FEMS Microbiology Letters 227 (2003) 311^317

Transcription

1 FEMS Microbiology Letters 227 (2003) 311^317 New use of broomcorn millets for production of granular cultures of aphid-pathogenic fungus Pandora neoaphidis for high sporulation potential and infectivity to Myzus persicae Li Hua a, Ming-Guang Feng a;b; b a Institute of Microbiology, College of Life Science, Zhejiang University, Hangzhou , PR China Institute of Applied Entomology, College of Agricultural Science and Biotechnology, Zhejiang University, Hangzhou , PR China Received 22 June 2003; received in revised form 25 August 2003; accepted 11 September 2003 First published online 2 October 2003 Abstract Glutinous broomcorn millets from the crop Panicum miliaceum were first used as substrate to produce granular cultures of Pandora neoaphidis, an obligate fungal pathogen specific to aphids. Carrying a water content of 36.5% after being steamed in a regular autoclaving procedure, millet grains of each 15 g (dry weight) in a 100-ml flask were mixed with 3 ml modified Sabouraud dextrose broth containing half a mashed colony of P. neoaphidis grown on egg yolk milk agar and then incubated at 20 C and a light/dark cycle of 12 h/12 h for 21 days. Based on individually monitoring conidial production potential of 20 millet grains sampled from an arbitrarily taken flask at 3-day intervals, the millet cultures incubated for 6^15 days were capable of producing 16.8^23.4U10 4 conidia per millet grain with conidial ejection lasting for up to 6 days. The cultured millet grains individually produced significantly more conidia than apterous adults of Myzus persicae killed by P. neoaphidis (8.4U10 4 conidia per cadaver) and sporulated twice longer. The modeling of time^dose^mortality data from bioassays on M. persicae apterae exposed to conidial showers from the cultured millet grains and the mycelial mats produced in liquid culture resulted in similar estimates of LC 50 (millets: 21.4, 7.3, and 4.9 conidia mm 32 on days 5^7 after exposure; mycelial mats: 22.1, 10.6, and 7.7 conidia mm 32 ) although the LT 50 estimated at a given conidial concentration was slightly smaller for the millet cultures than for the mycelial mats. This indicates that the millet grains cultured with P. neoaphidis produced conidia as infective as or slightly more infective to M. persicae than those from the mycelial mats. Based on the sporulation potential, infectivity, and ease and cost of the millet cultures, the method developed in this study highly improved in vitro cultures of P. neoaphidis and may adapt to culturing other entomophthoralean fungi for microbial control of insect pests. ß 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Entomophthorales; Entomophthoracea; Broomcorn millet; Granular culture; Sporulation and infectivity; Microbial control; Pandora neoaphidis; Panicum miliaceum; Myzus persicae 1. Introduction Entomophthoralean fungi (Zygomycotina: Entomophthorales) are well-known insect pathogens [1^3]. This group includes numerous species that frequently cause epizootics in populations of insect pests such as aphids and leafhoppers and are considered as important biocontrol agents of insect pests [4^11]. In the fungal group, Pandora neoaphidis (Remaudie're * Corresponding author. Tel./Fax: +86 (571) address: mgfeng@zju.edu.cn (M.-G. Feng). and Hennebert) Humber ( = Erynia neoaphidis Remaudie're and Hennebert) is an aphid-speci c pathogen that induces epizootics in populations of various aphid species worldwide [4^10]. The high potential of this fungal species in natural control of aphids has drawn much research attention. P. neoaphidis usually grows slowly and unsteadily on arti cial substrate such as egg yolk milk agar medium though it can be cultured in vitro [12]. As is characteristic of most entomophthoralean species, primary conidia of P. neoaphidis are forcibly ejected from conidiophores that grow out from host cadavers in nature or mycelial mats cultured in vitro. In previous attempts to use P. neoaphidis and other entomophthoralean fungi for aphid control, emphasis has been placed upon introducing foreign mycosiskilled cadavers to local host populations for colonization / 03 / $22.00 ß 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi: /s (03)

2 312 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^317 [13] or releasing laboratory-prepared cadaver powder into host populations for an epizootic [14,15]. Despite more or less success, the propagation of P. neoaphidis in vivo is too laborious and expensive for practical use [16]. The earlier e ort has also been directed toward the spray of P. neoaphidis mycelia, cultured in liquid medium, for induction of aphid epizootics under eld or greenhouse conditions but failed [17^19] largely due to the fragility of mycelia in environment [16]. The technical di culties in the propagation of fungal inocula biologically competitive to those discharged from mycosis-killed cadavers have largely slowed down in the past decades the progress in research and utilization of P. neoaphidis as well as other entomophthoralean fungi. Recently, laboratory trials to immobilize liquid-cultured Pandora mycelia into materials such as alginate matrix and polyacrylamide-starch gel [20^22] have somewhat progressed, making it possible to prepare gelatinized pellets that may eject conidia as natural cadavers to infect healthy aphids. Presently, however, methods involved in preparation of gelatinized P. neoaphidis pellets are not practical for use, leaving problems in post-culture desiccation and preservation to be resolved. Moreover, mycelial pellets from shake asks have been found to sporulate less abundantly than aphid cadavers [23]. Is it possible to nd an easier way to make pellet-like or granular cultures that biologically function as disease-killed cadavers and sporulate better? In the present study, we developed a novel method for easing propagation of P. neoaphidis inocula using small grains as substrate of solid culture. Sporulation potential was compared between the solid cultures and aphid cadavers. The infectivity of the cultures to the green peach aphid Myzus persicae (Sulzer) was also evaluated in parallel to that of mycelial mats produced in liquid culture. 2. Materials and methods 2.1. Fungal isolate and cultures P. neoaphidis F98028 used in this study was isolated from a M. persicae cadaver collected on cabbage in Hangzhou, Zhejiang Province, PR China in October The isolate was maintained on slants of Sabouraud dextrose agar supplemented with 0.5% (v/v) sesame oil and 0.1% (w/v) sucrose fatty acid esters (an emulsi er commonly used in food industry and available in food additive stores) at 3 C in the dark and recovered twice a year [12]. To initiate the study, colonial pieces taken from one slant culture were recovered on 60-mm Petri dish plates of Sabouraud dextrose egg yolk milk agar at 20 C and light/dark cycle of 12 h/12 h. Incubated for 1 week, the fungal colonies on plates were ready for the following use Granular cultures Glutinous broomcorn millets, yielded by the crop Panicum miliaceum L. in northern China and purchased from a local grocery store, were used as solid substrate to culture P. neoaphidis. Millet grains of each 15 g (dry weight) in 100-ml asks were soaked in hot water (V80 C) for V30 min, followed by washing and leaching. Then, the grains were autoclaved for 15 min at 121 C and cooled to ambient temperature. Each ask of the autoclaved grains was inoculated with a homogenized mixture of 3 ml Sabouraud dextrose broth, modi ed by supplementing 0.5% (v/v) sesame oil and 0.1% (w/v) sucrose fatty acid esters [12], and half a plate colony. Plugged with vent stoppers, all asks were incubated for up to 21 days in an incubator at 20 C and light/dark cycle of 12 h/12 h Collection and counts of conidia To estimate conidial production potential of the millets cultured over time periods, 20 millet grains arbitrarily taken from a ask at 3-day intervals during the 21-day incubation were individually placed at the center of small plates (13 mm in diameter) containing 2% agar only. Each of the plates was then inverted onto a 10-mm-high concave bottom (13 mm in diameter) containing 200 Wl of 0.5% dodecyl sodium sulfate solution, which, functioning as a surfactant, may inactivate fungal conidia quickly but has little e ect on conidial morphology [24]. This closed device consisting of the upper plate and the lower concave bottom where a small volume of the surfactant solution was enough to ll its inner conic part was especially designed for convenient spore collection. The primary conidia ejected from each grain or aphid cadaver fell freely down to the surfactant solution in which they were well suspended and inactivated without morphological change and secondary germination. All sampled millet grains xed on the upper plates in the devices were then incubated at 20 C and a light/dark cycle of 12 h/12 h. The lower concave bottoms were replaced daily for examination until the sporulation in the devices was not detectable. From each of the concave bottoms daily moved from the devices, 15 1-Wl suspension samples were pipetted for conidial counts of each cultured grain using a standard hemocytometer under microscope. A sum of the daily counts gave an estimate of overall potential for conidial production of each grain Bioassays for infectivity to aphids About 100 millet grains cultured with P. neoaphidis for 15 days were uniformly distributed on a 90-mm Petri dish containing 2% agar only. Maintained for 24 h at 20 C and a light/dark cycle of 12 h/12 h, these grains turned to sporulating status and were ready for inoculating aphids.

3 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^ Fig. 1. P. neoaphidis culture grown on broomcorn millets steamed in a regular autoclaving procedure. A: Millet grains unshelled. B: Shelled millet grains available in market. C: Cultured millet grains heavily wrapped with P. neoaphidis mycelia. D: Outgrowths and primary conidia ejected from a P. neoaphidis-killed aphid on glass slide in moist Petri dish, showing the conidial halo around the cadaver. E: The dust halo around a cultured millet grain on glass slide. F: Details of the dust halo under microscope (not stained), showing primary conidia typical for P. neoaphidis. Bar scales: 1 mm for A^E and 20 Wm for F. Sporulating mycelial mats from liquid culture [12,25^27] were also bioassayed for comparison. The bioassay system in this study referred to previous reports [25^27]. For inoculation, colonies of M. persicae apterae ( 9 2-day-old after last ecdysis) on detached cabbage leaves were separately exposed to the shower of the conidia ejected from the sporulating millet grains (assay 1) or mycelial mats (assay 2) on agar plates. Exposed to the shower for di erent periods of time, conidial concentrations ( þ S.E.M.) from low to high were 1.7 ( þ 0.4), 5.8 ( þ 1.5) and 25.5 ( þ 5.9) conidia mm 32 in assay 1, and 3.9 ( þ 0.9), 8.8 ( þ 1.0), 25.3 ( þ 1.2) and 40.4 ( þ 2.8) conidia mm 32 in assay 2. Showered under each concentration were three to ve aphid colonies (replicates) on detached leaves, each including 33^50 apterae. Three or ve aphid colonies not receiving conidial shower were included as blank control in the two independent bioassays completed within 1 month. After exposure, all aphid colonies were individually maintained for 7 days on detached leaves in moist Petri dishes at 20 C and light/dark cycle of 12 h/12 h, and examined daily for counts of deaths with typical P. neoaphidis syndrome. Aphid cadavers, whenever found, were individually examined for veri cation of P. neoaphidis infection under microscope immediately or after overnight maintenance in moist Petri dishes if the syndrome was not clearly visible. To compare conidial production potential of a P. neoaphidis-killed cadaver with that of each millet grain cultured as above, 20 fresh aphid cadavers arbitrarily taken from the deaths in assay 1 were individually mounted within the spore collection devices. Daily counts of conidia discharged from each cadaver were made as above for the cultured millet grains Data analysis Variances among the daily and cumulative counts of primary conidia produced by the millet grains cultured for di erent time lengths were analyzed using analysis of variance (ANOVA) procedures. Following [25,26,28], each set of bioassay data was tted to a time^dose^mortality model, generating the estimates of parameters and associated variances for the e ects of time and dose and their Fig. 2. Comparison of conidial production potential of P. neoaphidis cultures grown on broomcorn millets for 3^21 days (number of conidia per millet grain) to that of M. persicae cadavers (CA) killed by the same fungal species (number of conidia per cadaver). The bars with different lowercase letters di ered signi cantly in height (P ) based on ANOVA. Error bars: S.E.M.

4 314 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^317 Fig. 3. The potential and timing pattern of conidial production from P. neoaphidis cultures grown on broomcorn millets (number of conidia per millet grain) and cadavers of M. persicae apterae killed by P. neoaphidis (number of conidia per cadaver). A^G: P. neoaphidis cultures grown on millet grains for 3^21 days (harvested at 3-day intervals). H: Fresh aphid cadavers. White bars: daily counts of primary conidia with S.E.M. Shading bars: cumulative counts of primary conidia with S.E.M. interaction. All analyses, modeling, and computations were conducted using DPS software [29]. 3. Results and discussion 3.1. Cultures grown on broomcorn millets Broomcorn millet grains are ovate in shape and yellowish in color with the small size of 2.5^3.2U2.0^2.6U1.4^ 2.0 mm, weighing 3^10 g per 1000 grains (Fig. 1A,B). Soaked and steamed by regularly autoclaving, the millet grains became swollen and loose, including water content of 36.5% (estimated after drying at 80 C for 48 h). During the period of incubation at the regime concerned, P. neoaphidis mycelial growth on the millets was visible as the millet grains turned whitish. The mycelial layer on the millet grains became dense as the incubation progressed, and heavily wrapped the grains in 1 week or so (Fig. 1C). The grains wrapped with dense mycelia contained a water content of 54.9% when ready, and sporulated very abundantly while maintained on non-nutritional substrate under moist conditions (Fig. 1E,F). The conidia ejected from the cultured millet grains formed a layer of dustlike conidial halo, which was typical for P. neoaphidis (Fig. 1D) and other entomophthoralean fungi as well. Obviously, P. neoaphidis was successfully grown on the millet grains despite its fastidious requirements for nutrients. Conidia were discharged from the grains just as from aphid cadavers killed by the fungal pathogen. Thus, each

5 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^ of the cultured millet grains biologically functioned as mimic to a mycosis-killed cadaver Potential and timing pattern of conidial production Conidial production of the cultured millet grains in the spore collection devices was monitored for 6 days. The sums of daily counts varied signi cantly among the agedi erent cultures (F = 14.89; df = 6, 114; P ), but di ered little among the 20 millet grains at a given culture age (F = 1.32; df = 19, 114; P = 0.19). The 15-day-old millet culture was capable of producing up to 23.4 þ 1.3U10 4 conidia per millet grain. This estimate was signi cantly greater than those for the millet grains at the rest culture ages except for the 12-day-old culture (19.4 þ 1.5U10 4 conidia per millet grain), as shown in Fig. 2. Too young or too old millet cultures displayed less potential for conidial production but on average never produced fewer conidia than aphid cadavers (8.4 þ 1.0U10 4 conidia per cadaver). The millet grains cultured for 6^15 days individually produced primary conidia 2.0^2.8 times more than cadavers of M. persicae apaterae infected by P. neoaphidis. The millet grains with di erent culture ages to some degree di ered in timing pattern of sporulation on nonnutritional agar plates (Fig. 3). The younger cultures tended to produce conidia earlier than the older ones. For instance, conidial counts obtained on the rst day took 50^60% in total for the 3- and 6-day-old cultures but only 25^35% for the v 9-day-old cultures. Moreover, sporulation potential released during the rst three days varied less among the cultures (taking 88^92% in total) except the oldest one (95%). The rest 8^12% of the total potential was released during the last three days. Regardless of the culture ages, there remained several thousands of conidia discharged from each of the millet grains on day 6 despite very low proportions (1.2^3.3%) in total. Afterwards, very few conidia were collected, not a ecting the sums of conidial counts. In contrast, fresh M. persicae cadavers released 99.8% of their sporulation potential within the rst three days, i.e., 78.6% on day 1, 19.9% on day 2, and 1.3% only on day 3, respectively. Thus, the millet grains cultured with P. neoaphidis tend to sporulate much longer than the aphid cadavers Infectivity to aphids The time^dose^mortality data from the two bioassays on M. persicae apterae exposed to the showers of P. neoaphidis conidia discharged from the sporulating millet grains or mycelial mats produced in liquid culture are shown in Fig. 4. Deaths attributed to P. neoaphids infection did not appear until the third day and mostly occurred on days 4^7 after exposure. The development of aphid mortality apparently depended on the concentrations of conidial showers. All deaths in aphid colonies Fig. 4. Trends in cumulative mortalities of M. persicae apterae after exposure to di erent concentrations of conidial showers (number of conidia mm 32 ) from P. neoaphidis cultures grown on broomcorn millets (upper panel) or in liquid medium (lower panel). Error bars: S.E.M. receiving conidial showers displayed typical P. neoaphidis syndrome under microscope. On day 7 after exposure, the cumulative mortalities attributed to P. neoaphidis were 16.7, 61.3 and 94.2% at 1.7, 5.8 and 25.5 conidia mm 32 in assay 1, and 38.5, 51.1, 74.8 and 91.8% at 3.9, 8.8, 25.3 and 40.4 conidia mm 32 in assay 2, respectively. In blank control, natural aphid mortality was as low as 5.7% in assay 1 and 4.1% in assay 2, but none of the deaths was attributed to any fungal infection in examinations. The data displayed in Fig. 4 t well to the time^dose^ mortality model [25,26,28], generating an estimated dose e ect parameter of 2.04 ( þ 0.03) for assay 1 and of 1.65 ( þ 0.01) for assay 2. The time e ect parameters ( þ S.E.M.) for days 3^7 after exposure were estimated as þ 0.42, þ 0.33, þ 0.24, þ 0.21 and þ 0.24 for assay 1, and þ 0.54, þ 0.19, þ 0.17, þ 0.17 and þ 0.19 for assay 2, respectively. All t-tests for estimates of the parameters were signi cant (P ). The Hosmer^Lemeshow test for heterogeneity of the goodness of t [28] was insignificant for the modeling of each data set (assay 1: df =9, C = 8.79, P = 0.46; assay 2: df =7, C = 10.46, P = 0.16), indicating an accepted homogeneity for the time^dose^ mortality modeling. Thus, the estimates of the parameters were used to compute virulence indices of LC 50 and LT 50 for both bioassays, as shown in Fig. 5. For assay 1, the LC 50 was estimated as 124.8, 21.4, 7.3, and 4.9 conidia mm 32 for days 4, 5, 6, and 7 after exposure to conidial shower, respectively. The corresponding estimates for assay 2 were 175.0,

6 316 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^317 Fig. 5. Trends in estimates of logarithm-scaled LC 50 (number of conidia mm 32 ; upper panel) and LT 50 (lower panel) based on the modeling of time^dose^mortality data from bioassays of P. neoaphidis on M. persicae apterae exposed to conidial showers from the sporulating millet grains (solid line) or from mycelial mats produced in liquid culture (dashed line). Error bars: S.E.M. 22.1, 10.6, and 7.78 conidia mm 32, respectively. Di erences in LC 50 between the two bioassays tended to be small after the fourth day though apparently spanned widely on day 3. On the other hand, the lowest conidial concentration to kill 50% of the aphids (LT 50 ) was 5.0 conidia mm 32 in assay 1 and 8.0 conidia mm 32 in assay 2, both taking 6.9 days. As the concentrations increased, the estimates of LT 50 for both bioassays tended to approach to each other, e.g., 5.3 and 5.5 days at 15 conidia mm 32 and both 4.5 days at 50 conidia mm 32. Obviously, both bioassays displayed a similar infectivity of P. neoaphidis to M. persicae though conidia from the cultured millet grains seemed to be slightly more infective than those from mycelial mats obtained in liquid culture Conclusions and implications The results presented above indicate that the broomcorn millets are excellent substrate to produce granular cultures of P. neoaphidis that are capable of discharging primary conidia to infect M. persicae as mycosis-killed cadavers. Since Pandora species in Entomophthorales are usually insect-speci c pathogens and, due to fastidious requirements for nutrients, often grow slowly even on highly nutritive media such as Sabouraud dextrose egg yolk milk agar, it is likely that the millet grains can also be applied to easing cultures of other fungi such as Erynia, Furia, and Zoophthora in Humber s classi cation system [1]. Advantages of growing Entomophthorales on the millet grains over the use of an arti cially de ned medium are enormous. The method is easy and inexpensive. All that is needed is to cook the widely available millet grains in a regular autoclaving procedure, introduce fungal inocula to the cooked grains and then let them grow for several days. Based on the potential and timing pattern of the cultured millet grains for conidial production on non-nutritional substrate, millet cultures of P. neoaphidis with greater sporulation potential resulted from incubation of 6^15 days at 20 C in this study. Furthermore, the cultured millet grains on average sporulated signi cantly more and much longer than aphid cadavers though the millet grains are slightly larger than aphid cadavers (usually V2 mm). This strongly suggests that the millet cultures be highly potential for use in study and application of entomophthoralean fungi for microbial control. Comparing virulent indices of the millet cultures with those of the mycelial mats from liquid culture (Figs. 4 and 5) further supports this consideration. Since the granular cultures directly grown on the millet grains can be produced more readily, but much less costly, than gelatinized pellets [20^22], we highly recommend the use of the millet grains for in vitro cultures of entomophthoralean fungi such as Pandora and Zoophthora species as potential biocontrol agents of insect pests. The broomcorn millets are commonly known as nutritionally rich small grains, including 9.6% proteins, 0.9% lipids, 76.3% carbohydrates and plenty of minerals and vitamins. As early as in late 1970s, egg yolk was found to greatly improve P. neoaphidis cultures in vitro although none of its components proved essential for the fungal growth [30^32]. The e ect of egg yolk to improve entomophthoralean growth may relate to its naturally appropriate composition of utilizable proteins and lipids. However, due to insolubility in liquid medium, the egg yolk is usually used to improve colonial growth only on agar media. In a recent study, replacing the egg yolk and milk with very low level of fully emulsi ed vegetable oil in Sabouraud dextrose egg yolk milk agar has also resulted in good cultures of P. neoaphidis and Zoophthora radicans [12]. In this study, the considerably fast growth of P. neoaphidis on the small millet grains implies that the composition of their nutrients may meet needs to culture entomophthoralean fungi as fastidious as P. neoaphidis. At this time, however, it is unknown what nutrients of the millet grains are really essential to the fungal growth. Further work is needed for exploring a way to process the cultured millet grains for preservation and application. Acknowledgements We wish to express our sincere thanks to Y. Liang, C.

7 L. Hua, M.-G. Feng / FEMS Microbiology Letters 227 (2003) 311^ Chen, S.H. Ying, and J.H. Xu (Institute of Microbiology, Zhejiang University) for technical assistance. This work was partially supported by the Natural Science Foundation of China ( ), the Special Fund for Graduate Study Programs in Chinese Universities ( ), and the Cheung Kong Scholars Programme, Ministry of Education, China. References [1] Humber, R.A. (1989) Synopsis of a revised classi cation for the Entomophthorales (Zygomycotina). Mycotaxon 34, 441^460. [2] Li, Z.Z. (2000) Flora Fungorum Sinicorum Vol. 13: Entomophthorales, 168 pp. Science Press, Beijing. [3] Feng, M.G. (2002) Microbial control of insect pests with entomopathogenic fungi in China: a decade s progress in research and utilization. In: Advances in Microbial Control of Insect Pests (Upadhyay, R.K., Ed.), pp. 213^234. Kluwer Academic/Plenum Publishers, New York. [4] Feng, M.G., Johnson, J.B. and Kish, L.P. (1990) Survey of entomopathogenic fungi naturally infecting cereal aphids (Homoptera: Aphididae) of irrigated grain crops in southwestern Idaho. Environ. Entomol. 19, 1534^1542. [5] Feng, M.G., Johnson, J.B. and Halbert, S.E. (1991) Natural control of cereal aphids (Homoptera: Aphididae) by entomopathogenic fungi (Zygomycetes: Entomophthorales) and parasitoids (Hymenoptera: Braconidae and Encyrtidae) on irrigated spring wheat in southwestern Idaho. Environ. Entomol. 20, 1699^1710. [6] Feng, M.G., Nowierski, R.M., Johnson, J.B. and Poprawski, T.J. (1992) Epizootics caused by entomophthoralean fungi (Zygomycetes: Entomophthorales) in populations of cereal aphids (Hom.: Aphididae) in irrigated small grains of southwestern Idaho, USA. J. Appl. Entomol. 113, 376^390. [7] Pickering, J., Dutcher, J.D. and Ekbom, B.S. (1989) An epizootic caused by Erynia neoaphidis and E. radicans (Zygomycetes, entomophthoraceae) on Acyrthosiphon pisum (Hom., Aphididae) on legumes under overhead irrigation. J. Appl. Entomol. 107, 331^333. [8] Wilding, N. and Perry, J.N. (1980) Studies on Entomophthora in populations of Aphis fabae on eld beans. Ann. Appl. Biol. 94, 367^378. [9] Wraight, G.S., Wraight, S.P., Carruthers, R.I., Magalhaes, B.P. and Roberts, D.W. (1991) Description of a Zoophthora radicans (Zygomycetes: Entomophthoraceae) epizootic in a population of Empoasca kraemeri (Homoptera: Cicadellidae) on beans in central Brazil. J. Invertebr. Pathol. 58, 311^326. [10] Steinkraus, D.C., Boys, G.O. and Rosenheim, J.A. (2002) Classical biological control of Aphis gossypii (Homoptera: Aphididae) with Neozygites fresenii (Entomophthoralis: Neozygitaceae) in California cotton. Biol. Control 25, 297^304. [11] Montserrat, M., Castao'e, C. and Santamaria, S. (1998) Neozygites parvispora (Zygomycotina: Entomophthorales) causing an epizootic in Frankliniella occidentalis (Thysanoptera: Thripidae) on cucumber in Spain. J. Invertebr. Pathol. 71, 165^168. [12] Feng, M.G. and Xu, Q. (2001) A simple method for routine maintenance and preservation of entomophthoraceous cultures. J. Invertebr. Pathol. 77, 141^143. [13] Milner, R.J., Soper, R.S. and Lutton, G.G. (1982) Field release of an Israeli strain of the fungus Zoophthora radicans for biological control of Therioaphis trifolii f. maculata. J. Aust. Entomol. Soc. 21, 113^118. [14] Wilding, N., Mardell, S.K. and Brobyn, P.J. (1986) Introducing Erynia neoaphidis into a eld population of Aphis fabae: form of the inoculum and e ect of irrigation. Ann. Appl. Biol. 108, 373^385. [15] Wilding, N., Mardell, S.K., Brobyn, P.J., Wratten, S.D. and Lomas, J. (1990) The e ect of introducing the aphid-pathogenic fungus Erynia neoaphidis into populations of cereal aphids. Ann. Appl. Biol. 117, 683^691. [16] Wilding, N. and Latteur, G. (1987) The Entomophthorales-problems relative to their mass production and their utilization. Med. Fac. Landbouwwet. Rijksuniv. Gent 52, 159^164. [17] Latteur, G. and Godefroid, J. (1983) Trial of eld treatment against cereal aphids with mycelium of Erynia neoaphidis (Entomophthorales) produced in vitro. In: Aphid Antagonists (Cavalloro, R., Ed.), pp. 2^10. A.A. Balkema, Rotterdam. [18] Latteur, G., Destain, J. and Godefroid, J. (1984) Research on the feasibility of producing in vitro an entomophthorale biopreparation based on conidia or mycelium to control cereal aphids. In: CEC Programme on Integrated and Biological Control, Final Report 1979/1983 (Cavalloro, R. and Piavaux, A., Eds.), pp. 407^416. EUR 8689 B [19] Latge, J.P., Silvie, P., Papierok, B., Remaudie're, G., Dedryver, C.A. and Rabasse, J.M. (1983) Advantages and disadvantages of Conidiobolus obscurus and Erynia neoaphidis in the biological control of aphids. In: Aphid Antagonists (Cavalloro, R., Ed.), pp. 20^32. A.A. Balkema, Rotterdam. [20] Shah, P.A., Aebi, M. and Tuor, U. (1998) Method to immobilize the aphid-pathogenic fungus Erynia neoaphidis in an alginate matrix for biocontrol. Appl. Environ. Microbiol. 64, 4260^4263. [21] Shah, P.A., Aebi, M. and Tuor, U. (1999) Production factors involved in the formulation of Erynia neoaphidis as alginate granules. Biocontr. Sci. Technol. 9, 19^28. [22] Liu, Z.Q., Feng, M.G. and Xu, J.H. (2001) Gelatinization of Pandora delphacis mycelia produced in liquid culture and aphidicidal activity. Mycosystema 20, 353^357. [23] Bonner, T.J., Pell, J.K. and Gray, S.N. (2003) A novel computerized image analysis method for the measurement of production of conidia from the aphid pathogenic fungus Erynia neoaphidis. FEMS Microbiol. Lett. 220, 75^80. [24] Ying, S.H. and Feng, M.G. (2001) Optional components for emulsi- able suspensions of Beauveria bassiana conidia. Acta Phytophyl. Sin. 28, 345^351. [25] Feng, M.G., Liu, C.L., Xu, J.H. and Xu, Q. (1998) Modeling and biological implication of time^dose^mortality data for the Entomophthoralean fungus, Zoophthora anhuiensis, on the green peach aphid Myzus persicae. J. Invertebr. Pathol. 72, 246^251. [26] Xu, J.H. and Feng, M.G. (2000) The time^dose^mortality modeling and virulence indices for two entomophthoralean species, Pandora delphacis and P. neoaphidis, against the green peach aphid, Myzus persicae. Biol. Contr. 17, 29^34. [27] Feng, M.G. and Johnson, J.B. (1991) Bioassay of four entomophthoralean fungi (Entomophthorales) against Diuraphis noxia and Metopolophium dirhodum (Homoptera: Aphididae). Environ. Entomol. 20, 338^345. [28] Feng, M.G. and Poprawski, T.J. (1999) Robustness for the processing of bioassay data of microbial control agents based on time^dose^ mortality model. Subtrop. Plant Sci. 51, 36^38. [29] Tang, Q.Y. and Feng, M.G. (2002) DPS data Processing System for Practical Statistics, 648 pp. Science Press, Beijing. [30] Latge, J.P. (1981) Comparaision des exigencies nutrionnelles des Entomophthorales. Ann. Microbiol. (Inst. Past.) 132B, 299^306. [31] Latge, J.P. and de Bie'vre, C. (1976) In uences des liquids et acides gras jaune d oeuf sur la croissance et la sporulation des Entomophthorales. Ann. Microbiol. (Inst. Past.) 127A, 261^274. [32] Latge, J.P., Remaudie're, G. and Papierok, B. (1978) Un nouveau milieu pour la croissance et la sporulation d Entomophthorales pathoge'nes d aphids. Ann. Microbiol. (Inst. Past.) 129B, 463^476.