension. Spore density

ranged from 0 to lo6 spores cmP2 (four levels). Soil water contents were lo%, 15% or 18% at sowing, and average (min-max) soil water contents over a period of 3 wk were 9% (&lo%), 14% (13-15%) or 17% (16-18%), respectively. The experiment had a two-factor (soil water content and spore density) randomised block design with five replicates. The logit-log model was fitted per block to data of the final day of observation (28 days after sowing). The proportions of emerged plants, ED5,,, bED, LDS0 and bLD on the final day were analysed statistically.

Effect of soil type (Expt 5) Seeds of C. album were sown into soils (four types) infested with a spore suspension. Spore

density ranged from 0 to 3 x lo6 spores cm-' (six levels). Soil types were sand 1, sand 2, clay and peat and soil water contents were 15%, 15%, 35% or 70% at sowing for sand 1, sand 2, clay and peat, respectively, and 12-18%, 14-21%, 3 4 4 2 % or 60-75% afterwards, respectively. The experiment had a two-factor (soil type and spore density) randomised

Control of C. album by soil application of A. caulina 347

block design with four replicates. The logit-log model was fitted per block to data of the final day of observation (22 days after sowing). The proportions of emerged plants, ED50, bED, LDso and bm on the final day were analysed statistically. On the final day, plants that had not died were uprooted, washed, and dried at 105°C for two days. Plant dry weight was determined, and analysed statistically.

Effect of time of sowing (Expt 6) Infested and uninfested sand 1 were stored in the greenhouse at a determined soil water

content of 15%. Seeds of C. album were sown into the soils on the day the soil was infested (day 0), and 7 or 14 days later. Spore densities were 0 or 4x10’ spores cm-2. After sowing, the soil water content was 12-18%. The experiment had a two-factor (time of sowing and spore density) randomised block design with five replicates. The proportions of emerged, affected and dead plants 32 days after sowing were analysed statistically.

Results

Host-specificity Disease symptoms developed on seedlings of C. album and S. oleracea that emerged from

infested soil, but symptoms were far more severe on C. album than on S. oleracea. A. caulina was isolated from diseased C. album and S. oleracea tissues. Seedlings of B. vulgaris, 2. mays, T. aestivum and P. sativum cultivars that emerged from infested soil showed neither disease symptoms nor growth reduction.

On C. album, the first symptoms were observed a few days after emergence. Symptoms consisted of an abnormal olive-green coloured shoot or necrotic spots on cotyledons and hypocotyls. The olive-green plants often wilted and died within 1 wk after emergence. Necrotic spots on cotyledons of C. album were round to irregularly shaped and had a yellow- brown colour. Cotyledons of affected plants often curled downward and shrivelled. Necrotic spots on hypocotyls were round to oval shaped and had a grey-brown colour. Hypocotyl spots appeared most frequently near the soil surface. Spots that girdled the hypocotyl resulted in plant death. Pycnidia were observed on dead plant tissue.

On S. oleracea, the first symptoms were observed 2 wk after emergence. Affected plants were pale or carried irregularly shaped red-brown spots on cotyledons. Plant growth was retarded, but plants were not killed.

Effect of spore density Plant emergence was not influenced by spore density (P > 0.05) in either experiment; the

experimental mean proportions of emerged plants were 0.76 (Expt 2a) and 0.59 (Expt 2b). Proportions of affected plants increased with time for soils containing spores, and increased faster at higher spore densities (Figs 1A and 2A). The proportions of affected plants on the final days of observation were influenced by spore density (P < 0.001). Proportions of dead plants increased with time for soils containing spores, and increased faster at higher spore densities (Figs 1B and 2B). The proportions of dead plants on the final days of observation were influenced by spore density (P < 0.001). Regression analysis with the monomolecular model showed significant relationships between disease incidence and time, and mortality and time (P<0.05). Parameter a was not significantly influenced by spore density, and therefore was at determined values of 6 for disease incidence or 8 for mortality. Estimated values of parameter r and goodness-of-fits are given in Table 1. ED50 and LD50 on the final days of

348 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

1 .o

0.8

0.6

0.4

0.2

0.0 0 7 14 21 28 35

0.8 B

'I I I

0.6

0.4

0.2

0.0 'I I I

0 7 14 21 28 35

Time (days after sowing)

Fig. 1. Monomolecular curves describing average disease incidence (A) on, and mortality (B) of Chenopodrum album plants when Ascochyra caulina spores were incorporated into the soil at four densities (2 x lo6 - 0, 1 x 10' - 0, 1 x lo4 - A and 2 x lo3 - A spores cm-').

observation were 3 . 8 ~ 1 0 ~ or 1.9x103, and 4 . 5 ~ 1 0 ~ or 0 . 4 ~ 1 0 ~ spores cm-2 when spores were incorporated into the soil or sprayed onto the soil surface, respectively (Tables 2 and 3).

Effect of sowing depth Plant emergence was neither influenced by sowing depth nor by spore density (P > 0.05);

the experimental mean proportion of emerged plants was 0.60. The proportion of affected plants and proportion of dead plants 30 days after sowing (Table 4) were influenced by spore density (P < 0.001), but not by sowing depth (P > 0.05).

Control of C . album by soil application of A. caulina 349

Table 1. Rate parameters rd and r , of a monomolecular model fitted to data of disease incidence on, and mortality of Chenopodium album plants when Ascochyta caulina spores

were incorporated into the soil (Expt 2a) or sprayed onto the soil surface (Expt 2b) Disease incidence

Spore density f ?

Experiment (spores cm-’) rd SEa R2adj.b

2a 2 x lo6 0.124 0.013 82 1 lo5 0.119 0.013 86 1 x lo4 0.038 0.011 72 2 x lo3 0.0055 0.0006 54

2b 2 x lo6 0.545 0.024 54

2 lo4 0.0465 0.0039 55 2 1 6 0.121 0.012 71

2 lo3 0.0073 0.0006 46

a Estimates of standard errors. Percentages of variance accounted for by the regressions.

Mortality

rm

0.055 0.027 0.0061 0.0015

0.283 0.012 0.0042 0.0003

SE

0.006 0.004 0.0014 0.0005

0.012 0.0009 0.0009 0.0001

R*adj.

60 31 18 13

86 67 14 7

Effect of soil water content Plant emergence was neither influenced by soil water content nor by spore density (P >

0.05); the experimental mean proportion of emerged plants was 0.78. EDs0 and LD50 28 days after sowing (Tables 2 and 3) were influenced by soil water content (P < 0.001). The slope, parameter bLD was influenced by soil water content (P < O.Ol), but parameter bED was not (P > 0.05).

Effect of soil type Plant emergence was neither influenced by soil type nor by spore density (P > 0.05); the

experimental mean proportion of emerged plants was 0.78. ED50 and LD50 22 days after sowing (Tables 2 and 3) were both influenced by soil type (P < 0.001). Parameters bED and bLD were not influenced by soil type (P > 0.05). The experiment was terminated earlier than the other experiments because size-mediated plant competition was expected in some treatments. Plant dry weight 22 days after sowing (Fig. 3) showed an interaction between soil type and spore density (P < 0.001). Dry weight of plants grown on sand 1, sand 2 and clay

Table 2. Parameters a and b of a logit-log model fitted to data of disease incidence on Chenopodium album plants grown on soils infested with Ascochyta caulina spores, and spore densities that caused symptoms of 50% of the plants (ED5o), on final days of observation (days

after sowing)

Experiment Days after EDSO

Treatment sowing a b ( x lo3 spores cmp2) Spore density, 2a - Spore density, 2b - Soil water content 9%

14% 17%

Soil type Sand 1 Sand 2 Clay Peat

Standard errors of the means are in parentheses.

32 30 28 28 28 22 22 22 22

-6.9 (0.6) 1.9 (0.1) -7.6 (0.7) 2.4 (0.2)

-10.5 (0.5) 2.3 (0.1) -8.3 (0.6) 2.0 (0.1)

-10.0 (1.1) 2.6 (0.3) -8.4 (1.3) 2.4 (0.4) -6.0 (0.5) 1.6 (0.1) -6.5 (0.4) 1.5 (0.1) -8.3 (0.9) 2.0 (0.3)

3.8 (0.9)

42.4 (4.8) 18.2 (5.1) 6.5 (0.8) 3.5 (1.4) 5.8 (1.2)

23.8 (3.0) 14.8 (3.0)

1.9 (0.2)

350 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

0 7 14 21 28 35

1 .o

0.8

0.6

0.4

0.2

0.0 0 7 14 21 28 35

Time (days after sowing)

Fig. 2. Monomolecular curves describing average disease incidence (A) on, and mortality (B) of Chenopodium album plants over time when Ascochytn cnulinn spores were sprayed onto the soil surface at four densities ( 2 x lo6 - 0, 2 x lo5 - 0. 2 x lo4 - A and 2 x lo3 - A spores cm-*).

was significantly influenced by spore density over the whole range of spore densities, while dry weight of plants grown on peat showed a significant weight reduction only at the highest spore density tested.

Effect of time of sowing Plant emergence was neither influenced by time of sowing nor by spore density (P > 0.05);

the experimental mean proportion of emerged plants was 0.86. The proportion of affected

Control of C. album by soil application of A. caulina 35 1

Table 3. Parameters a and b of a logit-log model fitted to data of mortality of Chenopodium album plants grown on soils infested with Ascochyta caulina spores, and spore densities that caused mortality of 50% of the plants (LD50), on final days of observation (days after sowing) Experiment Treatment Days after U b LD50

sowing ( x lo5 spores cm-’1 Spore density, 2a - Spore density, 2b - Soil water content 9%

14% 17%

Soil type Sand 1 Sand 2 Clay Peat

32 30 28 28 28 22 22 22 22

-7.6 (0.9) 1.4 (0.2) -14.0 (0.6) 3.1 (0.1) -19.9 (2.1) 3.3 (0.4) -24.0 (2.4) 4.1 (0.5) -26.4 (2.3) 4.7 (0.4) -9.2 (0.7) 1.8 (0.1)

-11.3 (1.3) 1.8 (0.3) * *

-14.9 (0.6) 2.4 (0.1)

4.5 (1.2) 0.4 (0.1)

14.0 (2.7) 5.8 (1.3) 3.8 (0.3) 1.4 (0.1)

43.6 (23.8)

25.5 (5.5) *

Standard errors of the means are in parentheses. * No regression analysis done because mortality did not exceed 0.5; maximum mortality was 0.2.

plants 30 days after sowing (Table 5 ) was influenced by spore density (P < O.OOl), but not by time of sowing (P > 0.05). The proportion of dead plants 30 days after sowing was influenced by both spore density and by time of sowing (P < 0.001).

Discussion This study reveals a thusfar unknown infection route within the C. album - A . caulina

pathosystem. Spores of Axaulina in soils may cause infections of emerging plants, and infections may result in systemic disease development. We feel that the infection route is probably not an important step within the natural life cycle of A. caulina, but we lack knowledge for a sound evaluation. However, the importance of the observation is that this infection route may be utilised within a biological control strategy of C. album.

Disease development of C. album seedlings resembled disease development of mature C. album plants (Van der Aa & Van Kesteren, 1979; Boerema, Van Kesteren & Loerakker, 1985). The necrotic spots on cotyledons and hypocotyls probably resulted from infections during seedling emergence when plant tissues were exposed to spores. Cotyledon tissues

Table 4. Effect of sowing depth; disease incidence on and mortality of Chenopodium album plants grown on soils infested with Ascochyta caulina spores, 30 days after sowing

Spores density Proportion of Proportion of Sowing depth (spores cm-’1 affected plants dead plants

0.1 0 0.00 (0.00) 0.00 (0.00) 0.5 0 0.01 (0.01) 0.00 (0.00) 1.0 0 0.00 (0.00) 0.00 (0.00)

0.1 0.5 1 .o

2 x lo6 1.00 (0.00) 0.98 (0.02)

2 x lo6 1.00 (0.00) 0.92 (0.05) 2 x 106 1.00 (0.00) 1.00 (0.00)

Standard errors of the means are in parentheses.

352 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

0.04

h

0.03 L - - u E 3 0.02 - C

E

0.01

0 0 1 E2 1 E3 lE4 1E5 1E6 1E7

Spores cm-’

Fig. 3. Effect of soil type (Sand 1 - 0, Sand 2 - 0, Clay - A and Peat - A) and spore density of Ascochyta caulina in the soil on average plant dry weight of Chenopodium album plants 22 days after sowing. Error bars indicate standard errors of the means.

covered by the seed coats during emergence never showed necrotic spots. The abnormal olive-green colour of seedlings was not observed previously in this pathosystem. It points to the systemic activity of a toxin (Capasso et al., 1991). Disease development was not limited to C. album. S . oleracea was also affected by soil applications of A. caulina, but symptoms were less severe. This observation corresponds with results of a host-specificity test with older plants (Kempenaar, 1995).

Time courses of disease incidence and mortality could be described by the monomolecular model. The model was selected because it is convenient for description of time courses of disease incidence of soil-borne diseases (Zadoks & Schein, 1979). A sigmoidal model could be used as well, but overall, we obtained the best regressions with the monomolecular model. The regressions provided meaningful rate parameters. Parameter values of r greater than 0.1 can be considered high, and were estimated for the higher spore density treatments. Spore density in the soil had a large effect on both disease incidence and mortality of C. album. The infection rates were higher when more spores were applied to the soil. The method of spore application, incorporation into the soil or spraying onto the soil surface, had little effect on disease incidence and mortality. From the LDS0 values it can be concluded that spray application of spores provided greater plant mortality than soil incorporation. However, the estimated mortality rates showed that differences between the two application methods were small. Sowing depth did not significantly affect disease incidence and mortality in the range of 0.1-1 cm. Under field conditions C. album mainly emerges from depths of 0-0.5cm (Van den Brand, 1985). Soil water content, soil type and time of sowing had significant effects on disease incidence and mortality, but they did not exhibit limitations to the use of A. caulina as a soil-applied mycoherbicide.

Some of the potential of soil application of mycoherbicides for the control of weeds has been shown elsewhere. The first registered soil mycoherbicide, Devine’ for the control of Morreniaordonata, is a suspension of chlamydospores and mycelia of Phytophthorapalmi-

Control of C . album by soil application of A. caulina 353

Table 5. Effect of time of sowing; disease incidence on and mortality of Chenopodium album plants grown on soils infested with Ascochyta caulina spores, 30 days after sowing

Time of sowing relative Spores density Proportion of Proportion of to soil infestation (spores cm-’) affected plants dead plants

0 7

14

0 7

14

0 0.00 (0.00) 0.00 (0.00) 0 0.01 (0.01) 0.00 (0.00) 0 0.00 (0.00) 0.00 (0.00)

4 x lo5 1.00 (0.00) 0.99 (0.01) 4 x lo5 1.00 (0.00) 0.99 (0.01) 4 x lo5 0.95 (0.01) 0.69 (0.03)

Standard errors of the means are in parentheses.

vora. It showed sufficient control when applied to soils pre-emergence (Ridings, 1986). Promising control by soil applications of potential mycoherbicides has been demonstrated for at least six other pathosystems (Abbas, Boyette, Hoagland & Vesonder, 1991; Boyette, Abbas & Connink, 1993; Boyette, Templeton & Oliver, 1984; Vogelsang, Watson & Hurle, 1994; Walker, 1981; Weideman & Templeton, 1988). The active ingredients of the (potential) mycoherbicides were spores plus mycelium, either in a suspension or in a granulate, indicating some of the options of how weed pathogenic fungi can be applied to soils.

Host specificity and efficacy of control are the two major criteria that determine the potential of a mycoherbicide (Charudattan, 1989). If A. caulina is further developed into a soil-applied mycoherbicide, more plant species would have to be tested for a sound evaluation of host specificity. If S. oleracea is considered in a rotation, a detailed risk analysis has to be conducted. This analysis requires information on survival of A. caulina in soils. Based on the data presented, we are optimistic about the potential of A. caulina as a soil-applied mycoherbicide. High levels of control were achieved under a broad range of greenhouse conditions. Approximately lo9 to lo1’ spores mP2 were required for 50% mortality of emerged C. album plants. These densities are high but not unrealistic. Sublethally infected plants were retarded in growth considerably, and would be less competitive than healthy plants (Massion & Lindow, 1986; Kempenaar et al., 19963). We feel that if an economic method can be developed for increasing the amount of inoculum of A. caulina in soils, the method has potential in current weed control strategies.

Acknowledgements This research was carried out in the frame work of the Multi Year Crop Protection Plan

(Meerjarenplan Gewasbescherming) of the Dutch Ministry of Agriculture, Nature Manage- ment and Fisheries. We acknowledge the Ministry for the financial support. We also acknowledge J Kartalska and P F J M Horsten for technical assistance, and J C Zadoks for comments on this paper.

References Abbas H K, Boyette C D, Hoagland R E, Vesonder R F. 1991. Bioherbicidal potential of Fusarium

Boerema G H, Van Kesteren H A, Leerakker W M. 1985. Vermeldenswaardige schimmelaantasfin- moniliforme and its phytotoxin, Fumonisin. Weed Science 39:673-677.

gen in de periode 1980-1984. Gewasbescherming 16:103-119 (In Dutch).

354 C KEMPENAAR, R WANNINGEN AND P C SCHEEPENS

Boyette C D, Abbas H K, Connink W J. 1993. Evaluation of Fusarium oxysporum as a potential bioherbicide for sicklepod (Cassia obtusifolia), Coffee Senna (C. occidentalis), and Hemp Sesbania (Sesbania exaltata). Weed Science 41:678-681.

Boyette C D, Templeton G E, Oliver L R 1984. Texas Gourd (Cucurbita texana) control with Fusarium solani fsp. cucurbitae. Weed Science 32:649-655.

Campbell C L, Madden L V. 1990. Introduction to Plant Disease Epidemiology. New York: John Wiley & Sons. 532 pp.

Capasso R, Evidente A, Randazzo G, Vurro M, Bottalico A. 1991. Analysis of Cytochalasins in culture of Ascochyta spp. and in infected plants by high performance liquid and thin layer chromatography. Phytochemical Analysis 2:87-92.

Charudattan R 1989. Assessment of efficacy of mycoherbicide candidates. In Proceedings of the Seventh International Symposium on Biological Control of Weeds, pp. 455464. Ed. E S Delfosse. Rome: Instituto Sperimentale per la Patologia Vegetale (MAF).

Holm L G, Pluckett D L, Pancho J V, Herberger J P. 1977. Chenopodium album L. In The World’s Worst Weeds (Distribution and Biology), pp. 84-91. Honolulu: University Press of Hawaii.

Holocomb G E. 1982. Constraints on disease development. In Biological Control of Weeds with Plant Pathogens, pp. 61-72. Eds R Charudattan and H L Walker. New York: John Wiley & Sons.

Kempenaar C. 1995. Studies on biological control of Chenopodium album by Ascochyta caulina. Ph.D. Thesis, Wageningen Agricultural University, The Netherlands. 128 pp.

Kempenaar C, Horsten P F J M, Scheepens P C. 1996 . Spore germination and disease development of Chenopodium album after application of Ascochyta caulina as a mycoherbicide. European Journal of Plant Pathology 102:143-153.

Kempenaar C, Horsten P F J M, Scheepens P C. 19966. Growth and competitiveness of Chenopodium album after application of Ascochyta caulina in field crops. Weed Science 44:609-614.

Massion C L, Lindow S E. 1986. Effects of Sphace[otheca holci infection on morphology and competitiveness of Johnsongrass (Sorghum halepense). Weed Science 34:883-888.

McCullagh P, Nelder J A. 1989. Generalized Linear Models. London: Chapman & Hall. 511 pp. Payne R W, Lane P W, Ainsley A E, Bicknell K E, Digby P G N, Harding S A, Leech P K, Simpson

H R, Todd A D, Venier P J, White R P, Gower J C, Tunnicliffe Wilson G, Paterson L J. 1987. Genstat 5: Reference Manual. Oxford: Clarendon Press. 749 pp.

Ridings W H. 1986. Biological control of stranglervine in citrus - A researcher’s view. Weed Science

Schroeder D, Mueller-Schaerer H, Stinson C A S. 1993. A European weed survey in 10 major crop systems to identify targets for biological control. Weed Research 33:449-458.

Templeton G E. 1992. Potential for developing and marketing mycoherbicides. In Proceedings of the First International Weed Control Congress, pp. 264-268. Melbourne, Australia.

Van den Brand W G M. 1985. Biologie en ecologie van melganzevoet (Chenopodium album), Rapport 47 (in Dutch). Lelystad (NL): Research Station for Arable Farming and Field Production of Vegetables.

Van der Aa H A, Van Kesteren H A. 1979. Some pycnidial fungi occurring on Atriplex and Chenopodium. Persoonia 10: 267-276.

Vogelsang S, Watson A K, Hurle K. 1994. The efficacy of Phomopsis convolvulus against field bindweed (Convolvulus arvensis) applied as a pre-emergence bioherbicide. Zeitschrift f i r Pflantzenkrankheiten und Pflantzenschutz XIV:253-260.

Walker H L. 1981. Granular formulation of Alternaria macrospora for control of Spurred anoda (Anoda cristata). Weed Science 29:342-345.

Weideman G J, Templeton G E. 1988. Efficacy and soil persistance of Fusarium solani f.sp. cucurbitae for control of Texas Gourd (Cucurbita texana). Plant Disease 72:36-38.

Zadoks J C, Schein R D. 1979. Epidemiology and Plant Disease Management. New York: Oxford University Press. 427 pp.

(Received 29 April 1996)

34 (Suppl. 1):31-32.