Modern agriculture is highly dependent on the use of chemical pesticides to control plant pathogens. Fungicides and fumigants commonly have drastic effects on the soil biota, as they are intentionally applied at much higher rates than herbicides and insecticides ( Fraser 1994). These methods are time-consuming and uneconomical, pollute the atmosphere, and are environmentally harmful, as the chemicals build up in the soil ( Nannipieri 1994). Furthermore, the repeated use of such chemicals has encouraged the development of resistance among the target organisms ( Goldman et al. 1994 ). This has resulted in the use of ever-increasing amounts of pesticides and has prompted the search for new strategies of pest control to reduce or eliminate the use of pesticides ( Cook & Granados 1991; Lorito et al. 1994 ). For instance, integration of biocontrol agents with reduced doses of chemical agents has the potential to control plant pathogens with minimal impact on the environment ( Chet & Inbar 1994).
Trichoderma spp. have been extensively studied as potential biocontrol agents (e.g. Lynch 1990; Papavizas 1992). However, some studies have also shown that Trichoderma spp. can stimulate the growth of a number of vegetable and bedding plant crops (e.g. Baker 1988; Lynch et al. 1991a,b ). Lynch et al. (1991a,b) investigated the effect of Trichoderma on the growth of lettuce, and its ability to control damping-off diseases caused by Rhizoctonia solani and Pythium ultimum. They investigated whether a number of Trichoderma strains had a direct effect on lettuce establishment and growth in the absence of pathogens. It was found that the fungal treatments reduced the emergence time of seedlings compared to the controls. From their results, and those of Ousley et al. (1994) , they concluded that specific Trichoderma strains have the potential to consistently increase plant growth ( Lynch et al. 1991a ). The prospects for control of Pythium damping-off of lettuce with Trichoderma, Gliocladium, and Enterobacter spp. (E. cloacae) were also studied. The bacterial antagonists did not perform as consistently as the fungi. Furthermore, the view of synergism between fungal and bacterial antagonist shown by Kwok et al. (1987) was not shared by Lynch et al. (1991b) . However, they demonstrated that, to thoroughly eliminate the adverse effect of the damping-off pathogen, a threshold level of T. harzianum antagonist was needed.
Ousley et al. (1993) investigated the effect of autoclaving plant growth promoting Trichoderma isolates on the germination and growth of lettuce. Some strains turned out to be good plant growth promoters even after autoclaving. It was inferred that seedling growth promotion by Trichoderma is a balance between growth inhibiting and growth promoting properties, and autoclaving some strains alters this balance.
Few commercial preparations of Trichoderma are available for controlling plant diseases ( Coley-Smith et al. 1991 ), possibly because the nonbiological methods of control used at present are more reliable. For Trichoderma preparations to provide realistic alternatives to chemical control, their efficiency and reliability need to be improved and knowledge of their ecology and effects on the soil ecosystem needs to be enhanced. The integration of the use of biocontrol and plant growth promoting strains of Trichoderma may be a useful asset in the commercialization and marketing of such Trichoderma isolates. Therefore, the assessment of strains for both of these assets in vivo is essential. Besnard & Davet (1993) found a number of Trichoderma strains that were simultaneously plant growth promoters (tomato and cucumber) and biocontrol agents (Pythium); however, they used a sterile potting medium and did not use fresh soil in their experiments.
Wild type isolates of Trichoderma with known biocontrol activities are assessed here for their effect upon pea growth and their antagonistic activity against large Pythium ultimum inocula; the effects of the Trichoderma inocula upon the indigenous soil microflora and soil enzyme activities in the presence and absence of Pythium were also assessed.
MATERIALS and METHODS
The soil used was sandy loam of the Holiday Hills series, taken from Merrist Wood Agricultural College (Surrey), and had been under permanent pasture for at least 15 years. The analysis of the soil, conducted at the University of Surrey, was pH 5·4, particle ratio 10 : 9 : 81 clay : silt : sand, and organic matter content 1·6% by weight.
Trichoderma strains were collected from various sources and tested for biocontrol activity in plate assays against Pythium. The experiment described below was conducted twice, the first with 10 Trichoderma strains chosen from plate assays, replicated three times (results not shown) and the second with five strains chosen from the first experiment, replicated five times (data presented here). The five strains selected were: Trichoderma harzianum strains TH1 (IMI 275950), T4 (IMI 298372), T12 (IMI 298373) and N47 (IMI 288110), and Trichoderma pseudokoningii strain To10, which was obtained from M. Welland of Abies seeds and is available in the University of Surrey culture collection.
Material from stock cultures was grown on potato dextrose agar (PDA) at 25 °C for 3 d (primary plates). Mycelial discs (5 mm diameter) from primary plates were then used to inoculate secondary PDA plates, from which spore suspensions were prepared after 7 days’ growth at 25 °C. Suspensions were prepared by detaching the spores from the surface of the colonies into sterile distilled water using a glass spreader. The concentration of each Trichoderma strain was adjusted to 106 spores ml−1 by dilution and direct counting using a haemocytometer. The spores of each Trichoderma strain were suspended in a 0·75% guar gum solution in which the pea seeds were imbibed. Control seeds were imbibed in sterile guar gum.
Pythium ultimum (IMI 308273) was obtained from CABI Biosceince. Material from stock cultures was grown on plates of PDA at 25 °C for 3 d (primary plates). Four 5-mm disks were cut and placed in a flask containing 95 g of sand, 5 g of organically grown porridge oats and 20 ml of distilled water, all previously autoclaved twice. The flasks were incubated at 25 °C for 3 weeks before being homogenized in a blender and mixed with coarsely sieved soil at a concentration of 3%.
Pythium inoculated or uninocuated soil (150 g) was placed in experimental microcosms consisting of 210-mm high acetate cylinders, slotted between the top and base of plastic 90 mm diameter Petri dishes, creating semienclosed microcosms ( Naseby & Lynch 1998). Each Trichoderma treatment and controls were replicated five times in the presence and absence of Pythium. Each microcosm consisted of eight imbibed seeds, planted at a depth of approximately 1 cm below the soil surface. Twenty-five millilitres of water was added to each microcosm before they were placed in a random design into a growth chamber (Vindon Scientific) set at a 16-h photoperiod with a day/night temperature regime of 21 °C/15 °C, respectively. The relative humidity was maintained at 70%. The experiment was conducted twice: first with 10 Trichoderma strains, replicated three times (results not shown), and second with five strains chosen from the first experiment, replicated five times (data presented here).
Sampling and analysis
After 21 days’ growth, the microcosms were harvested, and the following plant measurements were made: the plant shoot and root weight (wet and dry), number of nodules, length of each root, the number of lateral roots and the number of lesions. Rhizosphere soil (closely associated with the plant roots) samples were collected and stored at 4 °C. The samples were subsequently assayed for soil acid and alkaline phosphatase, urease, β-glucosidase, N-acetyl glucosaminidase, cellobiosidase and chitobiosidase, by the methods of Naseby & Lynch (1997).
A 1-g fresh root sample was taken from each replicate and macerated in 9 ml of sterile quarter-strength Ringers solution using a pestle and mortar. One gram of rhizosphere soil from each replicate was also suspended in 9 ml of sterile quarter-strength Ringers solution. Filamentous fungal populations were quantified by plating a 10-fold dilution series of each root macerate or soil suspension onto 10% malt extract agar containing 100 μg g−1 streptomycin and 50 μg g−1 rose bengal. Plates were incubated at 20 °C for 5 d before enumeration. P1 medium ( Katoh & Itoh 1983) was used for the enumeration of root-indigenous fluorescent Pseudomonas; plates were incubated at 25 °C and enumerated after 5 d growth. Tryptone soya agar (10%) was used for the enumeration of total culturable bacteria.
VP agar ( Lumsden et al. 1990 ), based on potato dextrose agar, was used to enumerate Pythium, and contained the following supplements: vancomycin (200 mg l−1), pimaricin (10 mg l−1), penthanodichlorobenzene (100 mg l−1), streptomycin (50 mg l−1) and rose bengal (2·5 mg l−1). The media of Papavizas & Lumsden (1982) was used to enumerate Trichoderma spp., and consisted of 500 ml of V8 (vegetable juice) centrifuged with 7·5 g calcium carbonate, from which 200 ml of the supernatant was made up to 1 l with distilled water and 15 g of agar was added. After autoclaving, the following antibiotics were added to the media: 100 μg g−1 of neomycin sulphate, bacitracin, penicillin, and chrononeb, 25 μg g−1 of chlorotetracycline hydrochloride, 20 μg g−1 of nystatin, 500 μg g−1 of sodium propionate and alkylaryl poliether alcohol (2 ml l−1).
Data were analysed using SPSS for Windows (SPSS Inc.) by means of a one-way anova, and subsequently differences between treatments (multiple comparisons) were determined using least significant differences (LSD).
Biocontrol and plant growth
Pythium inoculation reduced the emergence of pea seedlings ( Table 1) and this effect was significantly suppressed by all the Trichoderma strains except strain To10, with strains T4 and N47 providing the greatest control.
Emergence, wet shoot and root weights (g) of pea plants as affected by Trichoderma
inocula and Pythium ultimum
Trichoderma strains TH1, T4, N47 and T12 significantly increased the wet shoot weights by 51%, 41%, 45% and 13%, respectively, compared with the Pythium-treated control ( Table 1). Trichoderma strains TH1 and T4 resulted in a significantly greater wet root weight in the presence of Pythium compared with the Pythium control ( Table 1). Similarly, inoculation with Trichoderma strains TH1 and T4 as well as strain T12 resulted in significantly greater dry root weights ( Table 2). Pythium inoculation significantly increased the wet and the dry shoot/root ratio ( Tables 1 and 2), indicating a dramatic effect on root growth. Strain N47 significantly increased the wet shoot/root ratio compared with the Pythium control.
Dry shoot and root weights (g) of pea plants as affected by Trichoderma
inocula and Pythium ultimum
Pythium inoculation significantly reduced the root length, the number of lateral roots and nodules ( Table 3). Inoculation with all the Trichoderma strains significantly reduced the number of lesions caused by Pythium and all but strain To10 increased the root length. All the Trichoderma inocula significantly increased the number of lateral roots in comparison with the Pythium control, but did not significantly affect the number of nodules per root system.
Mean root length (cm) and number of lateral roots, nodules and lesions per root system as affected by Trichoderma
inocula and Pythium ultimum
Plant growth promotion
In the absence of Pythium, strains TH1 and T12 significantly increased the emergence of pea seedlings compared with the control, which is also consistent with the increased emergence with Pythium present ( Table 1). Strain N47 resulted in similar pea emergence both in the presence and absence of Pythium, and significantly increased the wet shoot weight by 15% in the absence of Pythium. However, none of the Trichoderma strains significantly affected the dry weight ( Table 2). Strain T4 and N47 resulted in significantly greater wet root weights than the control, by 21% and 8%, respectively. Trichoderma strain T4 had a significantly lower wet shoot/root ratio than strains N47, TH1 and T12.
Trichoderma strains TH1 and N47 significantly increased the root length, the number of nodules per root system and the number of lateral roots per root system, in the absence of Pythium ( Table 3), whilst strain To10 significantly reduced the number of lateral roots.
Pythium inoculation significantly increased the root and rhizosphere soil bacterial populations and the root fluorescent Pseudomonas population ( Table 4). All the Trichoderma inocula reduced the soil total bacterial populations in the absence of Pythium; however, only strain TH1 reduced the soil bacterial population in the presence of Pythium.There were no significant differences in the fluorescent Pseudomonas populations among the Trichoderma treatments in the absence of Pythium. The only effect of Trichoderma in the presence of Pythium was with the inoculation of strain To10, which resulted in a significant increase in the Pseudomonas population.
Log total soil and root bacteria and fluorescent Pseudomonad populations as affected by Trichoderma
inocula and Pythium ultimum
Pythium inoculation significantly increased the root and rhizosphere soil fungal populations ( Table 5). The root fungal populations in the absence of Pythium were significantly greater than the control with the inoculation of all the Trichoderma strains except strain TH1, whilst strains TH1, T4 and T12 significantly increased the soil fungal population in the absence of Pythium. The rhizosphere soil and root Trichoderma populations were greater in the Pythium-treated soil than in noninfested soil ( Table 5). The Trichoderma populations of the two controls (without Trichoderma inoculation) were below the level of detection of around 103 cfu g−1. The root and rhizosphere soil Trichoderma populations did not significantly differ with the different inocula in the absence of Pythium. However, strain To10 resulted in the greatest root Trichoderma population, and strain T4 the greatest rhizosphere soil population, in the presence of Pythium. Inoculation with Trichoderma strains T12, N47 and T4 resulted in significantly lower Pythium populations than the control and strain TH1 ( Table 5).
Log fungal populations as affected by Trichoderma
inocula and Pythium ultimum
Soil enzyme activities
There were significantly higher enzyme activities in soil infected with Pythium ( Table 6), in most cases by a factor of 3 or more. Inoculation with a number of Trichoderma strains significantly reduced the effect of Pythium. Strains TH1, T4, N47 and T12 significantly reduced β-glucosidase, NAGase and chitobiosidase activities relative to the Pythium control. Strains TH1, T4 and T12 significantly reduced the alkaline phosphatase activity, whilst strains TH1, T4 and N47 significantly reduced the urease activities, relative to the Pythium control. However, strain To10 did not affect any of the enzyme activities relative to the Pythium control and none of the strains affected the acid phosphatase and cellobiosidase activities relative to the Pythium control. The only significant effects with respect to the control in the absence of Pythium were increases in NAGase and β-glucosidase activities with the inoculation of strain To10.
Soil enzyme activities in the rhizosphere of pea plants inoculated with Trichoderma
and Pythium ultimum
lant growth measurements were used to assess the potential impact of the different inocula on crop production. Pythium ultimum had a significant effect on all the plant measurements, which is in accordance with the fact that it is a major pathogen of pea and is most destructive at the seedling stage ( Kommedahl et al. 1981 ). Pythium inoculation reduced pea emergence, and this effect was suppressed by all the Trichoderma strains except strain To10, with strains T4 and N47 providing the greatest control. These results are in concordance with the work of Cook (1994), who suggested that the ability of Pythium to rapidly colonize the host plant before other fungi is an essential part of their pathogenicity. Therefore, seed inoculation with Trichoderma can dramatically reduce the effect of Pythium.
All the Trichoderma strains except strain To10 significantly improved the growth of plants in the presence of Pythium and significantly reduced the damaging effect of the pathogen. Trichoderma strains TH1, T4 and N47 had the greatest overall effects, resulting in increased plant weights and root lengths, and reduced root lesions. This is corroborated by previous work where strain TH1 was shown to control Pythium damping-off of lettuce ( Lumsden et al. 1990 ), and increase both plant stand and plant fresh weight ( Migheli et al. 1994 ).
The fact that Trichoderma strains T4, N47 and T12 significantly reduced the Pythium population demonstrates that these Trichoderma strains have antagonistic activity towards the pathogen, which is related to an improvement in plant production. Baker (1989) suggested that the time-course of mycoparasitism indicates that mycoparasitism of Pythium cannot be the mechanism of antagonism when the agent is applied to seed, and the production of an antibiotic (‘routing factor’) by Trichoderma spp. was the causal factor in biological control of Pythium spp. However, three (TH1, N47 and T12) of the five strains used retarded the growth of Pythium by the production of soluble diffusable metabolites and subsequently totally overgrew and mycoparasitized the Pythium in plate assays (data not shown). The other two strains (To10 and T4) did not totally overgrow the Pythium but retarded the growth of Pythium by the production of soluble diffusable metabolites. This suggests that the reduction in the Pythium population is related to the mycoparasitic activity of the Trichoderma rather than the production of metabolites.
Plant growth stimulation (in the absence of Pythium) has been shown with a number of Trichoderma strains ( Ousley et al. 1994 ) on other plants such as lettuce, petunia and marigold. In this study, the Trichoderma strains were preliminarily selected for biocontrol properties before plant growth stimulation. It is evident from these results that inoculation with strains TH1, T4, N47 and T12 resulted in varying degrees of plant growth promotion, whilst strain To10 had, overall, a slight detrimental effect on plant growth. Strain N47 was the most consistent plant growth promoter, resulting in significant increases in most of the plant growth measurements reported, and was the only strain to increase the plant shoot weight.
The increased growth response of plants caused by T. harzianum depends on the ability of the fungus to survive and develop in the rhizosphere ( Kleifield & Chet 1992). A possible mechanism for increased plant growth is an increase in nutrient transfer from soil to root, which is supported by the fact that Trichoderma can colonize the interior of roots ( Kleifield & Chet 1992). Many microbial treatments of plants have in the past been shown to have positive or negative influence on plant growth ( Lynch et al. 1991a ). However, Ousley et al. (1993) found that autoclaving Trichoderma inocula did not remove growth-promoting properties and suggested that seedling growth promotion by Trichoderma could be a balance between growth inhibition and growth promotion properties, with the balance altered in some strains by autoclaving. The slight inhibitory effect of Trichoderma strain To10 may be related to the production of volatile pentyl and pentenyl-pyrones by T. harzianum ( Lumsden et al. 1990 ) which, besides being fungistatic and effecting biocontrol action ( Tables 1 and 2), can have phytotoxic side-effects at high doses.
Inoculation with strain T4 resulted in a large increase in root weight, but did not effect the shoot weight, which resulted in a large decrease in the shoot/root ratio. The only other strain to increase the root weight was strain N47, which also increased the shoot weight and therefore did not affect the shoot/root ratio. The conversion into shoot/root ratio has been used extensively in the past ( Clark & Reinhard 1991) and has been suggested to be an indicator of plant stress, whereby the lower the shoot/root ratio (or the higher the root/shoot ratio) the more stressed the plant. Causes of plant stress include nutrient limitations (including oxygen, Drew & Lynch 1980), and therefore a decrease in shoot/root ratio may indicate such stress, although pathogenic effects do not necessarily comply with this rule. It should be recognized, however, that such stressed plants may be more effective in acquiring water and nutrients as a result of the expanded root system. Thus this is a positive adaptive response to such stresses and this could be a useful trait in low-nutrient or dry soils.
The increased root and rhizosphere soil bacterial populations and the root fluorescent Pseudomonas population in the presence of Pythium is due to the pathogenic effect of the pathogen causing nutrient leakage from the root. Only strain TH1 reduced this effect, which is due to this strain causing the greatest reduction in the effect of the pathogen on plant growth, indicating that the pathogen caused less root damage in the presence of strain TH1. All the Trichoderma inocula reduced the soil total bacterial populations in the absence of Pythium, however, this effect may be due to the Trichoderma inocula and fungal population taking up an increased proportion of the soil niche, indicated by increased soil fungal populations. Therefore, there would be increased competition for limiting nutrients, leading to a decrease in the bacterial population in comparison with the control and an increase in fungi such as Trichoderma, which are strong soil colonizers.
Fluorescent Pseudomonads are natural biocontrol agents found in the soil, and these results indicate that the introduction of the Trichoderma strains (without Pythium) did not adversely affect these useful populations. The only effect of Trichoderma in the presence of Pythium was with the inoculation of strain To10, which resulted in a significant increase in the Pseudomonas population. This is likely to be a result of the slightly deleterious effect of this strain causing increased root leakage/damage, which allows a greater population of aggressive rhizosphere/root colonizers such as fluorescent pseudomonads.
The increase in root and rhizosphere soil fungal populations in the presence of Pythium again is due to the root damage caused by the pathogen and the resulting nutrient leakage. The increases in root and soil fungal populations with most Trichoderma inocula in the absence of Pythium are in part due to the addition of the inocula. This is indicated by the Trichoderma populations being undetectable in the controls (below 103 cfu g−1 root or soil) and therefore, the Trichoderma inocula had an additive effect on the total fungal populations. The lack of an increase in the fungal population with the TH1 inocula in the absence of Pythium, and an actual decrease in the presence of Pythium, may be related to the potent antagonistic properties of this strain against a wide range of fungi ( Lynch 1987).
The fact that the Trichoderma populations of the two controls (without Trichoderma inoculation) were below the level of detection allowed direct quantification of the different inocula without the aid of genetic markers. The greater rhizosphere soil and root Trichoderma populations in Pythium-treated soil is again due to the increase in available carbon sources from the damaged roots. All the strains were recovered from the soil and root at similar levels in the absence of Pythium, indicating similar colonization abilities in this soil. However, strain To10 resulted in the greatest root Trichoderma population in the presence of Pythium, which may be related to this treatment having the lowest root mass and therefore the greatest damage and carbon leakage.
The microbial population results revealed that the soil infected with Pythium had significantly greater bacterial and fungal populations than uninfected soil. This is likely to be due to increased leakage of carbon compounds from the diseased roots. The results also show that microbial population measurements can be indicative of the extent of damage caused to the plant by the pathogen.
Measurement of soil enzyme activities may be useful for gaining a greater understanding of the nature of perturbations caused to ecosystem function, and has been used as an indicator of the effect of microbial inoculation ( Naseby & Lynch 1998). Soil enzyme activities have also been used as an indicator of carbon leakage from roots ( Naseby & Lynch 1998; Naseby et al. 1999 ). The large increase in enzyme activities found in the presence of Pythium therefore indicate a dramatic increase in carbon and nutrient leakage from roots due to root damage. The C cycle enzyme activities such as β-glucosidase, NAGase and chitobiosidase are directly related to carbon availability, whereas P cycle enzyme activities are inversely related to P availability ( Tabatabai 1982; Tadano et al. 1993 ). In conditions of high C availability, such as root leakage, P is a more limiting nutrient, and demand increases, resulting in an increase in phosphatase activity ( Naseby & Lynch 1998). Therefore, the Pythium must have caused a decrease in the available phosphate, thus causing an overall increase in activity. The decrease in available P took the form of an increase in the available carbon in the rhizosphere (by root leakage). Urease activity (N cycle) was also increased by Pythium infection, which again indicates increased C availability, as shown by Naseby et al. (1999) .
The trend found in the alkaline phosphatase activity was not repeated with the acid phosphatase activity. However, the majority of the acid phosphatase activity may be of a different origin to the alkaline phosphatase. Acid phosphatase is mostly of plant and associated fungal origin ( Tarafdar & Marschner 1994), whereas the alkaline phosphatase is more likely to be of microbial origin. If this is the case, then the effects of the inocula upon acid and alkaline phosphatase can, in some circumstances, be independent. This is supported by the work of Naseby & Lynch (1997), where rhizosphere acid phosphatase did not show significant differences with the inoculation of bacteria or addition of substrates, and did not show a trend with soil depth. The acid phosphatase activity would be more dependent upon the nutritional status of the plant; therefore the damage caused by Pythium, resulting in the loss of C from the roots, would reduce the P requirement of the plant.
Strains TH1, T4, N47 and T12 reduced the effect of the Pythium on soil enzyme activities by varying degrees. All four strains reduced the β-glucosidase, chitobiosidase and NAGase activities, indicating a reduction in carbon leakage from the root. Furthermore, the reduction in alkaline phosphatase activity with strains TH1, T4 and T12 also indicates a reduction in available C, as P is less limiting with these treatments than with the Pythium control. Urease activity (N cycle) was also reduced by strains TH1, T4 and N47 in relation to the Pythium control, which again indicates a comparative reduction in available C. The reduction in a number of enzyme activities by strains TH1, T4, N47 and T12 therefore indicates a reduction in plant damage and subsequent C leakage caused by Pythium, and is also related to increases in plant growth described earlier with these strains. This is supported by the fact that inoculation with strain To10 resulted in similar enzyme activities to the Pythium control, and this strain did not offer any protection to the plant, as shown by the plant growth measurements.
As few of the enzyme activities were affected by most of the Trichoderma strains in the absence of Pythium, it follows that these inocula did not have a large effect on the nutrient status of the rhizosphere or nutrient exudation/leakage from the root ( Naseby & Lynch 1998). The only strain to increase the β-glucosidase activity was strain To10, which also increased the NAGase activity; this effect may be due to a slight detrimental property of this strain, as shown by the decrease in the number of lateral roots and a small insignificant decrease in the root weight. Therefore, strain To10 may have damaged the root system, causing a slight increase in C leakage, which increased C cycle enzyme activities. The enzyme activity results demonstrate that such measurements are sensitive indicators of the effect of Pythium on plant roots and of the protective effect of the Trichoderma.
Overall, strain N47 had the greatest beneficial characteristics, as it consistently improved the pea growth measurements in the absence plant pathogens and also had an antagonistic effect against Pythium ultimum which reduced the damage caused to the pea plants by the pathogen. Strain T4 also improved several of the plant growth measurements and had good biocontrol properties, whilst strain TH1 was the best biocontrol agent. The dual properties of these strains improve the commercial application, giving them an advantage over single action inocula, especially in the absence of plant pathogens.