Studies on the pathogenicity of Metarhizium anisopliae (Metch.) Sorok. to Schistocerca gregaria (Forsk.) with particular reference to the gut.

Abstract

A study has been made of the interaction between the entomo-pathogenic fungus Metarhizium anisopliae and the gut of the desert locust Schistocerca gregaria. This was undertaken in an attempt to define the reasons as to why the insect gut is generally a poor environment for germination and penetration by parasitic fungi. Conidia fed to 5th instar locusts appeared in the faeces soon after the completion of the test meal and numbers peaked at 13/4 h, slightly less than the food transit time through the gut of a continuously fed insect. By 41/2 h over 90% of conidia had been voided from the gut. Since the mean conidial germination time, in vitro, is 11 - 12 h, then most of the conidia had passed through the gut without germinating. However, approximately 1% of the ingested conidia were still resident in the gut after 24 h, which is theoretically long enough for germination and for penetration to have occurred. The viability of gut-passaged conidia was reduced compared to controls. This was not due to deleterious gut pH or osmolarity. Phytochemicals in the diet were not implicated and the gut was found to be predominantly aerobic. Conidia placed in locust gut fluid in vitro were inhibited from germinating, whether the fluid was used neat or diluted 1:3 with nutrient broth. However, no inhibition occurred when conidia were preincubated for 6 h in nutrients prior to exposure to gut fluid; thus conidial swelling was the sensitive phase. In contrast to gut fluid from conventional locusts (i.e. parasitised "stock") and parasite-free locusts, gut fluid from axenic insects was not inhibitory to conidial germination. Passage through the axenic gut also did not significantly affect conidial viability. There was no significant difference in 1,3-glucanase or "chitinase" (endochitinase) activity between conventional and axenic insects guts which otherwise might have accounted for the difference in the effects of axenic and conventional guts on conidia of M. anisopliae. Indeed previous work has suggested that axenic locusts are physiologically comparable to conventional locusts (Hunt, 1982). Therefore it was concluded that the detrimental effect of the locust gut on Metarhizium conidia was best accounted for on the basis of an antifungal toxin produced by the gut bacteria. Consistent with this hypothesis, incubation of Entero-bacter agglomerans (the dominant hindgut bacterium) with plant homogenate produced a filtrate which was inhibitory to Metarhizium conidial germination. The "antifungal toxin" was heat labile (5 mins at 100°C) and had a low molecular weight (<1200). Oral inoculation of starved or fed conventional insects with Metarhizium conidia failed to produce an infection via the gut wall. However infection of starved axenic insects was achieved by this route. Conidia germinated in the gut and produced blasto-spores. These lodged between rectal pads of the hindgut where they germinated and proceeded to invade the gut wall. A massive host melanic response in the gut epithelium failed to stop the fungus from spreading throughout the body and killing the insect. Presoaking prior to providing a suitable nutrient source halved the time to germination with respect to control conidia. The rate of swelling and germ tube formation was proportional to the length of the presoak time up to a maximum at 20 h. The physiological basis of the phenomenon was not discovered. However, measurements with the Coulter counter showed that no swelling took place during the presoak period and apparently no leached autogermination inhibitor was present in spent presoak fluid. The effect of presoaking on the rate of germination in vitro was also found to apply to conidia germinating in vivo on cuticle of larvae of the tobacco hornworm, Manduca sexta. Appressorial formation was also promoted. The pathogenicity of Metarhizium conidia to 2nd instar M. sexta was similarly enhanced by presoaking the conidia. A study on germination of Metarhizium conidia showed that germination was promoted most effectively by a nutrient containing both N and C (viz. peptone) and that the swelling phase was a metabolic event, not just a period of passive water uptake. A preliminary S.E.M. study revealed that preincubated conidia germinated faster than fresh conidia on locust cuticle. The relative merits of presoaking and preincubation, as methods for reducing germination time of conidia, are discussed particularly with regard to their practical use in the field situation.

Date of Award	1984
Original language	English
Awarding Institution	University of Bath