MATERIALS AND METHODS

Preparation of Plant Extracts
Acacia gourmaensis A. Chev (bark) and Eclipta alba (L.) Hassk. (whole plant) were collected in 2006 at Kamboinsé (Burkina Faso). Collected materials were air-dried and then reduced into powder using a mortar. The powders were sieved and only particles smaller than 2.0 mm were used. In each case, 25 g of powder were mixed with 100 mL of distilled water and the mixture was incubated at 25-30 °C for 20 h. Extracts were subsequently filtered with cheesecloth and centrifuged at 3000 rpm for 5 min to remove plant debris. Finally, supernatants were filtered with a coffee filter before use.

Two commercial chemicals (Benlate T20 and Apron star 42 WS) were used as positive controls for sorghum and pearl millet, respectively. Benlate T20 was composed of 20% Benomyl and 20% Thirame while Apron star 42 WS contained 20% Thiamethoxam, 20% Metalaxyl-M and 20% Difenoconazole.

Sorghum and Pearl Millet Seed Samples
Sorghum cultivar Sariaso 11 and pearl millet cv. IKMP 5 were used in this study. Seed samples of these two cultivars were collected from farmers` saved-seed lots in 2006. Preliminary tests indicated that collected seeds were infected by the following fungi: Fusarium moniliforme Sheldon (syn F. thapsinum Klittich, Leslie, Nelson et Marasas sp. nov.), Curvularia lunata (Wakk.) Boedijn, Phoma sorghina (Sacc.) Boerema, Dorenbosch and Van Kesteren, Colletotrichum graminicola (Ces.) Wilson (syn C. sublineolum P. Henn) and Exserohilum rostratum (Drechsler) Leonard and Suggs.

Seed Treatment
Naturally-infected seeds of sorghum and pearl millet were first immersed for 20 h in plant extracts at 25-30 °C in the dark. Then they were air-dried for another period of 20 h. Control treatments included non-treated seeds, seeds immersed in distilled water and seeds treated with commercial chemicals i.e., Benlate T20 (5 g kg^-1 of seed) and Apron star (2.5 g kg^-1 of seed).

Testing the Effect of Aqueous Extracts on Seed Health
For each crop species, 200 seeds per treatment were tested for seed health using the standard blotter method reported by Mathur and Kongsdal (2003). Seeds were sown on moistened blotters in Petri dishes (25 seeds per Petri dish) and incubated for 7 days at 20-25 °C under alternating cycles of 12 h of near-ultraviolet light and 12 h of darkness. Incubated seeds were examined individually and the growing fungi were recorded.

Testing the Effect of Aqueous Extracts on Seedling Emergence, Seedling Mortality and Plant Growth
For each crop species, 200 seeds per treatment were sown in plastic pots containing sterilized soil (25 seeds per pot). The soil was saturated with water before sowing. The pots were placed at 25-30 °C under alternating cycles of 12 h daylight and 12 h of dark. The pots were watered regularly to maintain the moisture. The number of emerged seedlings as well as that of subsequently dead seedlings was recorded 15 Days After Sowing (DAS). To evaluate plant growth, seedling fresh weight was recorded at 30 DAS.

Field Condition Tests
The effect of the aqueous extracts on seedling emergence, seedling mortality, plant vigour and grain yield was evaluated by sowing the treated seeds and growing plants under field conditions. In total, three trials for sorghum and four trials for pearl millet were conducted in farmers` fields. All trials were implemented using a randomised complete bloc design with four replicates. In each plot, mineral fertilizer consisting of Nitrogen-Phosphate-Potassium (NPK 14-23-14) was applied at 100 kg ha^-1 at sowing. Urea (50 kg ha^-1) was also applied 30 days post-emergence. In any case, seeds were sown in 6 m-long rows (6 rows par plot) with distances of 0.80 m between rows and 0.40 m between holes in the same row. Data were recorded from the four central rows of each plot. These included:

Statistical Analyses
The data recorded were analysed by ANOVA procedures using Statistical Analysis System, version 8. Significant differences between treatment means were determined using Least Significant Difference (LSD) (Gomez and Gomez, 1984).

RESULTS

Laboratory Tests
Laboratory tests were aimed at assessing the effects of seed treatment with aqueous plant extracts on the control of seed contamination by fungi, the emergence and mortality of seedlings and the seedling growth.

Treatments of Sorghum Seeds
Plant extracts and the commercial chemical Benlate T20 had significant effects on controlling seed infection by P. sorghina (Table 1). Only distilled water gave results which were similar to those from non-treated seeds. Infections by P. sorghina were reduced by 27.0, 72.0 and 64.0%, respectively with extracts of A. gourmaensis and E. alba and the chemical control Benlate T20. A. gourmaensis and Benlate T20 also lowered C. graminicola incidence on seeds by 68.5 and 75.1%, respectively. Among the treatments, Benlate T20 was the only one effective against C. lunata (93% reduction of incidence). None of the treatments had significant effect on F. moniliforme. Compared to non-treated seeds, seedling emergence and mortality were reduced by the chemical control only. Treatment of seeds with Benlate T20 lowered seedling emergence by 15.8% but preserved all the emerged seedlings from death. Seed treatment with A. gourmaensis and E. alba stimulated the weight of the seedlings. Seedling growth was most stimulated by extract of A. gourmaensis which led to the heaviest seedling weight.

Treatments of Pearl Millet Seeds
Extracts of A. gourmaensis and E. alba effectively controlled seed infection by all fungi, except E. rostratum. Compared to non-treated seeds, seed infections were reduced at rates ranging between 69 to 86.4% and 52.1 to 72.0%, respectively (Table 2). The chemical Apron star was effective in controlling seed infection by F. moniliforme and P. sorghina only. None of the treatments showed significant effects on seedling emergence and mortality. However, as observed in sorghum, the treatment of seeds with A. gourmaensis and E. alba increased seedling weight, the most biomass promoting treatment being that with A. gourmaensis.

Field Trials
Field trials were conducted to assess the effect of seed treatment with plant extracts and chemical fungicides on emergence and mortality of seedlings, plant vigour and grain yield.

Table 1:	Effect of treatment of sorghum seeds with aqueous plant extracts on incidence of four seed-borne fungi, emergence and mortality of seedlings and plant biomassa
aTested fungi were Colletotrichum graminicola, Curvularia lunata, Fusarium moniliforme and Phoma sorghina; Means in columns followed by the same letter(s) are not significantly different (p = 0.05) according to LSD test; b Percentage of reduction compared to non-treated seeds; c - Non evaluated

Table 2:	Effect of treatment of pearl millet seeds with aqueous plant extracts on incidence of four seed-borne fungi, emergence and mortality of seedlings and plant biomassa
Tested fungi were Curvularia lunata, Exserohilum rostratum, Fusarium moniliforme and Phoma sorghina; Means in columns followed by the same letter(s) are not significantly different (p = 0.05) according to LSD test; bPercentage of reduction compared to non-treated seeds; c- Non evaluated

Table 3:	Effect of treatment of sorghum seeds with aqueous plant extracts on emergence and mortality of seedlings, plant vigour and grain yield
aPlant vigour ratings based on a 1-10 scale, where: 1 = poor and 10 = excellent; Means in columns followed by the same letter(s) are not significantly different (p = 0.05) according to LSD test

Table 4:	Effect of treatment of pearl millet seeds with aqueous plant extracts on emergence and mortality of seedlings, plant vigour and grain yield
aPlant vigour ratings based on a 1-10 scale, where: 1 = poor and 10 = excellent; Means in columns followed by the same letter(s) are not significantly different (p = 0.05) according to LSD test

Effects of Treatments of Sorghum Seeds
Seed treatment with plant extract and chemical Benlate T20 had significant effects on emergence of seedlings (F = 2.85; p = 0.04). Treatments of the seeds with extracts of A. gourmaensis and E. alba resulted in 80.2 and 70.6% of seedling emergence, which were significantly higher than the proportion of seedlings emerged from non-treated seeds (66.4%) (Table 3). The chemical Benlate T20 had no marked effect on seedling emergence. In all treatments, seedling mortality was low (0.6 to 1.6%) and no effect of the treatments were noticed (F = 0.79; p = 0.56). By contrast, plant vigour ratings were significantly higher in seedlings emerged from all treated seeds (7.1 to 8.3) than in the case of non-treated control seeds (F = 14.17; p<0.01). Overall, plant yield was also significantly affected by treatments of the seeds (F = 2.61; p<0.05). However, only treatments with extracts from E. alba induced a clear increase in grain yield. With these treatments, average yield reached 2.5 t ha^-1 while in all other cases yields did not exceed 2 t ha^-1.

Effects of Treatments of Pearl Millet Seeds
As indicated in Table 4, treatments of pearl millet seeds with plants extracts and commercial chemical Apron star resulted in significant effects on emergence (F = 2.86; p = 0.04) of seedlings, plant vigour (F = 8.42; p<0.01) and grain yield (F = 2.01; p<0.05). Proportions of emerged seedlings from treated seeds of pearl millet varied between 80.4 to 88.6% and only the commercial chemical control performed significantly better than the non-treated seeds. Effects of seed treatments on seedling mortality were less apparent (F = 2.92; p = 0.06) and the highest mortality (9.1%) was induced by treatments with plant extract from E. alba. More clear-cut seed treatment effects were observed on plant vigour. All seed-treating products significantly induced higher plant vigour compared to the non-treated seeds. Moreover, the commercial chemical Apron star performed better than the two plants extracts. As for grain production, only seed treatments with plant extract from E. alba and Apron star gave yields which were significantly higher than that of plants from non-treated seeds.

DISCUSSION

Low seed quality is a serious problem in crop production, especially in most developing countries. For instance, the use of infected seeds without any treatment against infecting pathogens is often responsible for the decrease in seedling emergence (Hofs et al., 2004). This effect may even continue to act on plant growth, therefore affecting plant vigour and yield. Fungi such as Fusarium moniliforme, Phoma sorghina, Curvularia lunata, Colletotrichum graminicola and Exserohilum rostratum were frequently isolated from many sorghum and pearl millet seed samples collected in different locations of Burkina Faso (Zida et al., 2008). In the present study, aqueous extracts of A. gourmaensis and E. alba showed effectiveness in controlling several sorghum and pearl millet seed-borne fungi. The fungus P. sorghina was controlled by both plant extracts in sorghum seeds. However, plant extract from A. gourmaensis had more fungicidal properties because it was the only one capable to control C. graminicola. The development of C. lunata was efficiently inhibited by the chemical Benlate T20 only. This result indicated that Benlate T20 was more effective than the two plant extracts in controlling sorghum seed-borne fungi.

By contrast, plant extracts were more efficient than the chemical Apron star in the control of fungi in pearl millet seeds. Treatments of the seeds with Apron star had no significant effect in inhibiting C. lunata, in the contrary of plant extracts. F. moniliforme was efficiently controlled by plant extracts in pearl millet seeds but not in those of sorghum. This result suggests the occurrence of variants of the fungus with different levels of resistance to fungicides. Depending on the plant species from which seeds F. moniliforme was isolated, Kini et al. (2002) recently found specific genetic variants of the fungus in the country.

Both A. gourmaensis and E. alba extracts promoted plant growth. Seedling weight after germination was enhanced upon treatment of the seeds by either extract, regardless of the crop species. Similar increase in seedling fresh weight after treatment of rice seeds with fungal and bacterial bio-control agents have been reported (Mishra and Sinha, 2000). Extracts from A. gourmaensis were most efficient in promoting plant growth. The effect of both plant extracts on plant growth is possibly favoured by their upstream effects on controlling seed-borne fungi. However, other factors may be involved in the process, as in some cases plant growth was promoted while extracts were not efficient in controlling seed pathogens.

Field tests confirmed the beneficial effects of treating both sorghum and pearl millet seeds for the control of seed-borne fungi. On the one hand, emergence and vigour of seedlings were generally increased in both crops when the seeds were treated either with plant extracts or chemical products. On the other hand, treatments of the seeds did not affect the mortality of emerged seedlings significantly. Consequently, grain yields were increased due to treatments of the seeds. Treatments of sorghum or pearl millet seeds with plant extracts from E. alba led to the most increase in yield. Altogether, plant extracts showed an overall good performance in promoting seed health against infections by fungi and enhancing grain yields.

The fungicidal effects of E. alba extracts reported here are consistent with results obtained by Kaushal et al. (2002) and Venkatesan and Ravi (2004). These authors have found high inhibitory effects of extracts of E. alba against fungi such as Macrophomina phaseolina, Candida tropicalis, Rhodotorula glutini and Candida albicans. The antifungal activity of aqueous bark extracts from Acacia gourmaensis are reported for the first time. The inhibitory effects of both A. gourmaensis and E. alba on fungi can be attributed to active chemical components within the plants. However, more investigation is needed to identify the chemical nature of these components.

Overall, results obtained in this study underlined the possible used of aqueous extracts of E. alba and A. gourmaensis by smallholder farmers to control several seed-borne fungi. These include fungi infecting pearl millet seeds (Phoma sorghina, Curvularia lunata and Fusarium moniliforme) and those infecting sorghum seeds (P. sorghina and Colletotrichum graminicola). Extract from both E. alba and A. gourmaensis may be used as alternatives to commonly used chemicals for seed treatment against fungi. Both plant species are widely distributed in sub-Saharan Africa. However, further investigation is needed in order to determine optimal concentrations of the extracts. Investigation on the mammal toxicity of the extracts is also required. Testing the effect of extracts from organs other than bark, especially for A. gourmaensis may also be useful for better preservation of this species.

ACKNOWLEDGMENTS

The authors are grateful to the Danish Seed Health Centre for Developing Countries (DSHC) for providing laboratory equipment and funds for this work. The authors are also thankful to the International Foundation for Science for providing financial support and complementary equipment as research grant given to Mrs Zoma-Zida P. Elisabeth. They wish to acknowledge Drs. Jan Torp and Marianne V. Bengtsson (DSHC, Denmark), Dr Hamidou Traoré and Dr Oumar Traoré (INERA, Burkina Faso) for reviewing the manuscript.