Journal of Invertebrate Pathology 120 (2014) 67–73

Contents lists available at ScienceDirect

Journal of Invertebrate Pathology

journal homepage: www.elsevier .com/ locate/ j ip

Heat-induced post-stress growth delay: A biological trait of manyMetarhizium isolates reducing biocontrol efficacy?

http://dx.doi.org/10.1016/j.jip.2014.05.0080022-2011/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author. Current address: Department of Plant and Environmen-tal Sciences, University of Copenhagen, Thorvaldsensvej 40, stair 2, 3rd floor, B321,Frederiksberg 1871, Denmark.

E-mail address: chadaltonkeyser@gmail.com (C.A. Keyser).1 Current address: Instituto de Patologia Tropical e Saúde Pública, Universidade

Federal de Goiás, Goiânia, GO 74605-050, Brazil.2 Current address: Instituto de Pesquisa e Desenvolvimento, Universidade do Vale

do Paraíba, São José dos Campos, SP 12244-000, Brazil.

Chad A. Keyser ⇑, Éverton K.K. Fernandes 1, Drauzio E.N. Rangel 2, Donald W. RobertsDepartment of Biology, Utah State University, Logan, UT 84322-5305, USA

a r t i c l e i n f o

Article history:Received 17 March 2014Accepted 27 May 2014Available online 6 June 2014

Keywords:Heat toleranceEntomopathogenic fungiMetarhiziumVegetative growthBioassay

a b s t r a c t

The habitats of many pest insects have fluctuating climatic conditions. To function effectively, the patho-gens of these pests must be capable of infecting and developing disease at a wide range of temperatures.The current study examines ten Metarhizium spp. isolates as to their ability to recover normal metabolicactivity after exposure to high temperature for several hours daily; and whether such recovery, with atleast some isolates, requires a temporary repair (‘‘retooling’’) period. Fungal colonies were exposed to40 �C for 4 h or 8 h followed by 20 h or 16 h at 28 �C, respectively, for three consecutive days. Growthrates during treatments were compared to control plates (constant 28 �C) and to plates with growth stop-page by cold treatment (4 h or 8 h at 5 �C per day). All ten isolates survived 3 days of cycled heat treat-ment and resumed normal growth afterward; some isolates however, were considerably more negativelyaffected by heat-cycling than others. In fact, some isolates underwent greatly reduced growth not onlyduring 8 h heating, but also some hours after cessation of heat treatment. This phenomenon is labeledin the current study as ‘‘post-stress growth delay’’ (PSGD). In contrast, all isolates stopped growing during8 h cold treatments, but immediately recommenced growing on return to 28 �C. The delay in recommenc-ing growth of some isolates after heat treatment amplifies the effect of this stress. In addition to the stud-ies on the effects of heat cycling on fungal cultures, the effects of imposing such temperature cycling onfungal infection of insects was documented in the laboratory. Three Metarhizium isolates were bioassayedusing Galleria mellonella larvae. Treated insects were placed at daily temperature regimes matching thoseused for the in vitro fungus rate-of-growth study, and insect mortality recorded daily. For all three isolatesthe levels of insect mortality at the highest-heat dose (40 �C at 8 h daily) significantly reduced infection.Fluctuating temperatures are likely to be a factor in most pest–insect habitats; therefore, the presenceand level of PSGD of each isolate should be a primary consideration in selecting field-appropriate fungalisolates.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Problematic outbreaks of insect pests often occur in warm cli-mates, such as those commonly found in Africa, Australia, China,Mexico, South America and the United States (Faria and Wraight,2007). In these regions, pest outbreaks frequently occur duringperiods when temperatures are fluctuating and can range from10 to over 40 �C within a single day. High (40 �C) daytime

temperatures are known to be a major hindrance to growth ofsome strains of fungal biological control agents like Metarhizium(Metschnikoff) Sorokin (Clavicipitaceae: Hypocreales) (McClatchieet al., 1994; Moore et al., 1996). It is generally assumed that a fun-gal isolate is unlikely to infect insects at temperatures above itsupper growth threshold; but as long as the fungus is not killed,infection and disease development are expected to recommenceduring cooler periods of the day (Rangel et al., 2010). Conse-quently, temperature tolerance studies usually focus on conidialsurvival, including thermal death points or relative germinationrates after thermal exposure (Arthurs and Thomas, 2001; Rangelet al., 2005; Zimmermann, 1982). This information is useful indetermining which strains have a predisposition to survive poten-tially high field-temperatures; but does not highlight the potentialof non-lethal negative effects on the efficacy of a fungus resultingfrom the heat exposure.

68 C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73

Foster et al. (2010, 2011) noted that the M. brunneum isolateF52 (Novozymes, Salem, NC, USA) showed significant promise asa biocontrol agent for the Mormon cricket in laboratory tests; how-ever, when applied as a spray or bait treatment in field-cage trialsthe isolate failed to cause significant mortality. They explained thepoor performance in the field was primarily due to fluctuatingtemperatures outside the optimal growth range for that isolate.They estimated that the optimal growth rates (>50% maximum)for M. brunneum ranged between 18 and 30 �C. Using surrogates(thermistors in liquid in 1.5-ml tubes) for Mormon crickets, theyobserved that this optimal range occurred 0–11.8 h per day duringtheir field trial. They concluded that the temperature was the pri-mary cause of the reduced mortality; they speculated, however,that with sufficient time mortality eventually would occur becausethe fungi would have opportunity to grow during the optimaltemperature growth periods.

Several mechanisms for surviving heat stress have beenobserved in fungi (Butler and Day, 1998; Guerzoni et al., 2001;Rothschild and Mancinelli, 2001); however, these functions oftendivert metabolic resources. In fact, a recent transcriptonic analysisof heat-stressed (38 �C) M. robertsii (ARSEF 23) conidia reported2722 up-regulated and 788 down-regulated genes (Wang et al.,2014). Reduced or delayed fungal growth (e.g., conidial germina-tion and mycelial extension) is likely; and this in turn, may therebydelay initiation and/or expansion of insect infection. After a cold orhot period that is sufficiently extreme to arrest fungal growth, anisolate’s ability to immediately begin growing or recommencinggrowth on return to growth-permissive temperatures is cruciallyimportant for disease development in that environment. Accord-ingly, an important step in selecting fungal strains with high prob-ability of success as biological control agents is to evaluate eachstrain’s growth responses to temperatures normally present inthe target pest’s environment, including how quickly each straincan recommence growing after heat shock.

If significant growth-rate reduction due to heat-induced post-stress trauma occurs in a fungal isolate, and if the high-heat expo-sure re-occurs on a daily basis, this may inhibit the infection ofinsects by the entomopathogen. This is because, with many fungalstrains, negative effects from heating may continue for severalhours after treatment, which may render these strains ineffectiveas biological control agents in that area at that time. The researchreported here documents the effects of high-temperature stress onboth: (1) mycelial growth of ten Metarhizium spp. isolates in vitro(agar medium) and (2) virulence of three of these isolates towardwax moth larvae [Galleria mellonella (Lepidoptera, Pyralidae)].

2. Materials and methods

2.1. Fungal isolates

Ten isolates of Metarhizium spp. were examined (Table 1): two(ARSEF 324 and ARSEF 1095) were received from the USDA-ARSCollection of Entomopathogenic Fungal Cultures (ARSEF) (US Plant,Soil and Nutrition Laboratory, Ithaca, NY, USA). Bischoff et al.(2009) identified ARSEF 324 as Metarhizium acridum (Ma) andARSEF 1095 as M. brunneum (Mb). The remaining eight isolateswere obtained from western USA soils. Soil was collected and pla-ted on selective media as described by Fernandes et al. (2010b). Inshort, field collected soil was suspended in distilled water; vor-texed and then pipetted onto a selective medium; growth wasobserved weekly; and colonies identified as Metarhizium were cul-tured on potato dextrose agar (Difco Laboratories, Sparks, MD,USA) supplemented with 1 g l�1 yeast extract (Technical, Difco)(PDAY) adjusted to pH 6.9. These isolates are preserved in the Don-ald W. Roberts culture collection (DWR) (Utah State University,

Logan, Utah), and the USDA-ARS collection in Ithaca, NY (Table 1).Preliminary identification based primarily on the DNA a elongationfactor as described by Bischoff et al. (2009) indicated that DWR200, DWR 203, DWR 312 and DWR 313 are M. guizhouense (Mg);DWR 261 is M. brunneum (Mb); and DWR 338, DWR 346 andDWR 356 are M. robertsii (Mr) [Fernandes et al. (unpublished)].Stock cultures were maintained at 4 �C in test-tube slants ofPDAY. Fresh conidia of each isolate were produced on 23 ml PDAYmedium in 95 � 15 mm polystyrene Petri dishes (Fisherbrand�,Pittsburg, PA, USA). Cultures were incubated for 14 days at28 �C in the dark prior to harvest of conidia for experimentinitiation.

2.2. Effects of temperature on vegetative growth rates

2.2.1. Culture at constant temperaturesThree trials (repetitions) were conducted to estimate the

growth rates at eight temperatures of the ten Metarhizium isolates.Conidia were harvested with a microbiological loop and suspendedin 0.01% Tween 80. Two microliters of the suspension was placedin the center of a Petri dish (23 ml PDAY 95 � 15 mm polystyrene).All of the plates were held for 24 h at 28 �C to allow conidial germi-nation, and then the plates were moved to their respective testtemperatures, viz., 20 ± 1 �C, 24 ± 1 �C, 28 ± 1 �C, 32 ± 1 �C,36 ± 1 �C, or 40 ± 1 �C. Colony growth was determined by measur-ing (mm) across two intersecting diameters for each colony every24 h for 10 days using a carbon-fiber composites digital caliperwith 0.1 mm resolution, accuracy ±0.2 mm (Fisher Scientific,Hampton, NH, USA). The initial measurement of each fungal colonywas made at 24 h (i.e., before the first temperature treatment) andthis amount was subtracted from each subsequent measurementof that colony.

2.2.2. Culture at fluctuating temperaturesThree trials (repetitions) were set up identically with the above

constant temperature tests, except that the fungal colonies wereallowed to develop for 72 h at constant 28 �C, and then movedfor 3 days to one of five daily temperature treatments; namely(1) control (plates held at constant 28 ± 1 �C); (2) cold treatmentfor 4 h (plates placed at 5 ± 1 �C for 4 h, then moved to 28 ± 1 �Cfor 20 h); (3) cold treatment for 8 h (plates placed at 5 ± 1 �C for8 h, then moved to 28 ± 1 �C for 16 h); (4) heat treatment for 4 h(plates placed at 40 ± 1 �C for 4 h, then moved to 28 ± 1 �C for20 h); and (5) heat treatment for 8 h (plates placed at 40 ± 1 �Cfor 8 h, then moved to 28 ± 1 �C for 16 h). After the three 24-hcycles, all plates were returned to constant 28 ± 1 �C and held for4 days. Colony growth was determined by measuring (mm) acrosstwo diameters for each colony every 24 h using a carbon-fiber-composites digital caliper as previously mentioned. The initialmeasurement of colony size on each plate was made at the endof the first 72 h (before the first temperature treatment), and thisamount was subtracted from each subsequent measurement ofthat colony.

2.3. Bioassay

To estimate the effect of fluctuating temperatures on infectionof insects, a virulence assay was conducted while maintainingthe 4 or 8 h per day of heat or cold treatment described abovefor the length of the experiment (15 days). Three fungal isolateswere selected: DWR 203 (Mg), DWR 312 (Mg), and DWR 346(Mr). Conidia used for assays were harvested from 14-day-old cul-tures grown on PDAY plates (polystyrene, Petri dishes,95 � 15 mm, Fisherbrand�) incubated in the dark at 28 ± 1 �C.Harvested conidia were suspended in 40 ml of 0.01% Tween 80solution in 50-ml tubes (Modified polystyrene, Corning Inc.,

Table 1Metarhizium spp. isolates with their initial source (host insect) and geographic origin (with GPS coordinates).

Isolate Host/substrate Origin Coordinates

M. acridumARSEF 324 Orthoptera Australia �S 19�42.219 �E 145�46.264

M. brunneumARSEF 1095 Lepidoptera Austria �N 47�30.584 �E 14�33.02DWR 261 (ARSEF 8365) Soil Talkeetna, Alaska, USA N 62� 08.906 W 150� 02.303

M. guizhouenseDWR 200 (ARSEF 7847) Soil Snowflake, Arizona, USA N 34� 42.924 W 110� 01.860DWR 203 (ARSEF 8364) Soil Winslow, Arizona, USA N 35� 03.225 W 110� 35.388DWR 312 (ARSEF 8366) Soil Logan, Utah, USA N 41�46.465 W 111� 46.248DWR 313 (ARSEF 8357) Soil Sedona, Arizona, USA N 34�57.271 W 111� 45.148

M. robertsiiDWR 338 (ARSEF 8358) Soil Jacob Lake, Arizona, USA N 36� 43.555 W 112� 07.522DWR 346 (ARSEF 8367) Soil Snowflake, Arizona, USA N 34� 41.222 W 110� 06.199DWR 356 (ARSEF 9617) Soil Flaming Gorge, Utah, USA N 40� 51.225 W 109� 33.964

C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73 69

Corning, NY, USA). Each suspension was vigorously vortexed andconidial concentrations were estimated by hemocytometer counts.A conidial suspension with a concentration of 1 � 107 conidia ml�1

was prepared for each isolate, and used immediately.Fresh commercially produced last-instar G. mellonella larvae

(New York Worms, Long Island, NY) were used as the host insect.Ten larvae were placed in a 60 � 15 mm polystyrene Petri disheslined with a 5.5-cm P4 filter paper (Fisherbrand, porosity: medium– fine, flow rate: slow). The filter paper was moistened with 500 llof sterile distilled H2O. The dorsal region of each larva was inocu-lated topically with 10 ll of either a fungal suspension (test) orTween solution (non-fungus control). For each temperature/isolatecombination, thirty insects separated into three Petri dishes (10 ineach dish) were used. The daily temperature treatments were: (1)constant temperature (28 ± 1 �C); (2) cold treatment for 4 h(insects exposed to 5 ± 1 �C for 4 h, then moved to 28 ± 1 �C for20 h); (3) cold treatment for 8 h (insects exposed to 5 ± 1 �C for8 h, then moved to 28 ± 1 �C for 16 h); (4) heat treatment for 4 h(insects exposed to 40 ± 1 �C for 4 h, then moved to 28 ± 1 �C for20 h; and (5) heat treatment for 8 h (insects exposed to 40 ± 1 �Cfor 8 h, then moved to 28 ± 1 �C for 16 h). Treatments wererepeated for fifteen 24-h cycles. Insect mortality was assesseddaily.

2.4. Statistical analyses

For each colony, growth was summarized into a single value asgrowth over time using a linear regression model (Excel 2010,Microsoft office); the summary statistics from each repetitionwas then used for the statistical analysis. Growth rates for eachtemperature regime were analyzed separately using a mixedeffects model with each experiment repetition (trial) as the ran-dom factor; a treatment was considered significant if p-valueswere less than 0.05. The assumptions of normality and homogene-ity of variance were met without transformations. Post-hoc pair-wise comparisons of the mean rates at each temperature regimewere determined using the ghlt function in the multcomp packagefor R. P-values less than 0.005 (0.05 divided by ten pairwisecomparisons) were considered significant. Data analyses weregenerated using the RStudio (version 0.97.551) for Windows.

The effect of temperature regime on mortality from fungaltreatments of G. mellonella was evaluated by survival analysis(Kaplan and Meier, 1958). Insects surviving beyond day 14 wereconsidered censored. LT50s (time [days] to achieve 50% mortality)were generated based on the Kaplan–Meier model using the LIFET-EST procedure in SAS. Statistically significant differences betweenisolates within a temperature regime were calculated based on

the Cox Proportional Hazards Model using the PHREG procedurein SAS. P-values less than 0.008 (0.05 divided by six pairwisecomparisons) were considered as significant, and statisticallysignificant differences between temperature regimes within anisolate were also calculated using the same model with p-valuesless than 0.005 (0.05 divided by ten pairwise comparisons) beingconsidered significant.

3. Results

3.1. Growth rates at constant temperatures

The growth rates of the ten Metarhizium spp. isolates at the sixtemperatures varied considerably (Table 2). At 20 �C, growth rates(mm/day) were similar among many of the isolates; although,some were statistically different (F9,20 = 5.5, p = 0.001).

At 24 �C, there was slightly more variation in growth among theisolates than at 20 �C (F9,20 = 30.4, p < 0.0001). M. guizhouense iso-late DWR 203, which had the highest mean rate of growth at thistemperature, was statistically different from all other isolatesexcept DWR 356 (Mr) and DWR 346 (Mr). Both of the M. brunneumisolates had their highest growth rates at 24 �C.

There were significant variations among isolates at 28 �C(F9,20 = 33.3, p < 0.0001); all of the M. robertsii and all of the M. guiz-houense except DWR 312 reached their highest rate of growth atthis temperature. DWR 203 (Mg) attained the fastest mean rateof growth among all temperatures and isolates tested; and this ratediffered statistically from those of all other isolates except DWR356 (Mr) and DWR 346 (Mr) at this temperature.

At 32 �C, relatively distinct (F9,20 = 95.5, p < 0.0001) groups wereobserved with only ARSEF 324 (Ma) and DWR 356 (Mr) overlappingstatistically (see Table 2). Increasing culture temperature from28 �C to 32 �C decreased the rate of growth of all isolates, exceptDWR 312 (Mg) and ARSEF 324 (Ma). The growth rates of both ofthe M. brunneum isolates decreased to less than one half of theirgrowth rates at 28 �C.

The growth rates of all isolates were drastically reduced at36 �C, although some isolates still were able to produce a measur-able amount of hyphal extension. ARSEF 324, the only M. acridumisolate included in this study, had the highest rate of growth at36 �C of any isolate; but this was not statistically different fromDWR 203 or DWR 313 (F9,20 = 32.3, p < 0.0001). With many iso-lates, their daily growth was so minimal as to require several daysbefore change became measurable. DWR 261 (Mb) and ARSEF 1095(Mb) had total growth of only 1.7 and 2.3 mm after 10 days, respec-tively; while ARSEF 324 (Ma) had nearly 16 mm diameter coloniesafter 10 days. The highest temperature treatment, 40 �C, was too

Table 2The average rate of mycelial growth (mm/day) of ten Metarhizium spp. isolates kept at constant temperatures. Standard error based on three repetitions, letters designatestatistically significant groups based on p-values less than 0.005 within a column.

Isolate 20 �C 24 �C 28 �C 32 �C 36 �C 40 �C

M. acridumARSEF 324 2.73(±0.05)a 3.49(±0.08)ab 3.88(±0.01)ac 4.02(±0.04)bc 1.65(±0.09)d 0.24(±0.21)a

M. brunneumARSEF 1095 3.51(±0.4)ab 4.47(±0.04)cd 4.28(±0.02)bc 1.34(±0.06)a 0.1(±0.05)a 0.06(±0.02)aDWR 261 4.27(±0.07)b 4.62(±0.10)cd 4.14(±0.04)bc 1.43(±0.12)a 0.04(±0.01)a 0.03(±0.02)a

M. guizhouenseDWR 200 3.22(±0.18)ab 4.23(±0.12)bc 4.58(±0.17)c 3.3(±0.3)b 1.11(±0.16)c 0.0(±0.0)aDWR 203 4.17(±0.09)b 5.66(±0.02)e 6.42(±0.03)e 5.68(±0.18)e 1.24(±0.15)cd 0.03(±0.02)aDWR 312 3.01(±0.17)ab 3.16(±0.11)a 3.0(±0.26)a 3.69(±0.07)b 0.55(±0.19)ab 0.01(±0.01)aDWR 313 3.1(±0.56)ab 4.34(±0.36)c 4.65(±0.49)cd 3.69(±0.12)b 1.2(±0.21)cd 0.0(±0.0)a

M. robertsiiDWR 338 3.18(±0.06)ab 3.44(±0.23)ab 3.54(±0.07)ab 3.42(±0.13)b 0.24(±0.19)a 0.0(±0.0)aDWR 346 3.91(±0.18)ab 5.2(±0.03)de 5.59(±0.09)de 4.87(±0.07)d 0.99(±0.17)bc 0.04(±0.03)aDWR 356 4.12(±0.04)ab 5.24(±0.04)de 5.76(±0.04)e 4.77(±0.27)cd 0.86(±0.22)bc 0.02(±0.01)a

70 C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73

high to allow growth; and all isolates at this temperaturemeasured zero or nearly zero mm per day (F9,20 = 1.3, p = 0.28).

3.2. Cold treatment

Fungal colonies kept at 28 �C constantly and measured every24 h for 72 h grew faster than colonies treated to 4 h or 8 h at5 �C during every 24 h period for 3 days; with the differences pro-portional to the times spent each day at 28 �C. Significant differ-ences between the treatments after exposure to 5 �C for either4 h or 8 h cold per day were observed among all the fungal isolates(F9,20 = 100.6, p < 0.0001 and F9,20 = 32.8, p < 0.0001, respectively)(Table 3). After the 3-day fluctuating-temperature treatment per-iod, all colonies were allowed to grow for 4 days at 28 �C; andthe growth rates obtained with all isolates were the same as thoseof cultures kept at constant 28 �C temperature.

3.3. Heat treatment

The effect of daily 40 �C heat treatments for 3 days on radialgrowth of Metarhizium spp. colonies varied markedly betweenthe isolates (4 h treatment: F9,20 = 38.4, p < 0.0001; 8 h treatment:F9,20 = 80.2, p < 0.0001). DWR 312 (Mg) and ARSEF 324 (Ma), atboth 4 and 8 h treatments, as well as DWR 338 (Mr) at 4 h treat-ment, were not significantly affected by the heat treatment (i.e.,no lasting reduction in growth rates), resulting in growth rates

Table 3Average rates of mycelial growth (mm/day) of Metarhizium spp. during three daily exposure8 h at 40 �C. Standard error based on three repetitions, letters designate statistically signi

Isolate Constant 5 �C Treatment

28 �C 4 h

M. acridumARSEF 324 3.61(±0.06)ab 3.09(±0.04)a

M. brunneumARSEF 1095 4.71(±0.17)b 3.86(±0.11)bDWR 261 4.47(±0.01)b 3.96(±0.07)b

M. guizhouenseDWR 200 5.33(±0.12)c 5.49(±0.17)dDWR 203 6.71(±0.09)c 5.52(±0.14)dDWR 312 2.60(±0.12)ab 3.10(±0.11)aDWR 313 3.88(±0.64)c 5.50(±0.09)d

M. robbertsiiDWR 338 2.87(±0.05)a 2.62(±0.20)aDWR 346 5.61(±0.04)c 5.05(±0.04)cdDWR 356 5.66(±0.03)c 4.86(±0.04)c

similar to that of cold-treated plates (Fig. 1). With the rest of theisolates, being placed at 40 �C significantly reduced their growthrate in comparison to the growth rates observed with 5 �C treat-ments of similar times. After 4 h of 40 �C exposure, the M. brunne-um isolates had the highest percent growth-rate reduction (ARSEF1095 = 22.4% and DWR 261 = 34.6% reduction). During the 16 hperiod following the 8 h of 40 �C exposure, DWR 261 (Mb), DWR356 (Mr) and DWR 346 (Mr) had the highest growth rate reduction(78.3%, 80.1%, 87.2% reduction, respectively). Between the daily 8 hheat exposure, growth rates for DWR 261 (Mb), DWR 346 (Mr), andDWR 356 (Mr) dropped to less than 1 mm/day, effectively stoppinggrowth (Table 3).

3.4. Bioassay

G. mellonella larvae were highly susceptible to infection byMetarhizium; viz., all three isolates tested caused 100% mortalitywithin 5 days of conidia application at the constant 28 �C temper-ature regime (v2 = 381.16, df = 3, p < 0.0001). Mortalities with alltreatments (both temperature regimes and fungal isolates), weresignificantly different from mortality in non-fungus control groups(Table 4). All three isolates induced 50% mortality within 3 days atthe constant 28 �C temperature. Treatments at 5 �C for 4 h or 8 hevery 24 h period increased the time needed to reach 50% mortal-ity (LT50) for all the fungal isolates. The LT50s increased signifi-cantly with heat; both 4 h and 8 h at 40 �C each day increased

s to one of the following regimes: (a) constant 28 �C, (b) 4 h or 8 h at 5 �C, or (c) 4 h orficant groups based on p-values less than 0.005 within a column.

40 �C Treatment

8 h 4 h 8 h

2.32(±0.04)ab 3.10(±0.09)a 2.25(±0.07)c

2.93(±0.01)b 3.00(±0.16)a 0.91(±0.15)a2.99(±0.06)bc 2.59(±0.06)a 0.65(±0.09)a

3.71(±0.31)d 5.02(±0.02)c 2.87(±0.08)cd3.92(±0.16)d 5.08(±0.04)c 2.95(±0.12)d2.66(±0.05)b 3.17(±0.40)a 2.81(±0.10)cd3.74(±0.17)d 4.74(±0.13)bc 2.89(±0.05)d

1.92(±0.03)a 2.64(±0.11)a 1.61(±0.03)b3.90(±0.03)cd 4.11(±0.06)b 0.50(±0.04)a3.64(±0.05)b 4.17(±0.14)b 0.73(±0.23)a

Fig. 1. Relative reduction of growth rate of Metarhizium spp. mycelium duringexposure for three to 40 �C for either 4 h or 8 h per day as a percentage of the ratesof growth after exposure to 5 �C for the same amount of time. Bars represent thestandard error of three repetitions. Letters designate statistically significant groups(p < 0.05) between isolates; lowercase letters compare 4 h treatment group,uppercase compare 8 h treatment group.

C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73 71

the survival time by statistically significant amounts, as well asreducing the visible signs of infection for all isolates (e.g., cuticularmelanization, lethargy, etc.). Variation between the isolates at eachtemperature regime was less obvious, with no statistical differ-ences detected between the isolates at each temperature regime,except for 8 h at 40 �C each day. Also, high temperature had asignificant effect on untreated G. mellonella larvae (v2 = 35.66,df = 4, p < 0.0001); higher mortality was observed in the groupexposed to 8 h at 40 �C than any of the other groups.

4. Discussion

Many pest insects live in geographic regions that have climateswith wide temperature fluctuations; and, therefore, effective bio-logical control agents must both survive temperature extremes ofthese areas, but also must be able to resume growth during periodsof optimal growth temperatures. Based on the findings of thisstudy, many fungal isolates exhibit a post-stress growth delay(PSGD) that would likely inhibit disease development in insecthosts. With all ten isolates we used, colony growth ceased duringexposure of culture plates to cold temperatures (5 �C), that is tosay, the percent growth-rate reduction of the cold-treated plateswas proportional to the time spent away from the 28 �C. This sug-gests that in the field, 5 �C will induce temporary cession of fungalgrowth in target pest insects, e.g., conidia germination (infection)and/or disease development (mycelial growth), and thus a reduc-tion in speed and/or level of Metarhizium spp. fungal infections.Fernandes et al. (2008) observed that at 5 �C several strains of

Table 4The median time (days) required to reach 50% mortality (LT50) of Galleria mellonella larvae eor 8 h of heat or cold with the remainder of each day (20 h or 16 h) at 28 �C. Additionally, lbased on Cox proportional hazard model of survival data.

Daily temperature regime DWR 203 (Mg) DWR 3

Constant 28 �C 3 (3,3)A, a 3 (3,3)4 h at 5 �C 4 (4,4)B, a 4 (4,4)4 h at 40 �C 7 (6,7)C, a 7 (7,8)8 h at 5 �C 4 (4,4)B, a 4 (4,4)8 h at 40 �C 11 (10,13)D, a 14 (13

Statistical significant groupings:Uppercase letter indicates relationship of temperature regimes for each isolate (columnLowercase letters indicate relationship between isolates at each temperature (rows).

M. anisopliae s.l. had low levels of germination; however, whenmoved to 28 �C, all of the isolates attained nearly 100% germina-tion. In a later study, M. anisopliae s.l. and M. acridum strains wereobserved to have no conidial germination when left at 5 �C for15 days; but plates held at 10 �C for 15 days allowed germinationof several of the strains (Fernandes et al., 2010a). All isolates inthe current study had seriously inhibited fungal growth duringexposure to 5 �C, but they immediately resumed normal mycelialgrowth rates after transfer from cold temperatures to optimalgrowth temperatures; indicating that moderately cold tempera-tures did not induce a growth delay (PSGD).

Growth rates of all the isolates used in the current study weresharply reduced by increasing incubation temperature from32 �C to 36 �C; and further elevating the culture temperatureto 40 �C completely stopped mycelial growth. Growth inhibitionat 40 �C was noted in several other studies with entomopathogenicfungi (Brooks et al., 2004; Hallsworth and Magan, 1999; Ouedraogoet al., 1997; Smits et al., 2003). Deleterious effects of elevated tem-peratures, e.g., 30, 40, and 42 �C, on several Metarhizium spp. isclearly documented by Rangel et al. (2010). In the current study,M. acridum isolate ARSEF 324 along with M. robertsii isolatesDWR 203 and DWR 313 grew best at 36 �C. Ouedraogo et al.(1997) also found that ARSEF 324 had minimal, yet clearly visible,growth at 35 �C; and observed that, in general, M. acridum (=M.flavoviride) isolates were more heat tolerant than M. anisopliae s.l.isolates. In many areas where biological control agents might beused, 36 �C is likely to be a common air temperature; and bodytemperatures of target insects may be somewhat higher, especiallyamong orthopteran pests which are known to raise their body tem-peratures through basking in direct sunlight to ward off infection(behavioral fever) (Arthurs and Thomas, 2001; Blanford andThomas, 2001; O’Neill and Rolston, 2007; Ouedraogo et al., 2004;Thomas and Jenkins, 1997).

High temperatures have previously been shown to be extremelyharmful to conidia and mycelium of some Metarhizium strains(Rangel et al., 2005); and, in fact, germination counts were takenat 48 h in the present and several related research experimentsto obviate PSGD in comparing laboratory temperature-limitedresponses of numerous fungal strains. The majority of thermo-tol-erance studies involving fungal entomopathogens have focused onconidial survival, i.e., evaluating high-temperature endurance andcharacteristics which enhance survival (Leng et al., 2011; Liaoet al., 2013; Rangel et al., 2012). In one such study, Zimmermann(1982) noted a PSGD-type response in Metarhizium wherein coni-dia exposed to 45 �C for 30 min had low germination after 24 hincubation at 25 �C, but after incubation for 48 h, germinationincreased significantly (i.e., approached the level of untreated con-idia at 24 h). Also, the vigor of a colony’s growth after heat expo-sure was reported by Liu et al. (2009) to depend on the intensityof the stress treatment (i.e., temperature and time period). Accord-ingly, the phenomenon of PSGD has been observed previously,

xposed to Metarhizium spp. conidia, then held 14 days at constant 28 �C, or held at 4 hower and upper 95% confidence levels are provided. Statistically significant groupings

12 (Mg) DWR 346 (Mr) No-fungus control

A, a 3 (3,3)A, a >14 AB, bA, a 4 (4,4)B, a >14 A, bC, a 5 (5,8)D, a >14 A, bB, ab 5 (5,5)C, b >14 AB, c,15)D, b 12 (11,14)E, ab >14 B, c

s).

72 C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73

including in our own studies on the effects of physical stress onentomopathogenic fungi; but since the focus of these studies wason fungal survival, if the fungi recovered, the delay was not singledout as an important inhibitor of successful biological control withfungi.

The results from the current study clearly demonstrate that col-ony exposure to 40 �C for several hours causes delay of hyphalgrowth in some Metarhizium spp. strains; but, interestingly, has lit-tle effect on others. The difference in growth rates, for isolatesexhibiting short-term growth inhibition after exposure to 40 �Cfor either 4 or 8 h (in comparison to cold-treated colonies), isassumed to be due to the time needed to recover from heat damageversus the immediate resumption of growth by cold treated colo-nies when returned to a growth-permissive temperature (28 �C).This was most clearly evidenced during the 4-day post-treatmentperiod, wherein, when placed at a constant 28 �C, all the transientheat- and cold-treated cultures attained growth rates (mm/day)equivalent to those that had been left at 28 �C for the duration ofthe experiment. This return to pretreatment growth rates indicateslittle or no long-term growth inhibition by the heat or coldtreatments tested in this study.

A fungal isolate’s ability to survive exposure to environmentalstress is highly important to its success as a pest-control agent,and a PSGD reaction of an isolate potentially can be as harmfulto a biocontrol program as conidial mortality. If an isolate is unableto expeditiously recover after high-heat exposure, its mere survivalwill be inadequate for successful insect biocontrol becauserepeated (daily) exposure to heat stress will render it unable tocause full-blown disease development. The PSGD trait serves asan amplifier of a stress factor (heat, in the current study), becausethe negative effects continue for some hours after cessation of thestress. Several authors have observed a positive correlationbetween hypervirulence and rapid germination or vegetativegrowth of Metarhizium strains (Al-Aidroos and Roberts, 1978; Al-Aidroos and Seifert, 1980; Hassan et al., 1989; Inglis et al., 1996;Samuels et al., 1989). Conversely, delayed growth following stressexposure is likely to reduce virulence. Not all studies, however,have observed this correlation, e.g., Ouedraogo et al. (1997)reported that with M. acridum strains, hyphal growth was arelatively poor predictor of mortality levels.

Isolates DWR 312 (Mg) and ARSEF 324 (Ma) were not affectedby either 4 h or 8 h daily heat exposure, and maintained growthrates similar to those seen when these isolates were exposed tocold temperature regimes for the same times. Interestingly, ofthe isolates tested, these two (DWR 312 and ARSEF 324) wereamong the slowest growing at 28 �C. Perhaps their modest col-ony-expansion rates at optimum temperature (28 �C) are due totheir utilizing resources to protect against and/or repair damagecaused by environmental stresses. M. acridum strains, in general,are distinguished by their exceptional heat tolerance (Fernandeset al., 2010a).

The two isolates with the greatest growth delay following heatexposure were DWR 346 (Mr) and DWR 356 (Mr). Both of these iso-lates had high rates of growth compared to other isolates at 28 �C.This severe reduction in growth rate after high-heat exposurecould be devastating for fungal conidia applied in the field, sincethey would be rendered virtually useless during and even afterhot days (40+ �C).

Comparative virulence bioassays were conducted with threeMetarhizium isolates that were selected because of their differinglevels of growth-rate reduction following exposure to 40 �C for4 h or 8 h per day: DWR 312 had no growth-rate reductions,DWR 203 moderate reduction, and DWR 346 high reduction. Dueto its specificity toward orthopteran insects, M. acridum ARSEF324 was not included in this bioassay because it was based on alepidopterous insect, G. mellonella. Differences in LD50s between

the low temperature regimes and the high temperature regimeswere obvious; exposure to low temperature for either 4 h or 8 heach day produced insect-mortality rates similar to those withconstant 28 �C for all the isolates. On the other hand, exposure offungus-infected larvae to high temperature for 4 h or 8 h eachday significantly decreased mortality rates, in comparison to fun-gus-exposed larvae held at constant 28 �C. In fact, 100% mortalitywas not observed by the end of the experiment (14 days) in insectstreated with any of the isolates when they were exposed to 40 �Cfor 8 h daily. Although growth of DWR 346 on agar was consider-ably reduced (in comparison to DWR 312 and DWR 203) by daily8 h heat treatment, DWR 346 was not statistically slower at caus-ing mortality than either of these fungi. DWR 203, with its highgrowth rate, caused mortality faster than the slow growing DWR312 at the daily 8 h 40 �C temperature regime. While not all insectswere killed by the fungal treatments at the 8 h high-temperatureregime, those that survived showed severe signs of infection (i.e.,cuticle melanization and altered behavior). In all likelihood theseinsects, while technically alive, would not be active enough to beconsidered agricultural pests.

The current study did not evaluate the effects of fluctuatingtemperatures on insect metabolism and immunity. However, thisis likely to be an important factor to consider, since interactingorganisms or genotypes may have non-overlapping thermal-performance curves which can limit or enhance the infection/immunity response (Thomas and Blanford, 2003).

Environmental stresses can be debilitating factors that are likelyto strongly influence the success or failure of pest control programsutilizing fungal entomopathogens. Whether the stress is heat, UVradiation, nutrient deprivation, dehydration, or some otherunknown factor, the post-stress impacts on the fungal pathogensare likely to be quite varied. A fungal isolate that survives a stressfactor, but is unable to infect target pests due to its response to thestress (e.g., delayed growth) will have reduced potential as aneffective biological control agent in environments where thatstress factor is common. Temperatures that rise almost daily tostressful levels for many entomopathogenic fungi are a primaryconcern in using these fungi in many insect–pest habitats. There-fore, in addition to evaluating conidial survival in laboratorystudies, heat-induced PSGD should be a primary considerationfor selection of field-appropriate strains.

Acknowledgments

We thank Alessandra Fernandes, Carmina Moore, FabianaBernardo, and Scott Treat for their able assistance in conductingthe research described here. We also thank Patricia Golo andNicolai V. Meyling for assistance in editing the manuscript. We verymuch appreciate research funding provided by USDA/APHIS/CPHSTand the Utah Department of Agriculture and Food, Division of PlantIndustry. We express gratitude to Stefan Jaronski for his pointingout his insights into the negative effects of delayed germinationand growth of entomopathogenic fungi following exposure tostress in many pest–insect habitats. The statistical analyses wereconducted with the assistance of Susan Durham at USU,Department of Biology; and the survival analysis was done at theencouragement of Larry Jech, APHIS-USDA, Phoenix, AZ.

References

Al-Aidroos, K., Roberts, D.W., 1978. Mutants of Metarhizium anisopliae with increasevirulence toward mosquito larvae. Can. J. Genet. Cytol. 20, 211–219.

Al-Aidroos, K., Seifert, A.M., 1980. Polysaccharide and protein degradation,germination, and virulence against mosquitoes in the entomopathogenicfungus Metarhizium anisopliae. J. Invertebr. Pathol. 36 (1), 29–34.

Arthurs, S., Thomas, M.B., 2001. Effect of dose, pre-mortem host incubationtemperature and thermal behaviour on host mortality, mycosis and

C.A. Keyser et al. / Journal of Invertebrate Pathology 120 (2014) 67–73 73

sporulation of Metarhizium anisopliae var. acridum in Schistocerca gregaria.Biocontrol Sci. Technol. 11 (3), 411–420.

Bischoff, J.F., Rehner, S.A., Humber, R.A., 2009. A multilocus phylogeny of theMetarhizium anisopliae lineage. Mycologia 101 (4), 512–530.

Blanford, S., Thomas, M.B., 2001. Adult survival, maturation, and reproduction of thedesert locust Schistocerca gregaria infected with the fungus Metarhiziumanisopliae var. acridum. J. Invertebr. Pathol. 78 (1), 1–8.

Brooks, A.J., Muro, M.A.d., Burree, E., Moore, D., Taylor, D., Wall, R., 2004. Growthand pathogenicity of isolates of the fungus Metarhizium anisopliae against theparasitic mite, Psoroptes ovis: effects of temperature and formulation. PestManage. Sci. 60, 1043–1049.

Butler, M.J., Day, A.W., 1998. Fungal melanins: a review. Can. J. Microbiol. 44 (12),1115–1136.

Faria, M.R., Wraight, S.P., 2007. Mycoinsecticides and mycoacaricides: acomprehensive list with worldwide coverage and international classificationof formulation types. Biol. Control 43 (3), 237–256.

Fernandes, E.K.K., Rangel, D.E.N., Moraes, A.M.L., Bittencourt, V.R.E.P., Roberts, D.W.,2008. Cold activity of Beauveria and Metarhizium, and thermotolerance ofBeauveria. J. Invertebr. Pathol. 98 (1), 69–78.

Fernandes, E.K., Keyser, C.A., Chong, J.P., Rangel, D., Miller, M.P., Roberts, D.W.,2010a. Characterization of Metarhizium species and varieties based onmolecular analysis, heat tolerance and cold activity. J. Appl. Microbiol. 108,115–128.

Fernandes, E.K., Keyser, C.A., Rangel, D.E.N., Foster, R.N., Roberts, D.W., 2010b. CTCmedium: a novel dodine-free selective medium for isolating entomopathogenicfungi, especially Metarhizium acridum, from soil. Biol. Control 54 (3), 197–205.

Foster, R.N., Jaronski, S., Reuter, K.C., Black, L.R., Schlothauer, R., 2010. Explainingmycoinsecticide activity: poor performance of spray and bait formulations ofBeauveria bassiana and Metarhizium brunneum against Mormon cricket in fieldcage studies. J. Orthoptera Res. 19 (2), 303–313.

Foster, R.N., Jaronski, S., Reuter, K.C., Black, L.R., Schlothauer, R., Harper, J., Jech, L.E.,2011. Simulated aerial sprays for field cage evaluation of Beauveria bassiana andMetarhizium brunneum (Ascomycetes: Hypocreales) against Anabrus simplex(Orthoptera: Tettigoniidae) in Montana. Biocontrol Sci. Technol. 21 (11), 1331–1350.

Guerzoni, M.E., Lanciotti, R., Cocconcelli, P.S., 2001. Alteration in cellular fatty acidcomposition as a response to salt, acid, oxidative and thermal stresses inLactobacillus helveticus. Microbiology-Sgm 147, 2255–2264.

Hallsworth, J.E., Magan, N., 1999. Water and temperature relations of growth of theentomogenous fungi Beauveria bassiana, Metarhizium anisopliae, andPaecilomyces farinosus. J. Invertebr. Pathol. 74 (3), 261–266.

Hassan, A.E.M., Dillon, R.J., Charnley, A.K., 1989. Influence of acceleratedgermination of conidia on the pathogenicity of Metarhizium anisopliae forManduca sexta. J. Invertebr. Pathol. 54, 277–279.

Inglis, G.D., Johnson, D.L., Goettel, M.S., 1996. Effects of temperature andthermoregulation on mycosis by Beauveria bassiana in grasshoppers. Biol.Control 7 (2), 131–139.

Kaplan, E.L., Meier, P., 1958. Nonparametric estimation from incompleteobservations. J. Am. Stat. Assoc. 53, 457–481.

Leng, Y., Peng, G., Cao, Y., Xia, Y., 2011. Genetically altering the expression of neutraltrehalase gene affects conidiospore thermotolerance of the entomopathogenicfungus Metarhizium acridum. BMC Microbiol. 11, 32.

Liao, X., Lu, H.L., Fang, W., St Leger, R.J., 2013. Overexpression of a Metarhiziumrobertsii HSP25 gene increases thermotolerance and survival in soil. Appl.Microbiol. Biotechnol. 98, 777–783.

Liu, Q., Ying, S.H., Feng, M.G., Jiang, X.H., 2009. Physiological implication ofintracellular trehalose and mannitol changes in response of entomopathogenicfungus Beauveria bassiana to thermal stress. Antonie Van Leeuwenhoek Int. J.Gen. Mol. Microbiol. 95 (1), 65–75.

McClatchie, G.V., Moore, D., Bateman, R.P., Prior, C., 1994. Effects of temperature onthe viability of the conidia of Metarhizium flavoviride in oil formulations. Mycol.Res. 98, 749–756.

Moore, D., Douro-Kpindou, O.K., Jenkins, N.E., Lomer, C.J., 1996. Effects of moisturecontent and temperature on storage of Metarhizium flavoviride conidia.Biocontrol Sci. Technol. 6 (1), 51–61.

O’Neill, K.M., Rolston, M.G., 2007. Short-term dynamics of behavioralthermoregulation by adults of the grasshopper Melanoplus sanguinipes. J.Insect Sci. 7, 1–14.

Ouedraogo, A., Fargues, J., Goettel, M.S., Lomer, C.J., 1997. Effect of temperature onvegetative growth among isolates of Metarhizium anisopliae and M. flavoviride.Mycopathologia 137 (1), 37–43.

Ouedraogo, R.M., Goettel, M.S., Brodeur, J., 2004. Behavioral thermoregulation in themigratory locust: a therapy to overcome fungal infection. Oecologia 138 (2),312–319.

Rangel, D.E.N., Braga, G.U.L., Anderson, A.J., Roberts, D.W., 2005. Variability inconidial thermotolerance of Metarhizium anisopliae, isolates from differentgeographic origins. J. Invertebr. Pathol. 88 (2), 116–125.

Rangel, D.E.N., Fernandes, É.K.K., Dettenmaier, S.J., Roberts, D.W., 2010.Thermotolerance of germlings and mycelium of the insect-pathogenic fungusMetarhizium spp. and mycelial recovery after heat stress. J. Basic Microbiol. 50(4), 344–350.

Rangel, D.E., Fernandes, E.K., Anderson, A.J., Roberts, D.W., 2012. Culture ofMetarhizium robertsii on salicylic-acid supplemented medium inducesincreased conidial thermotolerance. Fungal Biol. 116 (3), 438–442.

Rothschild, L.J., Mancinelli, R.L., 2001. Life in extreme environments. Nature 409(6823), 1092–1101.

Samuels, K.D.Z., Heale, J.B., Llewellyn, M., 1989. Characteristics relating to thepathogenicity of Metarhizium anisopliae toward Nilaparvata lugens. J. Invertebr.Pathol. 53 (1), 25–31.

Smits, N., Briere, J.F., Fargues, J., 2003. Comparison of non-linear temperature-dependent development rate models applied to in vitro growth ofentomopathogenic fungi. Mycol. Res. 107, 1476–1484.

Thomas, M.B., Blanford, S., 2003. Thermal biology in insect–parasite interactions.Trends Ecol. Evol. 18 (7), 344–350.

Thomas, M.B., Jenkins, N.E., 1997. Effects of temperature on growth of Metarhiziumflavoviride and virulence to the variegated grasshopper, Zonocerus variegatus.Mycol. Res. 101, 1469–1474.

Wang, Z.-X., Zhou, X.-Z., Meng, H.-M., Liu, Y.-J., Zhou, Q., Huang, B., 2014.Comparative transcriptomic analysis of the heat stress response in thefilamentous fungus Metarhizium anisopliae using RNA-Seq. Appl. Microbiol.Biotechnol., 1–9. http://dx.doi.org/10.1007/s00253-014-5763-y.

Zimmermann, G., 1982. Effect of high temperatures and artificial sunlight on theviability of conidia of Metarhizium anisopliae. J. Invertebr. Pathol. 40 (1), 36–40.