Plant Pathology (1989) 38, 571-576

Infection of the pearl millet {Pennisetum americanum) inflorescence hy the downy mildew fungus (Sclerospora graminicola)
S. T. SEMISI and S. F. L. BALL*
Department of Agriculture, University of Reading, Earley Gate, Reading RG6 2.4T. i'K

Different developmental stages of the inflorescence of pearl millet {Pennisetum americanum} were inoculated with zoospores of the downy mildew fungus {Sclerospora graminicola). Individual florets within a panicle were infected, with resultant malformation of any floral organs that were not fully differentiated at the time of infection. 'Green-ear' symptoms resulting from hyperplasia and hypertrophy of the host tissues were accompanied by both sexual and asexual sporulation of the fungus on the malformed plant parts. No grain set occurred in affected florets, indicating that secondary inoculum was able to cause yield reductions even at late stages in the host development. Infcution of differentiated stigmas led to rapid dissolution and necrosis of tissue and prevented colonization by the pathogen. This failure suggests that seeds are unlikely to be infected internally. INTRODUCTION Downy mildew disease (DM) caused by Sclerospora graminicola can result in severe grain losses of pearl millet (Pennisetum americanum). Primary infection and recurrence of the disease is due to soil-borne oospores. Secondary infection and spread occurs in the field by water-borne and airborne sporangia and zoospores. Mycelium has been observed in seed in the scutellum and endosperm (Suryanarayana, 1962; Shetty et al.. 1980), although whether seed transmission occurs is still controversial (Williams. 1984). Diseased plants show symptoms only on the base of the first infected leaf, with subsequent leaves becoming progressively more chlorotic (Kenneth, 1981; Francis & Williams, 1983). The inflorescence may show partial or complete phyllody. Safeeulla (1976) reported teratological changes of the bristles, glumes, stamens and pistils. The evidence for seedling infection by both oospores in the soil and zoospores in airborne aerosols is circumstantial but well documented (King & Webster, 1970; Singh & Williams, 1980; Ball, 1983; Williams, 1984). although the actual mechanisms and pathway(s) of infection and Present address: Pathology Section, NIAB Seed Testing Station, Huntingdon Road, Cambridge CB3 0LE, UK. colonization have yet to be elucidated. It has been suggested by Singh & Williams (19S0) and Ball (1983) that the pathogen reaches meristems of young shoots and tillers, colonizes undifTerentiated tissues and thereafter causes disease on leaves and/or inflorescences Plants infected before panicle initiation produce no inflorescence. Later infection may affect the tillers only, with the main shoot escaping disease. Malformation and phyllody of the organs within the inflorescence is well-known ('green-ear' disease), but information on direct infection of the inflorescence is lacking. Generally it has been assumed that panicle malformation is the result of s\ stemic infection of seedlings and tillers. Some plants produce panicles that are only partially malformed with some normal grain set. In these instances we noticed that the malformations occurred only on the gynoecia that became grossly elongated green structures, although still clearly bifurcated at the style-tips. Sometimes these malformed stigmas occurred at the base of the panicle, sometimes towards the tips or randomly scattered over the panicle. This paper presents evidence of panicle infection by zoospores of S. graminicola in which plants infected during floral development became diseased. The possibility that /oospore infection may be able to reduce yields at even a comparatively late stage in panicle development has not been previously considcied in the literature. I he

572

S. T. Semisi and S. F. L. Ball

30C, respectively, with 14-h daylength. Supplementary feeding with macronutrients (NPK) was supplied at sowing time and every 10 days until completion of the experiments. Florat initiation When seedlings reached the 3-4 leaf stage they were thinned out, leaving two seedlings per pot. Floral initiation was induced by short days (Ong & Everard, 1979; Coaldrake & Pearson, 1986) by covering the plants with black polyethylene sheets between 20.30 and 08.30 hours for 14 consecutive days. The test plants were grown until the desired floral development stages were reached (Table 1). Pathogen maintenance and production An accession of 5. graminicola from Bengou, Niger was supplied as oospores in dried, fiinely ground, pearl millet material. Seeds of peari millet cultivar 7042 were soaked in oospore suspensions (4-5 x lO^ml) for 1 h. Five seeds were sown in each pot in which 1 ml of oospore suspension had been added to each sowing hole.

contribution of infection by zoospores to epidemic progress and yield losses could be significant because late infection gives no opportunity for yield compensation by surrounding plants (Williams, 1984). MATERIALS AND METHODS Hosts Growth and maintenance Seeds of pearl millet P. americanum cultivar 7042 (susceptible to downy mildew) were supplied by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Hyderabad, India. To minimize variation, the same seed batch was used throughout this investigation. Seeds were surface-sterilized for 10 min in 1 g/l mercuric chloride followed byfivewashes in sterile distilled water (5 min each). Excess moisture was drained off and the seeds were dried at 40 C for 48 h and stored in sterile petri dishes at 5 C until required. Only samples with germination of >80% were used. Seeds were sown into 15-cm diameter plastic pots containing peat-based compost and placed on a greenhouse bench. Night/day temperatures were maintained at an average of 24

Table 1. Effect of inoculation with Sclerospora graminicola zoospores at defined floral development stages of Pennisetum americanum cv. 7042 Number of days from sowing
17 22

Treatment
ti

Number of leaves on mainshoot

6-7 8-9

Inflorescence development stage at inoculation Panicle initiation Branch primordia Spikelet primordia Spikelet differentiation (empty glume) Spikelet differentiation (stamens) Spikelet differentiation (pistil) Spikelet differentiation (ovule/carpel) Boot (undifferentiated gynoecia; fused styllodia) Panicle emerged from flag leaf (style elongation) Full protogyny (stigmas fully extended from hermaphrodite dorets)

Disease symptoms Complete phyllody Complete phyllody Complete phyllody Complete phyllody Partial phyllody Partial phyllody Partial phyllody Partial phyllody Normal seed development Flortt abortion, poor seed set

t:
tj

U
ts
t6 t7

27 29
31 32 34 36

10-11 11-12 12 14 13-14 14-15 14 15 14-15 14-15

U
t|0

41 47

Floral infection by zoospores of pearl millet downy mildew

The pots were placed in polythene tunnels with thermostatically controlled gas heating, which maintained an average night air temperature of 18 + 2C, and with summer daylight, which usually exceeded 50000 lux. After approximately 21 days, diseased leaves could be seen on oosporeinoculated plants. Zoospore inoculum was prepared in humid chambers as follows. Petri dishes and their lids (9 cm diameter) were each lined with filter paper to which 7 and 2 ml sterile tap water, respectively, were added. Leaves bearing sporangia were collected from the tunnels on the evening prior to inoculation. These were surface-sterilized by wiping with paper towels soaked in 70% alcohol and allowed to dry, cut into pieces approximately 4 cm long and placed, dorsal sides up (four to five pieces), in each humid chamber. These were incubated at 20C for 5 h followed by automatic reduction of temperature to 4C for 8 h. The temperature was returned to 20C and after 1 h zoospores had been released from sporangia into the water. The zoospore suspensions were removed from the dishes by Pasteur pipette and placed in glass beakers. The zoospore concentrations were estimated using a Fuchs Rosenthal haemocytometer and adjusted when necessary to with sterile tap water. RESULTS Effect of the pathogen on inflorescence development

573

Treatments Floral infection of pearl millet was investigated by inoculating plants at the stages of floral development defined in Table 1, Samples of five plants for each treatment were dissected to confirm each floral development stage before inoculation. The inoculated plants were transferred to an isolated section of the glasshouse to avoid any possibility of either cross-infection or contamination. Individual plants for pre-panicle emergence treatments (ti-ts) were inoculated with a sterile hypodermic needle that injected about 15 ml of zoospore suspension into the stem in the vicinity of each developing panicle. Test plants for treatments t9-tio (panicle emerged from boot) were sprayed to run-off with zoospore suspensions from an atomizer, and the heads bagged to maintain high humidity. A few plants were left uncovered as controls to assess the effect of this procedure. The effect of inoculation on the inflorescence of pearl millet was monitored by visual assessment for any morphological abnormalities at panicle emergence.

Results are summarized in Table 1. Complete malformation of the inflorescence occurred when inoculation was carried out during panicle initiation, branch primordia initiation, spikelet primordia initiation and differentiation into the empty glume (t|-t4. Fig, la). The malformed tissues supported asexual sporulation under favourable environmental conditions and occasionally produced oospores. Infection during panicle initiation (ti) caused proliferation of vegetative tissues (phyllody) without further floral differentiation. Inoculation at branch primordia initiation, spikelet primordia initiation and differentiation into the empty glume (t;-t4) resulted in growth of abnormal appendages in spikelets within 5-7 days of inoculation. No seed developed on plants inoculated during these stages of floral development. Infection by zoospores was by direct penetration of host tissues followed by formation of vesicles that produced intracellular mycelium within 12 h. Intercellular mycelium with finger-like haustoda was observed within 48 h. with systemic colonization of host tissues both by intercellular and intracellular hyphae. Externally both the flag leaf and the adjacent leaf consistently showed basal chlorosis and asexual sporulation. Inoculation during spikelet differentiation into the stamen primordia. pistil primordia and gynoecial development stages (ts-tj) resulted in partial malformation of the panicle with some of the seed setting normally (Fig. lb). Unaffected areas of inoculated spikelets completed seed development either as distinct areas of the panicle or scattered among the diseased spikelets. No further differentiation of floral organs occurred on infected spikelets. which appeared malformed at emergence of the panicle. With inoculations at stage t; (spikelet differentiation into the ovules carpels), malformations were confined to these floral organs, the rest having already differentiated normally (Fig, Ic), Stigmas, although malformed, were still recognizably bifurcate at panicle emergence. The anthers that had differentiated very early in inflorescence ontogeny were found to be normal. During the boot stage (t) the pathogen caused malformation of the ovary and the fused styllodia. and reduced the numbers of stigmatic hairs Diseased spikelets were recognizable by larger.

574

S. T. Sentisi and S. F. L. Ball

H l J . I . I n H o r e s e e n e e s o l l ' < n i i i \ i i i i i > i i i i ) t i i h a i i i i n i i i i r e c l e i . t w i t h .S'l U r d y / ' o n i i ' r i t m i n u D l u ( a ) C o m p l e t e m a l f o r m a t i o n o f I h e i n l l o r e s c e n e e w i l h i m s e e d cle\(.'l(ipnienl ( l | l^l ( h i I'.irlial m a l l o r n i a l i o n a n d p h y l l o d y w i t h s o m e s e e d set after p a n i c l e e i i i e r ( ; e n e e ( K I J I c ) N i i r n i . i l . i n i l i e i s U I I I H W N ) e n i e r g i i n ! a l b a s e o f m a l f o r m e d s t i g m a s (t*). ( d ) Sterile HorttS p r c s e n i a l I h e l i p u l p a n i c l e ( l i p s i e i i l i l \ ' l ,is well a s iiiairiiiiiici.1 a n d n o r m a l florets o n p l a n t s i n o c u l a t e d a t t h e b o o t stage (IB)

Floral infection by zoospores of pearl millet downy mildew

more elongated ovaries and styllodia, and stigmas which always emerged 24-48 h before those in normal spikelets. No seed developed from diseased spikelets. In addition, inoculation at the late boot stage caused sterile florets to form at the tip of the panicle (Fig. Id). Normal seed development frequently occurred on the basal half of the panicles due to the fact that the inoculum had f"ailed to reach these areas. When the pathogen was specifically introduced to the basal part of the panicles and not to the tips, malformations occurred on the Horets in this area and not on the upper half of the panicles, which produced normal seeds. Occasionally, where the inoculum had run down to the base of the panicle during the boot stage, malformed and/or sterile florets occurred along the inoculum pathway. Microscopic observations showed that, 48 h after inoculation, mycelium was present in the ovary walls with hypertrophied cell development. Within 7-10 days after inoculation, the ovaries became completely deformed and supported asexual reproduction by the fungus. However, the vascular tissues in the pedicel remained intact and uncolonized. No malformation of the floral organs occurred on plants inoculated during panicle emergence from the flag leaf (t9). Developing stigmas enclosed within the glumes were well protected from the pathogen. Although the pathogen was observed to penetrate the glumes, no colonization of other host tissues occurred. These plants produced normal seed at maturity. Inoculation at full-protogyny (tio) did not cause malformation or phyllody. However, little or no seed-set occurred on these plants. Within 12-24 h after inoculation, brown lesions were observed on the tips of the stigmas of inoculated panicles. The whole stigmas became completely brown and withered within 48-72 h of inoculation. At this stage, uninfected spikelets in control inflorescences showed normal, turgid, white stigmas. Anthers emerged within the next 24-48 h and although pollen was available no seeds developed from infected spikelets. The ovaries disintegrated and eventually dehisced and became necrotic. Florets that escaped infection and colonization developed normal seeds. Throughout this investigation control plants for each floral development stage completed normal seed development, DISCUSSION Prior to this work there was no record of

575

zoospores of S. graminicola being able to infect the floral organs of P. americanum. It has been suggested that systemic infection of P. americanum seedlings by S. graminicota (from oospores or zoospores) follows invasion of the apical meristem and colonization of undifferentiated tissues with characteristic symptoms on the leaves and panicle phyllody (Williams. 1984), The work recorded here shows that infection and disease could be localized to individual florets, 5, graminicola zoospores can germinate and penetrate the inflorescence at any developmental stage before maturity. However, the extent of the resulting malformation of host tissues was dependent upon the developmental stage at infection, maturity of the host tissues and escape from infection clearly being two major factors in this context. Only when all the tissues were fully differentiated did the zoospores fail to colonize, as was shown during full protogyny when there was a hypersensitive host reaction to pathogen invasion. This host reaction is the first recorded instance of S. graminicola inoculum causing direct damage to fully differentiated tissue of pearl millet. Infection of individual spikelets on the pearl millet panicle always resulted in phyllody. malformation and/or sterility, Sclerospora graminicola colonized the tissue, which prevented further differentiation and caused malfonnation of differentiating organs or abortion of differentiated tissues. These results suggest that transmission of this disease from the interior of the seed is very unlikely. Host tissues were either colonized, in which case phyllody occurred with no seed set. or late infection of mature stigmas resulted in necrosis, destroying them as receptive organs for pollination. In the field, partially malformed panicles are a common sight, with abnormal stigmas protruding from diseased florets interspersed with spikelets which have normal seed. This has been found even in parts of the Southern Sahel in Africa, where conditions are usually hot and dry and not particularly conducive to spread of downy mildew. Clearly further field laboratory investigations will be necessary to resolve how such symptoms are caused. Lack of full understanding of the infection processes by this fungus is hampering the development of rational control strategies in spite of the great importance of this crop in the semi-arid regions The present work used only one susceptible cultivar but the same symptoms have been seen on other cultivars in the licld. Further studies should investigate if cultivar resistance to floral infection occurs and whether or not it is correlated with juvenile resistance

576

S. T. Semisi and S. F. L. Ball

nicola. CMI Descriptions of Pathogenic Fungi and Bacteria No. 770. Kenneth R.G. (1981). Downy mildews of graminaceous crops. In The Downy Mildews (Ed. by D. M. Spencer), pp. 367-394. Academic Press, London. King A.B. & Webster O.J. (1970) Downy mildew of sorghum in Nigeria. Indian Phytopathology 23, 342349. Ong C.K. & Everard E. (1979) Short day induction of flowering in pearl millet (Pennisetum typhoides) and its effect on plant morphology. Experimental Agriculture 15,401^10. Safeeulla K. M. (1976) Biology and Control of the Downy Mildews of Pearl Millet, Sorghum and Finger Millet. Wesley Press, Mysore, India. Shetty, H.S., Mathur S.B. & Neergaard P. (1980) Sclerospora graminicola in pearl millet and its transmission. Transactions of the British Mycologicat Society 74. 127-134. Singh S.D. & Wilhams R.J. (1980) The role of sporangia in the epidemiology of pearl millet downy mildew. Phytopathology 70, 1187-1190. Suryanarayana D. (1962) Occurrence of an unknown fungal mycelium inside the sound grains produced on partly formed green ears of bajra plants. Science and Culture 28, 536. Williams, R.J. (1984). Downy mildews of tropical cereals. Advances in Plant Pathology 2. 1-103.

ACKNOWLEDGEMENTS Seed and pathogen were supplied by staff and cooperators from ICRISAT Center, Hyderabad, India and the Sahelian Center, Niamey, Niger. This study was carried out under MAFF licence No. PHF 149 117 (87) (S2(66), issued under the Import and Export (Plant Health) (Great Britain) Order 1980 and the Plant Pests (Great Britain) Order 1980. The work was completed by the first author as part of his PhD programme funded by the GTZ (SGCPP), and Ministry of Agriculture, Western Samoa. Thanks are extended to Dr J. Barnett for help and advice with the electron microscopy and to B. Thurley for her technical assistance. REFERENCES
Ball S.L. (1983) Pathogenic variability of downy mildew {Sclerospora graminicola) on pearl millet. I. Host cultivar reactions to infection by different pathogen isolates. Annals of Applied Biology 102, 257-264. Coaldrake P.D. & Pearson C.J. (1986) Environmental influences on panicle differentiation and spikelet number of Pennisetum americanum. Journal of Experimental Botany 37, 865 875. Francis S.M. & Williams R.J. (1983) Scterosporagrami-