Mass P roduction, Yield, Quality, Formulation and Efficacy of Entomopathogenic Metarhizium anisopliae Conidia

Aims: The aim of this study was to develop an easy but robust method for mass production, to formulate and compare the efficacy of mass produced conidia of a local isolate Metarhizium anisopliae against aphids and oil - formulated conidia against whiteflies. Study Design : The randomized complete block design with replications. Place and Duration of Study: 2012 - 2014, Laboratory of Crop Protection Department, Faculty of Agricultural Engineering and Veterinary Medicine, Lebanese University. Methodol ogy: Isolate of fungal entomopathogen M. anisopliae (Metschn.) Sorokin (LIM1) was grown on cooked rice, wheat, vegetable peels and burgul in roasting bags to produce and harvest spore powder. The cultures were dried and total yield of harvested conidia was determined. After harvesting, spores were submitted to quality control to assess concentration, germination, purity, moisture content and pathogenicity against rose aphid, Macrosiphum rosae L., mealy plum aphid,


INTRODUCTION
Most of the vegetable crops in Lebanon are produced in greenhouses. Greenhouse production is more intensive than the open field production allowing rapid spread of problematic pests such as aphids and whiteflies. Their rapid reproductive rates and high resistance to insecticides [1] make them difficult to control. The potential for resistant populations of green aphid Myzus persicae (Sulzer) (Hemiptera: Aphididae) and whitefly Bemisia tabaci (Genaddius) (Hemiptera: Aleyrodidae) to develop as a result of intensive insecticides use has prompted scientists to conduct studies on integrated pest management (IPM) strategies in which biocontrol may play a significant role. The most promising alternatives to chemical use are the entomopathogenic fungi such as Beauveria and Metarhizium which kill a number of pest species. Production of adequate quantities of a good quality inoculum is an essential component of any biocontrol strategy. Recent advances in production, formulation and application of biological control agents in the fields and greenhouses have resulted in improvements of mycoinsecticide products such as Mycotal and Ago Biocontrol Verticillium (based on Verticilium lecanii), PreFeRal and Pac-Sin (based on Paecilomyces fumosoroseus), and BotanyGard, Ago Biocontrol Beauveria, Bea-Sin and Boveril PM (based on Beauveria bassiana ) [2]. These products have shown the capacity not only to suppress but also to provide a good control of whiteflies and aphids in both greenhouses and field crops [3][4][5][6][7][8]. However, there are numerous factors that continue to impede the development of commercial products. Strong dependence on environmental conditions, slow action, relatively high cost, negative interactions with commonly used fungicides, poor adulticidal activity, limited shelf life are those limitations that have to be overcome in order to progress the development of strategies for IPM.
Conidia are probably the most appropriate propagules for field use, as they are more infectious units and show greater stability under dry conditions and after application than hyphae or blastospores [9,10]. Herein we report on an investigation into mass production of M. anisopliae on solid substrates. The yield, quality, persistence and shelf-life of mass produced conidia are also discussed.

Mass production 2.1.1 Fungal isolates
Isolate of M. anisopliae (Metschn.) Sorokin (LIM1) originally isolated from soils in Lebanon [8] was tested in this study. Long term storage of a single spore culture was achieved by freezing conidia in 10% w/w Glycerol at -20°C.

Preparation of inoculums
When required conidia were re-hydrated by suspending in a small volume of sterile water, placed on SDA and incubated at 23±2ºC under darkness for 14 days. Following incubation conidia were harvested using a spatula and then suspended in 0.03% Tween 80 solution. In order to remove any hyphal fragments the resultant conidial suspension was filtered and used for further experimentation.

Preparation and inoculation of solid substrates
Whole grains of long rice (parboiled), wheat (durum) and fine burgul (a form of whole wheat that has been cleaned, steamed or parboiled, dried, and then ground into grains of several sizes) were tested as solid substrates. In order to mass produce conidia in the shortest period of time, the initial step of the common diphasic liquid-solid fermentation method was eliminated. ) were injected into each bag.
Upon needle withdrawal the puncture was immediately resealed. Inoculated bags were thoroughly massaged to mix inoculum with the substrate and then placed into plastic containers and incubated at 25°C until the substrates were fully colonized by fungal hyphae. At this point plugs of cotton wool were removed and the content of the bag was kneaded for substrate aeration and maximum sporulation. The experiment was set up as a completely randomized design with a factorial arrangement (4 x 4). The treatments comprised of 4 substrates and 4 different times of harvest, 1, 2, 3 and 4 weeks post inoculation and incubation. Each treatment combination was replicated five times and repeated twice.

Effect of Mass Production on Conidial Yields
The conidia were harvested at weekly intervals in order to estimate the optimal harvesting period. One g of substrate was drawn from each replicate, placed into 10 ml of 0.03% Tween 80 solution and shaken using a rotary shaker (380 rpm) for 20 min. Spore counts were made and yield estimated using Neubauer haemocytometer. At its maximal production, the bags were removed from the incubator, sides of the bags split open in the middle and kept at fluctuating room temperature of 21±4°C and average relative humidity of 50±5% to allow rapid dehydration of the substrates. After a week of drying the aerial conidia were dislodged from the substrata by gently mixing the grains and then placing them onto 150 µm aperture sieve nested over 106 µm aperture sieve over a collection tray. The set was covered with muslin and manually agitated for approximately 30 min. The resultant powders were collected from the trays and placed into sterile vials, weighed and then dried in silica gel desiccators at 5°C in vacuum for a week. These conidia were also used for quality control assessments, determination of C:N ratios, pathogenicity and shelf-life.

Effect of Mass Production on Conidial Quality
The quality of the mass produced conidia was assessed for its purity, moisture content, viability and growth.

Purity test and moisture content
Conidial suspension used for yield determination was also assed for its purity using Neubauer haemocytometer and light microscope with phase optics at 400x magnification. Contamination control (CC) and moisture content of produced powder were done according to procedures described in insect pathology manual [11].

Viability and growth tests
Spore germination for each treatment was performed according to Ibrahim et al. [12] before and after desiccation. Briefly, ten µl of conidial suspension in 0.03% Tween 80 (10 5 conidia ml -1 ) were dispensed onto the center of each of the 10 replicate Petri dishes containing 15 ml of SDA. All plates were sealed with parafilm and incubated in darkness at 25±1°C. After 24 h incubation the incidence of germinated conidia was recorded at 400 magnification using phasecontrast illumination microscope. Total of five hundred conidia (100 conidia per replicate) were examined. Conidia were considered to have germinated when the germ tube of any length was clearly visible. The remaining five replicates were examined every three days over the following 15 days when colony diameters were measured.

Effect of Mass Production on Conidial C:N Ratios
C:N ratios (carbon and nitrogen composition) from conidia produced on SDA, rice, wheat, vegetable peels and burgul were determined using modified methods of Springer-Klee for organic carbon [13] and Kjeldahl method for total nitrogen [14].

Determination of total nitrogen
Each sample of desiccated conidia (0.5 g) together with 3.35 g of catalyst (K 2 SO 4 /CuSO 4 .5H 2 O/Se at 100:10:1 ratio) and 10 ml of conc. H 2 SO 4 were digested for 30 min at 200°C and then for 1.5 h at 350°C in a fume cabinet. After digestion for each batch of samples at least one blank (no conidia) and one standard ((NH 4 ) 2 SO 4 ) were prepared. Recovery was determined between 95-105%. Distillation was carried out in the distillation unit.

Aphids
Mass produced conidia and freshly harvested conidia from SDA cultures were compared for their virulence against three species of aphids. Mixed instar nymphs of rose aphid Macrosiphum rosae L. (Hemiptera: Aphididae), melon aphid Aphis gossypii Glover (Hemiptera: Aphididae) and mealy plum aphid Hyalopterus pruni (Hemiptera: Aphididae) were collected from infested roses grown in the greenhouse, melon plants grown in the open fields and plum trees, respectively. Collected individuals were monitored for symptoms of natural fungal infection or parasitism and only healthy 3 rd instars nymphs were used in the pathogenicity assays. For each treatment, a leaf lamina of respective host plant was dipped into 5 ml of the prepared conidial suspension (10 7 conidia ml -1 ) for 10 sec and the excess removed by placing the leaf on sterile filter paper for 10 min. Each treated leaf then was placed on a moist 9 cm diam. Whatmans filter paper No. 2 in a 9 cm diam. plastic Petri dish. Thirty healthy aphids were placed on the treated leaf lamina and incubated at 25°±2°C with at 16:8 h (Light: Dark) photoperiod. Controls consisted of leaves immersed in 0.03% Tween 80 only. The treatments were replicated 5 times with ten aphids in each replicate. The number of dead aphids was recorded daily over a period of 7 days. In order to confirm that death was due to fungal infection, the cadavers were transferred to Petri dishes lined with moist tissue paper to encourage fungal growth and sporulation.

Effects of oil Formulations on M. anisopliae Conidial Viability and Shelf-life
Conidia of M. anisopliae isolate LIM1 obtained from burgul substrate were used in this experiment because they retained high viability after desiccation and were highly virulent against three different species of aphids.
Oil formulations were prepared by suspending 3 mg of desiccated conidia (4.5-5% moisture content) in 10 ml of each: fresh refined sunflower oil (Spinney's own brand, Spinney's supermarket, Beirut, Lebanon), extra virgin olive (home produced, Kfar Hammam, South Lebanon) and reutilized/recycled oil (mixture of vegetable oils previously used for deep frying potato, meat or fish, Zaatar W Zeit restaurant, City Centre, Hazmieh, Beirut, Lebanon) to give a final concentration of 10 9 conidia ml -1 . Filtered through sterilised muslin cloth and rested for 3 hr the stock oil suspensions were placed into 20 ml Universal glass bottles and thoroughly mixed for 1 min using vortex mixer (Lab Dancer S42, VWR (IKA), UK). To maintain moisture content between 4.5-5% five dry non-indicating silica gel beads were added to each treatment. All bottles containing oil suspensions and bottles containing 3 mg of pure dry conidia as a standard for general comparisons were wrapped in foil to prevent any exposure to light. Resultant formulations were stored at 4ºC, room (actual fluctuating) temperatures and 28ºC.
Viability tests were conducted by diluting stock oil formulations with respective oils and by adding 0.03% Tween 80 solution to dry conidia to desired 10 5 conidia ml -1 . Resultant dilutions were vortexed for 3 min to obtain homogeneous suspensions. Conidial germination for each treatment was assessed as described above [12] at 24 hr, 1, 4, 6, 9, 12 and 16 weeks.

Effect of Oil-formulated Conidia on Whiteflies under Field Conditions
Twenty seedlings of cucumber plants Cucumis sativus L. (Cucurbitaceae) variety "Hiba" and 20 seedlings of egg plants Solanum melongena L.
(Solonaceae) variety "Dream" were planted in 35 cm diameter pots. Potted plants were grown under filed conditions until formation of at least 3-5 true leaves and natural infestation by whiteflies Bemisia tabaci Gennadius (Homoptera: Aleyrodidae). Five potted plants were randomly assigned to each treatment and density of whiteflies was estimated one day before treatment by picking one middle leaf per plant. Eggs and nymphs of whiteflies were counted in the laboratory using a dissecting microscope. Leaf area was determined and densities of whiteflies were calculated as per cm 2 .
Treatments were consisting of control 1 (Tween 0.03% only), control 2 (Tween 0.03% + 60% sunflower oil), conidial suspension (Tween 0.03% + dry conidia that has been stored for 16 weeks at 4°C) and emulsified sunflower oil conidial formulation (Tween 0.03% + 60% sunflower oil formulated conidia that has been stored for 16 weeks at 4°C ). Each plant was sprayed with 15 ml of respective formulation using hand sprayer. Formulations containing fungal conidia were calibrated to give final concentration of 1 x 10 8 conidia ml -1 . Two weeks after treatment, one leaf from each replicate plant was sampled and density of eggs and nymphs was determined. To see if death has occurred and was due to fungal infection, the leaves carrying dead and alive individuals were transferred to Petri dishes lined with moist tissue paper to encourage fungal growth and sporulation. The number of surviving eggs and nymphs was analyzed using analysis of variance (TWO way ANOVA).

Statistical Analysis
All data were subjected to analysis of variance (TWO way ANOVA) using a statistical software package (SPSS Inc. 2013 v.22). Differences among treatments were compared using Tukey's mean separation test (P<0.05).

Effect of Mass Production on Conidial Yields
When isolate of Metarhizium anisopliae LIM1 was grown under standard conditions on different solid substrates and harvested at 1, 2, 4 and 6 weeks intervals, significant differences in conidial yield were found between the substrates. Highest total conidial yields (as determined by spore counts using Neubauer haemocytometer) were obtained after 4 weeks for rice, wheat, burgul and vegetable peel substrates (2.9 x 10 9 /g, 5.0 x 10 8 /g, 4.9 x 10 9 / g and 2.9 x 10 9 /g, respectively, Table 1). The spore concentration per gram of substrate was highly variable (P<0.001). This concentration was highest for burgul, followed by rice, vegetable peel and then wheat. These results were comparable with the spore powder weights where 16 times more powder was obtained from burgul than from wheat, 4.2 times more than from rice (Table 1). It was observed that after harvesting considerable amount of LIM1 spores still remained on the substrates. Following sieving, these substrates were washed out with 0.03% Tween 80 solution to extract the spores for counting and to calculate the total concentration for each treatment. The extraction of the remaining spores from the cultures showed that a highest number of spores (5.1 x 10 8 /g) remained on the rice (Table 1).

Purity test
Microscopic examination of spore powder extracted from each substrate revealed negligible 0.001% contamination with Aspergillum or Penicillium spp. only in wheat powder. Burgul culture was pure and free from any microbial or other contaminants. Rice spore powder contained insignificant amount of rice starch.

Viability and growth tests
The viability test of desiccated and nondesiccated conidia was conducted at 24 h only since germination counts at 48 hours were difficult to perform as they were overgrown with mycelia. Overall, the germination for all treatments before desiccation was over 85% and germination after desiccation was over 70%. The highest germination of 99.8% and 95.3% was recorded for conidia produced on burgul before and after desiccation, respectively. The lowest rate of germination was observed for conidia obtained from wheat (86.7% and 70.2%, respectively) ( Table 2). The speed of germination was directly reflected in speed of fungal growth, where fastest germinating conidia were also fastest in radial growth ( Table 2).

Effect of Mass Production on Conidial C:N Ratios
Mass produced conidia should always be tested for quality control. One of many attributes of such quality is endogenous C:N ratios. Dry conidia used in our study and grown on different substrates were tested for carbon and nitrogen content. The results obtained were expressed in percentages as a mean of three replicates and then C:N ratio was determined for each treatment. The results in Table 3 illustrate that conidia harvested from burgul contained the lowest amount of carbon (31.2%) and conidia from SDA showed the highest 47.7% quantity of carbon. Nitrogen content, on the other hand, did not significantly differ if grown on different substrates. C:N ratios of conidia harvested from burgul and vegetable peel were significantly lower (F 4,14 =33.186; P=0.001) than those obtained from SDA, rice or wheat (Table 3).

Effect of Mass Production on Conidial Virulence
Virulence of mass produced inoculum was tested against M. rosae and mealy plum aphid H. pruni and melon aphid A. gossypii (Fig. 2). Dry conidia of LIM1 isolate grown on burgul and rice were more effective against rose aphids killing 100% within 2 days post treatment. Conidia from the same substrates killed 83% and 63% of mealy plum aphids, respectively. The lowest mortality (16.7%) 2 days post treatment was observed for mealy plum aphids treated with inocula from wheat ( Fig. 1). However, 100% control of rose and mealy plum aphids were achieved within 7 days post inoculation for all treatments irrespective to culture substrate. Furthermore, more than 80% of dead aphids treated with conidia harvested from burgul, rice and vegetable peel have developed external conidiation on rose aphids' cadavers (Table 4). Rose aphids (Table 4) and mealy plum aphids (Fig. 1) treated with conidia produced on wheat, on the other hand, have shown only 65% and around 16.7% of external mycosis, respectively, suggesting of being less fit and aggressive than conidia from cultures mentioned ( Table 4 and Fig. 1).

Effects of Oil Formulations on M. anisopliae Conidial Viability and Shelf-life
Conidial viability test on SDA after 24h of incubation did not present significant differences between formulations resulting in germination values above 90% (data not presented) and indicating that none of oil formulations had any adverse effect on conidial germination.     Fig. 3A) and F 3,39 =1.4, P=0.261 (Fig. 3B), respectively).
Two weeks post-treatment densities of eggs and nymphs of whiteflies treated with Tween conidial suspension and oil formulation were slightly but not significantly lower than in both controls on both tested plants F 3,39 =1.503, P=0.232 (Fig. 3A) and F 3,39 =0.884, P=0.840 (Fig. 3B), respectively).

DISCUSSION
The data of this study show that burgul substrate was considerably better substrate for mass production of aerial conidia of M. anisopliae (LIM1) in terms of yield. Nutritional values of solid substrate would greatly influence the yield of conidia [15,16]. Compared to unenriched white rice, bulgul has more fiber and protein, a lower glycemic index, and higher levels of most vitamins and minerals [17]. In addition, utilisation of solid substrates by microorganisms is affected by many other factors such as particle size, shape, surface-to-volume ratio, crystallinity, and porosity of the substrate, all factors that can influence the accessible surface area to both organism and enzymes [18]. Also, production of conidia of M. anisopliae on coarse grain is significantly greater than that on whole grain [19]. Yypsilos [20] reported that burgul wheat supported a higher yield of better quality conidia compared to millet grains because burgul provided a more accessible substrate. This difference was marked in the case of millet where the hard shell of the grains could have considerably prevented the fungus from easy access when small volumes of water were used. This case could be applied to our study where rice and whole wheat grains were used.

Mortality (%) Treatments
In relation to what has been discussed, present study also suggests that the nutritional content of burgul wheat is more favourable for conidial production of higher quality compared to other substrates and that the physical properties of burgul wheat grains allowed the fungus to access more nutrients. In addition, it appears that interaction between culture substrate and moisture availability would also affect the conidial production of M. anisopliae [15]. When compared corn, millet and wheat grain with different water volumes it was found that highest conidial production was achieved on wheat at 1:1.5 (substrate : water) ratio [15]. In our study the ratio 1:0.8 was tested.  Although the total spore production of LIM1 was higher on burgul than on the other substrates, it is still below 1.0 x 10 10 spores/g which is the target level for production of reliable mycoinsecticidal products [21]. It was also suggested that when entompathogenic fungal spores are considered for industrial production it would be necessary to adapt or change production methods to increase spore production and harvest with greater efficiency [22,23]. In the case of LIM1 products based on burgul, it will be necessary to improve the extraction method to remove the spores that are retained on the burgul particles.
The speed of germination could be affected by conidial quality since the endogenous reserves of such inoculum are initially the sole source of nutrients available to conidia during the process of germination. There were significant compositional differences reported in conidia of B. bassiana, M. anisopliae and P. fumosoroseus [24,25] and in blastospores of B. bassiana [26] produced in different nutritional environments. Conidia of three M. anisopliae isolates [27] accumulated significantly higher levels of polyols and total proteins when grown on insect cuticles or nitrogen-limited medium (MM) than conidia cultured on YEA or SDAM media and contained higher amounts of total endogenous nitrogen. In the study reported here it was evident that spores containing lower amounts of C but higher quantities of N had also a greater germination rate. Moreover, studies with pathogenic fungi have suggested that carbon and nitrogen concentrations of nutritional conditions are responsible for the pathogenicity of resultant inocula [24,28,29]. Shah et al. [30] have recommended endogenous C:N ratio (<5.2:1) of conidia in combination of other parameter as a virulence indicator in quality control.
In this study inoculum produced on different substrates has shown different level of virulence against different species of aphids suggesting that any isolate considered for mass production should be tested against as many pest insects as possible. Also, the virulence of M. anisopliae inoculum was reported to be influenced by the culture medium [12,30]. For example, inoculum of B. bassiana from insect cadavers was shown to be less virulent than that produced on rice or other synthetic substrate [31]. Isolate LIM1 in our study was noticeably more aggressive towards all target insects if grown on burgul or rice. These conidia were also identified with lower endogenous C:N ratios.
Spores of M. anisopliae (LIM1) produced on burgul and formulated in different vegetable oils when subjected to short term storage have shown gradual drop in viability in temperature and oil type dependent manner. Storage at high temperatures resulted in a faster rate of decline than at cooler temperatures. Conidia of B. bassiana was also shown to survive better if maintained at 25°C than 30°C retaining viability above 24% after three weeks of storage [32]. Moore et al. [33] suggested that if more than three weeks of storage were required for minimal loss of viability then cooled storage would be necessary. High germination of our formulated spores after 24 h of incubation indicated that none of vegetable oils used had any initial adverse effect on conidial germination. Survival of spores over period of 16 weeks at room and 28°C, however, was greatly affected by the oil used. Rapid decline in viability was observed for olive oil followed by recycled vegetable oil. Olive oil contains more monounsaturated oleic acid and less polyunsaturated linoleic and linolenic acids than other vegetable oils and thus renders olive oil more resistant to oxidation. However, it is generally accepted that a warm region olive oil will be richer in polyunsaturated fatty acids than a cool region oil. Vegetable oils rich in polyunsaturated fatty acid are more prone to oxidation compared to the oils which are rich in monounsaturated fatty acids [34]. Since olive oil used in this experiment was extra virgin oil from South Lebanon, it is possible that this oil was richer in polyunsaturated than in monounsaturated fatty acids which made it more unstable and easily broken down by heat. When these essential fatty acids are destroyed and certain fat soluble vitamins (antioxidants) disappear the oil ages (oxidizes) very quickly. Rapid disintegration of olive oil could have been the reason for poor spore survival. Recycled vegetable oil, on the other hand, had been previously subjected to extremely high heat as it was repeatedly used for deep frying fish, meats and potatoes. When heated repeatedly, changes in physical appearances and chemical content of the oils will occur [35]. Heating causes the oil to undergo a series of chemical reactions like oxidation, hydrolysis and polymerization [36]. Changed chemical composition of recycled oil combined with effects of high storage temperatures could have been a contributing factor to accelerated spore deterioration. In contrast, fresh refined sunflower oil used in our experiment, presumably have undergone solvent extraction, de-gumming, neutralization and bleaching processes that made it more stable and heat resistant and thus considered more suitable for long term storage at room temperatures. Although good survival rate of dry LIM1 spore powder was achieved at room temperature over period of 16 weeks, there is scope for further improvements. Data from other studies identified spores of B. bassiana stored under ideal conditions of low moisture content and low temperature had survived for over two years [37]. Also, some commercial products based on B. bassiana isolates that currently used in agriculture [38] have shelf lives of 12 months or more at room temperature.
Dry and sunflower oil formulated conidia that have been stored for 16 weeks at 4°C were tested against eggs and nymphs of B. tabaci on cucumber and egg plants. Both, Tween and oil formulated conidial suspensions were to some extent but not significantly effective against white flies in comparison to controls on both tested plants. Ineffective results of our test under open field could be attributed to either very low relative humidity or insufficient number of spray applications. With respect to whiteflies in particular, effectiveness of many mycoinsecticides under field conditions is not as good as in protected agriculture [2]. This is attributed to unfavorable environmental conditions, economic constrains and high susceptibility of crops to mass migration of whiteflies from surrounding vegetation or harvested fields [2]. However, when incubated at optimal conditions, half of those eggs and nymphs that have died showed fungal growth and sporulation suggesting of pathogenesis taking place. Efficacy of this entomopathogen under field conditions, however, could be improved by increasing dosage, rate of applications or even changing formulants. For example, M. anisopliae formulated in Shellsol T paraffin oil was more effective against Phaedon cocleariae, mustard beetle than a mixture of Shellsol T and sunflower [39].

CONCLUSION AND RECOMMENDA-TIONS
Our studies have demonstrated that LIM1 isolate could be mass produced in large quantities on burgul substrate. The quality of produced inocula as measured by endogenous C:N ratios would be largely affected by substrate used for mass production. The quality would also influence the efficacy against target insects. Mass produced conidia of LIM1 could also be stored as dry or formulated in sunflower oil at room temperature up to 16 weeks. Four-month-old conidia could still be successful in controlling whiteflies. The authors strongly recommend to use the endogenous C:N ratio as quality control indicator for high quality inocula. Use of vegetable peels as potential substrate for mass production combined with a recyclable oil as formulant could be a low cost and environmentally sustainable technology for mass production of entomogenous fungi.