The effects of two entomopathogenic fungal endophytes, Beauveria bassiana and Purpureocillium lilacinum (formerly Paecilomyces lilacinus), were assessed on the reproduction of cotton aphid, Aphis gossypii Glover (Homoptera:Aphididae), through in planta feeding trials. In replicate greenhouse and field trials, cotton plants (Gossypium hirsutum) were inoculated as seed treatments with two concentrations of B. bassiana or P. lilacinum conidia. Positive colonization of cotton by the endophytes was confirmed through potato dextrose agar (PDA) media plating and PCR analysis. Inoculation and colonization of cotton by either B. bassiana or P. lilacinum negatively affected aphid reproduction over periods of seven and 14 days in a series of greenhouse trials. Field trials were conducted in the summers of 2012 and 2013 in which cotton plants inoculated as seed treatments with B. bassiana and P. lilacinum were exposed to cotton aphids for 14 days. There was a significant overall effect of endophyte treatment on the number of cotton aphids per plant. Plants inoculated with B. bassiana had significantly lower numbers of aphids across both years. The number of aphids on plants inoculated with P. lilacinum exhibited a similar, but non-significant, reduction in numbers relative to control plants. We also tested the pathogenicity of both P. lilacinum and B. bassiana strains used in the experiments against cotton aphids in a survival experiment where 60% and 57% of treated aphids, respectively, died from infection over seven days versus 10% mortality among control insects. Our results demonstrate (i) the successful establishment of P. lilacinum and B. bassiana as endophytes in cotton via seed inoculation, (ii) subsequent negative effects of the presence of both target endophytes on cotton aphid reproduction using whole plant assays, and (iii) that the P. lilacinum strain used is both endophytic and pathogenic to cotton aphids. Our results illustrate the potential of using these endophytes for the biological control of aphids and other herbivores under greenhouse and field conditions.
Citation: Castillo Lopez D, Zhu-Salzman K, Ek-Ramos MJ, Sword GA (2014) The Entomopathogenic Fungal Endophytes Purpureocillium lilacinum (Formerly Paecilomyces lilacinus) and Beauveria bassiana Negatively Affect Cotton Aphid Reproduction under Both Greenhouse and Field Conditions. PLoS ONE 9(8): e103891. https://doi.org/10.1371/journal.pone.0103891
Editor: Thomas L. Wilkinson, University College Dublin, Ireland
Received: February 14, 2014; Accepted: July 7, 2014; Published: August 5, 2014
Copyright: © 2014 Castillo Lopez et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was supported by Cotton Incorporated Grant and Good Neighbor Scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: This study was funded in part by a grant from the Cotton Incorporated Core Program (#12-387) to GAS. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.
Fungal endophytes can protect plants from a wide range of stressors including insect pests . In this study we refer to an endophyte as defined by Schulz (2005)  to be microorganisms (fungi or bacteria) found in asymptomatic plant tissues for all or part of their life cycle without causing detectable damage to the host. The need for the development of new strategies for the control of agricultural insect pests continues to increase due to factors such as development of insecticide resistance –. Here we focus on entomopathogenic fungal endophytes  and the ecological role these fungi can play in agricultural systems.
Entomopathogenic fungal endophytes have been isolated from a variety of different plant species and tissues, and can be inoculated to establish endophytically in a range of other plants to test for adverse effects, if any, on different insect herbivores  –. These entomopathogenic fungal endophytes are classified as non-clavicipitaceous ; referring to fungal endophytes that are usually horizontally transmitted. Clavicipitaceous endophytes, on the other hand, are found in grasses and are typically vertically transmitted, potentially leading to an obligate relationship and higher infection rates with their hosts –. Clavicipitaceous endophytes, named true endophytes, have been studied more extensively than non-clavicipitaceous species and are generally considered mutualistic. Evidence suggests that these fungal endophytes can significantly improve host plant tolerance to drought, insects, diseases, and nematodes, and in exchange, plants provide protection, nutrition and dissemination of the fungi .
A number of benefits to plants are also conferred by non-clavicipitaceous endophytes  –. As endophytes, several non-clavicipitaceous entomopathogens including Beauveria bassiana, Lecanicillium lecanii, Metharizium anisoplae and Isaria spp. can have negative effects on insect pests when in planta, antagonize plant pathogens and promote plant growth  . The activity of B. bassiana has received particular attention due to its negative effects on a variety of insect herbivores including the cotton aphid  –.
The fungus P. lilacinum is more widely known as Paecilomyces lilacinus, having undergone a recent taxonomic revision . To our knowledge there are no studies demonstrating P. lilacinum as an endophytic fungus causing negative effects on insect herbivores, but there are reports of it being pathogenic to a number of insects including Ceratitis capitata, Setora nitens, A. gossypii, and Triatoma infestans –. Both B. bassiana and P. lilacinum are commercially available for use as biocontrol agents, but P. lilacinum is mainly considered to be a nematophagous, egg-parasitizing fungus, specifically against root-knot nematode, Meloidogyne incognita, and several other nematode species including Radopholus similis, Heterodera spp, Globodeera spp –.
Cotton aphids, A. gossypii, have a broad range of host plants including cultivated cotton, causing damage directly by plant feeding and indirectly through virus transmission and physical contamination of cotton by honeydew production . Most commonly, A. gossypii is considered a mid- to late-season pest in cotton. However, extensive use of insecticides such as pyrethroids can decrease its natural enemy community, thereby contributing to the establishment of the aphid as a season-long pest across cotton production areas –. Chronic insecticide use for aphid control has also increased its resistance to several classes of insecticides –. Considering the increasing need for alternative insect management strategies in agricultural systems, we investigated the effects of two entomopathogens, B. bassiana and P. lilacinum, on the cotton aphid when present endophytically in cotton. Specifically, we tested: 1) the ability of B. bassiana and P. lilacinum to establish as endophytes in cotton seedlings when inoculated at the seed stage, and 2) the effects of these endophytes on cotton aphid reproduction using in planta feeding trials in both greenhouse and field environments.
Materials and Methods
Plants and endophytic fungi strains
The cotton seeds used for all experiments were variety LA122 (All-Tex Seed, Inc.). The P. lilacinum strain was isolated from a field survey of naturally-occurring fungal endophytes in cotton . This strain was confirmed to be P. lilacinum (formerly P. lilacinus) by diagnostic PCR and subsequent sequencing of the ribosomal ITS region using specific species primers . The B. bassiana was cultured from a commercially obtained strain (Botanigard, BioWorks Inc, Victor, NY). Stock spore solutions of each fungus were made by adding 10 ml of sterile water to the fungi cultured on potato dextrose agar (PDA) in 10 cm diameter petri dish plates and scraping them with a sterile scalpel. The resulting mycelia and spores were then filtered through cheese cloth into a sterile beaker. A haemocytometer was used to calculate the conidia concentrations of the resulting stock solutions. Final treatment concentrations were reached by dilution using sterile water.
Cotton seed inoculation
Seeds were surfaced sterilized prior to soaking in different spore concentrations by immersion in 70% ethanol for 3 minutes with constant shaking, then 3 minutes in 2% sodium hypochlorite (NaOCl) followed by three washes in sterile water, based on Posada et al. . The third wash was plated on PDA media to confirm surface sterilization efficiency. Seeds were then soaked for 24 hours in two different spore concentrations of the two fungi and sterile water was used as control. Spore concentrations for each fungus were zero (control), 1×106 spores/ml (treatment 1) and 1×107 spores/ml (treatment 2) based on inoculum concentrations used in previous studies of endophytic entomopathogens  –  . Beakers containing the seeds were placed in a dark environment chamber at 28°C until the next day for planting. Soaked seeds were planted in individual pots (15 cm diameter) containing unsterilized Metro mix 900 soil consisting of 40–50% composted pine bark, peat moss, vermiculite, perlite and dolomitic limestone (Borlaug Institute, Texas A&M). All plants were grown in a greenhouse at ∼25°C with natural photoperiod for the duration of the experiment. Pots were placed in a complete randomized design, watered as needed, and no fertilizer was applied throughout the experiments.
Confirmation of plant colonization by endophytic fungi
We have no reason to assume that 100% of the endophyte-treated plants are always colonized by the endophytes when inoculated as seed treatments. Given this constraint, we decided to use two detection methods simultaneously, PDA culturing and diagnostic PCR analysis, to positively confirm the presence of the target endophytes in the experimental plants from the greenhouse experiments, but not for our field experiments. At the end of each greenhouse trial, all treated and control plants were harvested, and each plant was cut in half longitudinally using a sterile scalpel. Fragments of leaves of 1 cm2, stems and roots of 1 cm length were plated on PDA media and placed in growth chamber at 28°C to check for presence of the endophytes. The other half of the plant was freeze dried and DNA was extracted utilizing the CTAB protocol . Species specific oligonucleotide primers for B. bassiana 5′CGGCGGACTCGCCCCAGCCCG 3′, 3′ CCGCGTCGGGGTTCCGGTGCG 5′  and P. lilacinum 5′ CTCAGTTGCCTCGGCGGGAA 3′, 3′ GTGCAACTCAGAGAAGAAATTCCG 5′  (Sigma-Aldrich, Inc St Louis, MO) were used for diagnostic PCR assays. PCR products were visualized on a 2% agarose gel to determine the presence of the inoculated fungal endophytes based on amplification of a DNA fragment of the expected size (positive control). Given the larger size of the plants utilized in our field trials and the impracticality of PDA plating and extracting genomic DNA from entire large plants, we did not test for the presence of the target endophytes in the experimental plants. Instead, we analyzed our data as treatment groups [control, B. bassiana (106), B. bassiana (107), P. lilacinum (106) and P. lilcainum (107)] with concentration effects nested within endophyte treatment and present our results as such.
Cotton aphid reproduction tests
A colony of A. gossypii was maintained on caged cotton plants in the same greenhouse as the experimental plants as described above. For all endophyte-aphid greenhouse trials, second instar nymphs were placed directly on to the experimental control and endophyte-treated cotton plants. Experimental and control plants with aphids were placed in individual clear plastic cages of 45 cm height and 20 cm diameter, then sealed on top with no-see-um mesh (Eastex products, NJ) to avoid aphid escape or movement between plants.
B. bassiana cotton aphid greenhouse experiments
Greenhouse assays of the effects of endophytic B. bassiana on cotton aphid reproduction consisted of three independent tests, each utilizing slightly different protocols. The first was initiated when plants were 13 days old (1st true leaf stage) with aphids allowed to feed for seven days on 10 plants per treatment group. For the second trial, we used older plants (20 days old/third true leaf stage) and aphids were left to reproduce for a longer period of time (14 days) on 10 plants per treatment. At the end of each trial, total aphid numbers were recorded on each individual plant. The third independent test consisted of only a single reproduction trial in which ten 2nd instar aphids were placed on 15 day old plants (second true leaf stage) and left to reproduce 14 days on 15 plants per treatment group, but the cohorts of aphids on each plant were sampled twice at 7 and then again at 14 days.
P. lilacinum cotton aphid greenhouse experiments
We conducted two replicate experiments testing for effects of endophytic P. lilacinum on cotton aphid reproduction utilizing the same reproduction test protocol for each trial. In these trials, ten 2nd instar aphids were left to reproduce on the same plants for 14 days consecutively and sampled twice at 7 and then again at 14 days. Ten 1st true leaf stage plants per treatment group were utilized for the first trial; 15 plants per treatment group were used for the second trial.
Cotton aphid field trials for both B. bassiana and P. lilacinum
During the summers of 2012 and 2013, experimental field trials were conducted at the Texas A&M University Field Station located near College Station in Burleson, Co., TX (N 30° 26′ 48′′ W 96° 24′ 05.12′′) at an elevation of 68.8 m. We utilized a randomized block design with five seed inoculation treatments (T1: Control, T2: B. bassiana 1×106, T3: B. bassiana 1×107, T4: P. lilacinum 1×106 and T5: P. lilacinum 1×107). Surface sterilized seeds were inoculated with the different treatments as described in our greenhouse assay protocol. Treatments were replicated six times, making a total of 30 plots in the field. Each plot was comprised of 4 rows of 16.6 m length and planted with 15 seeds per meter. For the aphid reproduction experiments, we utilized the same protocol during both field seasons whereby a total of 75 cone shaped metal framed cages (0.35 m of height) were randomly assigned to be placed over endophyte-inoculated and control plants (15 cages/treatment) and set up on May 17, 2012 and June 24, 2013, respectively (delayed experiment due to rain in 2013). Predators were eliminated if found prior to enclosing the caged plants with no-see-um mesh (Eastex products, NJ) to prevent aphid escapes and entrance of predators. Ten second instar aphid nymphs from the laboratory colony were placed on each plant and left to reproduce for 14 days. At the end of the experiment, cages were removed, the entire plant was bagged and brought back to laboratory for total aphid number counts.
Fungal pathogenicity experiment
To assess pathogenicity of both the P. lilacinum strain recovered in our endophyte survey of cotton , and the commercial B. bassiana strain utilized in our endophyte trials, we performed a cotton aphid survival experiment as per Gurunlingappa et al.  and Vega et al. 2008  with slight modification. The same spore concentrations used in our endophyte in planta experiment were used for this test for both endophytes (0, 1×106 and 1×107 spores/ml). Thirty 2nd instar aphids per treatment were dipped in spore solutions for 5 seconds, and then placed on fresh cotton leaves kept on moistened filter paper (to prevent drying out) inside 10 cm diameter petri dishes sealed with parafilm (Bemis flexible packaging, Neenah, WI). Ten aphids per petri dish were placed in three replicate petri dishes per treatment. Aphids were checked daily for mortality and dead aphids were removed, plated and incubated on PDA media to confirm emergence of the entomopathogens from aphid cadavers.
All data were tested for normality assumptions using a qqplot, Levene's homogeneity test and the Shapiro-Wilk normality test at alpha = 0.05 significance level. For the first independent B. bassiana greenhouse experiment, ANOVA and t-tests were performed to compare aphid reproduction differences among plants after 7 days of feeding. In the second and third B. bassiana tests, the data were non-normal and nonparametric Kruskal-Wallis and Mann-Whitney U tests were used. For both P. lilacinum greenhouse trials, a repeated measures ANOVA was performed with time as a repeated factor to test for differences in aphid numbers between plants after 7 and 14 days of reproduction because aphids on the same plants were sampled sequentially. Aphid field trials for both 2012 and 2013 were analyzed using ANOVA followed by pairwise comparisons (control vs. treatment). We conducted a combined ANOVA analysis of the field data across both 2012 and 2013 to test for year, treatment, and year by treatment effects. For the cotton aphid pathogenicity experiment, a Kaplan-Meier survival analysis was performed to compare the cumulative survival of treated vs. untreated control aphids. All analyses were conducted using SPSS 22 (IBM SPSS, Armonk NY).
Plant colonization by endophytic fungi
Our culturing results showed no fungal growth on the PDA plating of the third sterile water wash of either the surface sterilized seeds or plant samples, indicating the efficacy of our surface sterilization. Thus, we assume that the fungi growing in the media from surface-sterilized plant materials were endophytes that came from within plant tissues and not epiphytes from the plant surface. Utilizing combined PDA plating and diagnostic PCR detection methods revealed 30–45% more instances of positive endophytic colonization relative to PDA plating alone. B. bassiana was detected in 35% and 55% of the treated plants in the first (7 day) and second (14 day) greenhouse trials, respectively. For the third B. bassiana trial which consisted of using the same plants for both measurements of aphid reproduction at 7 and 14 days, B. bassiana was detected in 53.3% of the treated plants. In the P. lilacinum experiments, the target endophyte was detected in 55% and 45% of plants in the first and second trials, respectively.
B. bassiana cotton aphid greenhouse experiments
Our results were analyzed both as treatments (control, low and high concentration) and by confirmed positive colonization of plants by the target endophyte (colonized vs. uncolonized). In the first test, the mean number of cotton aphids per plant on B. bassiana treated plants was not significantly different from those on control plants after 7 days of reproduction when analyzed by treatment groups (F = 2.07; df = 2,29; P = 0.145), but was significantly different when analyzed by positive colonization of the endophyte (t-test; P = 0.014) (Fig 1a). In the second test, we observed a significant negative effect on reproduction of cotton aphids after 14 days when analyzed by treatment groups (Kruskal-Wallis = 6.744; P = 0.034) as well as by positive colonization of the endophyte (Mann Whitney U = 44; P = 0.004) (Fig 1b). In our third B. bassiana trial, there was no significant effect on the number of aphids per plant after 7 days when analyzed by treatment (Kruskal-Wallis = 4.74; P = 0.093), but there was a significant effect on aphids when analyzed by positive colonization by the endophyte (Mann-Whitney U = 60.50; P = <0.0001) (Fig 1c). Similarly at the end of the 14 days in the same experiment, there were no significant effects on the number of aphids when the data were analyzed by treatment (Kruskal Wallis = 3.069; P = 0.216), but a significant effect was observed when the data were analyzed by plant positive colonization by the endophyte (Mann Whitney U = 58; P<0.0001) (Fig 1d).
Cotton aphid reproduction on plants positively colonized by endophytic B. bassiana versus uncolonized plants after (a) 7 days in the first trial, (b) 14 days in the second trial, and (c) 7 and (d) 14 days successively in the third trial.
P. lilacinum cotton aphid greenhouse experiments
As with the B. bassiana trials above, we present the results of analyses categorizing the data as both treatment groups and positive versus negative colonization. In the first P. lilacinum trial, aphid numbers varied significantly with time (Repeated Measures ANOVA F = 60.40; df = 1,28; P = 0.0001), but no significant endophyte treatment effect was observed when data were analyzed by plant positive colonization (F = 0.026; df = 1,28; P = 0.873). However, when analyzed based on treatment groups, there was a significant effect of time (F = 69.56; df = 1,27; P<0.0001) as well as endophyte treatment (F = 140.48; df = 2,27; P = 0.049) (Fig 2a). After increasing our sample size in the second trial, we observed a significant effect of both time (F = 53.73; df = 1,42;P = 0.0001) and treatment when analyzed based on plant positive colonization by the endophyte (F = 8.05; df = 1,42; P = 0.007) (Fig 2c). Although there was a significant effect of time (F = 52.52; df = 1,41; P<0.000) on the number of aphids when we analyzed our data by treatment groups (control, low or high concentration), the effect of endophyte treatment was not significant (F = 0.546; df = 241; P = 0.583).
Cotton aphid field trials of both B. bassiana and P. lilacinum
In both 2012 and 2013 there was no effect of seed treatment spore concentration within each endophyte treatment (2012 Nested ANOVA, F = 1.95; df = 2,77; P = 0.149 and 2013 Nested ANOVA F = .935; df = 2,67; P = 0.398), therefore data from both concentrations were grouped for each endophyte in subsequent analyses. Across both years of the field trial, there was a significant effect of endophyte treatment (ANOVA, F = 7.31; df = 5,132; P = 0.001) and also a significant year effect (ANOVA, F = 17.43; df = 5,132; P<0.0001), but no endophyte by year interaction (ANOVA, F = 0.547; df = 5,132; P = 0.580). During the summer of 2012, there was a significant overall effect of endophyte treatment on the number of cotton aphids per plant at the end of 14 days of reproduction (ANOVA, F = 4.12; df = 2,73; P = 0.02). Follow-up pairwise comparisons revealed that there were significantly fewer aphids on cotton plants from B. bassiana-treated vs. control plots (P = 0.006). The difference in aphid numbers on plants in P. lilacinum-treated vs. control plots exhibited a similar but non-significant reduction (P = 0.085) (Fig 3a). Similarly in 2013, there was a significant overall effect of endophyte treatment on aphid reproduction at the end of 14 days (ANOVA, F = 3.13; df = 2,59; P = 0.05). Pairwise comparisons indicated that inoculation of plants with B. bassiana had a significant negative effect on aphid reproduction vs. control (P = 0.016), but only a non-significant trend was observed with P. lilacinum vs. the control (P = 0.086) (Fig 3b).
Cotton aphid survival experiment
There was no significant difference in aphid mortality between those treated with two different concentrations (1×106 or 1×107) of conidia solutions of each fungus. Thus, the data from both concentrations were pooled and analyzed together for each fungus. There was a highly significant increase in mortality between aphids treated with either P. lilacinum (60%) or B. bassiana (57%) vs. the controls (10%) (Kaplan-Meier, P<0.0001 for both fungi).
Our results provide the first report of the negative effects of two endophytic entomopathogenic fungi, B. bassiana and P. lilacinum, on cotton aphid reproduction when feeding on whole intact cotton plants inoculated as seed treatments. Importantly, we observed negative effects under both greenhouse and field conditions. We also provide the first evidence for an endophytic effect of P. lilacinum on herbivorous insect performance.
After analyzing our data based on positive plant colonization by the target endophyte, we found that aphid reproduction on cotton plants positively colonized by B. bassiana was reduced in three independent greenhouse trials. Although the results of our first trial testing the effects of P. lilacinum as an endophyte on aphid reproduction revealed only a significant effect of time but not treatment, we attributed this to a small sample size for the given effect size based on the results of power analysis (Power = 0.175) (Fig. 2b). After increasing the sample size in the second P. lilacinum trial, we observed a significant effect of both time and treatment on the reproduction of cotton aphid with lower aphid numbers on endophyte-colonized plants (Figs. 2c & 2d). Our greenhouse endophyte trial results using A. gossypii are similar to those of Martinuz et al.  in which whole squash plants were inoculated with Fusarium oxysporum as an endophyte via soil drench, resulting in negative effects on A. gossypii choice and performance. Similarly, Akello et al.  showed that Aphis fabae feeding on bean plants colonized independently by strains of either B. bassiana, Trichoderma asperellum or Gibberella moniliformis reproduced poorly compared to those on control plants. Both Martinuz et al.  and Akello et al  attribute the negative effects on aphid fitness to be due to chemical changes in the plant that were systemically induced by the presence of the endophyte, though the specific mechanism by which these fungi activated a systemic response within the plants was not investigated.
The ability of B. bassiana to establish as an endophyte across a range of plants has been well established [e.g., cotton, corn, bean, wheat, pumpkin, tomato ; coffee ); sorghum ; banana ; tomato ; jute  and pine . A number of plant-endophyte-insect interaction experiments, including a cotton aphid study by Gurunlingappa et al.  have been performed using cut leaf bioassays rather than whole intact plant experiments  –. Utilizing leaf cuts rather than whole intact plants can potentially cause release of allelochemicals due to direct plant damage that may have negative effects on insects that could obscure those caused by the presence of an endophyte . Alternatively, cutting plants and abscising leaves may induce changes in plant chemistry that alter the interaction between the endophyte and host in ways not observed in intact plants . Demonstrations of negative effects of endophytic entomopathogens including B. bassiana on herbivores in more natural whole plant feeding assays are relatively rare, but have been shown for a few species including aphids –. Similarly, there are only a few examples of negative effects on lepidopteran species caused by endophytic colonization by B. bassiana using whole plant assays including Ostrinia nubilalis and Helicoverpa zea  .
To our knowledge, there are no reports in the literature of negative endophytic effects of P. lilacinum on herbivorous insects. This is not surprising since this fungus was until recently thought to mainly have pathogenic properties against nematodes and not insects. Historically, P. lilacinum has been considered largely as a soil-born nematode egg parasite and used as a biocontrol agent against nematode pests such as root-knot, Meloidogyne incognita, and reniform, Rotylenchulus reniformis, nematodes –. However, recent evidence indicates that P. lilacinum can also be an entomopathogen –. Our results indicate that the P. lilacinum strain isolated from cotton by Ek-Ramos et al.  can negatively affect insect herbivores when present as an endophyte and that it is also pathogenic to insects. Interestingly, the same strain has also been observed to parasitize root-knot nematode eggs in simple lab bioassays and negatively affect nematode reproduction when present as an endophyte in in planta assays (W. Zhou, J.T. Starr and G.A. Sword, unpublished results).
The mechanisms by which herbivores can be negatively affected by clavicipitaceous obligate endophytes have been studied in a few different grass species and can vary from antixenosis and/or antibiosis mediated by constitutive production and or induction of secondary compounds produced by the plant – or secondary metabolites produced by the endophytes themselves   –,. It is important to mention that infection rates of natural populations of grasses by these endophytes can vary depending on the genetic and environmental background of the population and these factors can determine if this symbiosis goes from mutualistic to antagonistic –. Another hypothesis for the mechanism by which endophytes can negatively affect herbivores is based on the idea that endophytes can alter the phytosterol profiles of plants and compete with insects for these compounds which are essential for their development  . The mechanisms by which entomopathogenic endophytic fungi may protect plants from insect herbivores are unknown. Although these endophytes do produce secondary metabolites  , we do not know if this is the main cause for the negative effects on aphids when feeding on endophytically-colonized plants observed in our study. The literature also suggests a systemic response in the plant can be induced by the presence of some entomopathogenic endophytes including B. bassiana that confers resistance against plant pathogens –. Whether an induced systemic response accounts for the negative effects on insects observed in our study remains to be determined.
The mode of establishment and duration of presence of endophytic fungi in plants varies among the different plant-endophyte combinations tested to date  – –. In some cases, intentionally inoculated endophytes can be retained within plants for considerable amounts of time, including B. bassiana found for as long as eight months in coffee  or nine months in Pinus radiata . Our study indicates that B. bassiana and P. lilacinum were still present in cotton plants up to 34 days following inoculation as a seed treatment. This duration does not necessarily indicate that B. bassiana and P. lilacinum can only be present in cotton as endophytes for this period of time, but rather that we did not test for the presence/absence of the endophytes beyond 34 days. The average recovery success of the target endophytes used in our studies ranged from 35–55%. Though not a high colonization frequency, we were still able to detect negative effects on aphids feeding on plants colonized by the endophytes. We have not yet rigorously studied the endophytic colonization of cotton by P. lilacinum and B. bassiana, but P. lilacinum was primarily detected in the root tissues whereas B. bassiana was found mostly in the above ground tissues. Fungal endophytes are known to occur throughout an entire plant including leaves, stems, roots and reproductive parts, however, tissue specific presence in plants is not required for negative effects on target herbivores. For example, endophytic fungi inhabiting roots can negatively affect the performance and fitness of caterpillars feeding on above ground tissues , . Our results support this scenario given that P. lilacinum negatively affects aphids feeding on cotton leaves above ground, but is recovered more commonly from below ground root tissues.
The manipulation of endophytic fungi, many of which are completely unstudied, has the potential to protect plants from insect herbivores and other stress factors . We have provided novel evidence showing that the endophytic establishment in cotton of the entomopathogens B. bassiana and P. lilacinum when inoculated as seeds can adversely affect cotton aphid reproduction not only in greenhouse assays, but also under field conditions. Although we observed a significant year effect, this was due to differences in the total aphid numbers across years (Fig. 3a&b). Importantly, there was no year by endophyte treatment interaction effect. Our field results exhibited the same pattern of negative effects of endophytes on cotton aphids across years in both 2012 and 2013. The consistency of results across years under field conditions that can vary in variety of uncontrolled environmental variables (e.g. precipitation and temperature regimes) is particularly encouraging for the potential reliability of incorporating fungal endophyte manipulations into IPM strategies. Future directions of our work include testing these entomopathogenic endophytes against other insect and nematode herbivores along with phytohormone and transcriptomic analysis to investigate the mechanisms by which these endophytes confer protection to their plant hosts.
We would like to thank Cesar Valencia and Josephine Antwi for help provided during greenhouse and field trials both in 2012 and 2013. Dr. Steve Hague from the Soil and Crop Department at Texas A&M University provided critical logistical support for the field trials. Charlie Cook from All-Tex Seed, Inc. generously provided seeds for these experiments.
Conceived and designed the experiments: DCL KZS GAS. Performed the experiments: DCL. Analyzed the data: DCL GAS. Contributed reagents/materials/analysis tools: DCL GAS KZS MJER. Wrote the paper: DCL GAS.
- 1. Porras-Alfaro A, Bayman P (2011) Hidden fungi, emergent properties: endophytes and microbes. Annu Rev Phytopathol 49: 291–315.
- 2. Schulz B, Boyle C (2005) The endophytic continuum. Mycol Res 109: 661–686.
- 3. Gould F (1995) Comparison between resistance management strategies for insects and weeds. Weed Technology 9(4): 830–839.
- 4. Gassmann AJ (2009) Evolutionary analysis of herbivorous insects in natural and agricultural environments. Pest Manag Sci 65: 1174–1181.
- 5. Silva AX, Jander G, SamadiegoH, Ramsey JS, Figueroa CC (2012) Insecticide resistance mechanisms in the green peach aphid Myzus persicae (Hemiptera:Aphididae): A transcriptomic survey. PLOS ONE 7(6): 363–366.
- 6. Vega FE, Mark S, Goettel MS, Blackwell M, Chandler D, et al. (2009) Fungal entomopathogens: new insights in their ecology. Fungal Ecology 2: 149–159.
- 7. Gurulingappa P, Sword GA, Murdoch G, McGee PA (2010) Colonization of crop plants by fungal entomopathogens and their effects on two insect pests when in planta. BioControl 55: 34–41.
- 8. Rodriguez RJ, White JF Jr, Arnold AE, Redman RS (2009) Fungal endophytes: diversity and functional roles. New Pathologist 182(2): 314–330.
- 9. Hartley SE, Gange AC (2009) The impacts of symbiotic fungi on insect herbivores: mutualism in a multitrophic context. Ann Rev Entomol 54: 323–342.
- 10. Schardl CL, Leuchtmann A, Spiering MJ (2004) Symbioses of grasses with seedborne fungal endophytes. Annu Rev Plant Biol 55: 315–340.
- 11. Omacini M, Chaneton EJ, Ghersa CM, Muller CB (2001) Symbiotic fungal endophytes control insect host-parasite interaction webs. Nature 409: 78–81.
- 12. Jung HS, Lee HB, Kim K, Lee EY (2006) Selection of Lecancillium strains for aphid (Myzus persicae) control. Korean J Mycol 34: 112–118.
- 13. Jaber LR, Vidal S (2010) Fungal endophyte negative effects on herbivory are enhanced on intact plants and maintained in a subsequent generation. Ecol Entomol 35: 25–36.
- 14. Gange AC, Eschen R, Wearn J, Thawer A, Sutton B (2012) Differential effects of foliar endophytic fungi on insect herbivores attacking a herbaceous plant. Oecologia 168: 1023–1031.
- 15. Vega FE, Posada F, Aime MC, Pava-Ripoll M, Infante F, et al. (2008) Entomopathogenic fungal endophytes. Biological Control 46: 72–82.
- 16. Bing LA, Lewis LC (1991) Suppression of Ostrinia nubilalis by endophytic Beauveria bassiana. Environ Entomol 20: 1207–1211.
- 17. Posada FJ, Vega FE (2005) Establishment of the fungal entomopathogen Beauveria bassiana (Ascomycota: Hypocreales) as an endophyte in cocoa seedlings (Theobroma cacao). Mycologia 97: 1208–1213.
- 18. Posada F, Aime MC, Peterson SW, Rehner SA, Vega FE (2007) Inoculation of coffee plants with the fungal entomopathogen Beauveria bassiana (Ascomycota:Hypocreales). Mycol Res 111: 748–757.
- 19. Akello J, Dubois T, Coyne D, Kyamanywa S (2008) Endophytic Beauveria bassiana in banana (Musa spp.) reduces banana weevil (Cosmopolites sordidus) fitness and damage. Crop Prot 27: 1437–1441.
- 20. Powell WA, Klingeman WE, Ownley BH, Gwinn KD (2009) Evidence of Endophytic Beauveria bassiana in seed treated tomato plants acting as a systemic entomopathogen to larval Helicoverpa zea (Lepidoptera:Noctuidae). J Entomol Sci 44(4): 391–396.
- 21. Biswas C, Dey P, Satpathy S, Satya P (2011) Establishment of the fungal entomopathogen Beauveria bassiana as a season long endophyte in jute (Corchorus olitorius) and its rapid detection using SCAR marker. BioControl. DOI 10.1007/s10526-011-9424-0.
- 22. Gurunlingappa P, McGee PA, Sword GA (2011) Endophytic Lecanicillum lecanii and Beauveria bassiana reduce the survival and fecundity of Aphis gossypii following contact with conidia and secondary metabolites. Crop Prot 30: 349–353.
- 23. Luangsa-ard JJ, Houbraken J, van Doorn T, Hong SB, Borman AM, et al. (2011) Purpureocillium, a new genus for the medically important Paecilomyces lilacinum. FEMS Microbiol Lett 321: 141–149.
- 24. Imoulan A (2011) Natural occurrence of soil-borne entomopathogenic fungi in the Moroccan Endemic forest of Argania spinosa and their pathogenicity to Ceratitis capitata. J Microbiol & Biotech 27(11): 2619–2628.
- 25. Wakil W, Ashfaq M, Ghazanfar MU, Kwon YJ, Ullah E, et al. (2012) Testing Paecilomyces lilacinus, diatomaceous earth and Azadirachta indica alone and in combination against cotton aphid (Aphis gossypii Glover) (Insecta: Homoptera: Aphididae). African J Biotech 11(4): 821–828.
- 26. Rao NBVC, Snehalatharani A, Emmanuel N (2012) New record of Paecilomyces lilacinus (Deuteromycota: Hyphomycetes) as an entomopathogenic fungi on slug caterpillar of coconut. Insect Environ 17(4): 151–153.
- 27. Marti GA, Lastra CC, Pelizza SA, García JJ (2006) Isolation of Paecilomyces lilacinus (Thom) Samson (Ascomycota: Hypocreales) from the Chagas disease vector, Triatoma infestans Klug (Hemiptera:Reduviidae) in an endemic area in Argentina. Mycopathologia 162(5): 369–72.
- 28. Fiedler Ż, Sosnowska D (2007) Nematophagous fungus Paecilomyces lilacinus (Thom) Samson is also a biological agent for control of greenhouse insects and mite pests. BioControl 52(4): 547–8.
- 29. Sharma S, Trivedi PC (2012) Application of Paecilomyces lilacinus for the control of Meloidogyne incognita infecting Vigna radiate. Indian J Nematol 42(1): 1–4.
- 30. Kannan R (2012) Effect of different dose and application methods of Paecilomyces lilacinus (Thom.) Samson against Root Knot Nematode, Meloidogyne incognita (Kofoidand White) Chitwood in Okra. J Ag Science 4(11): 119–127.
- 31. Carrion G, Desgarennes D (2012) Effect of Paecilomyces lilacinus in free-living nematodes to the rhizosphere associates potatoes grown in the Cofre of Perote region, Veracruz, Mexico. Revista Mexicana de Fitopatologia 30(1): 86–90.
- 32. Khan MR (2012) Management of root-knot disease in eggplant through the application of biocontrol fungi and dry neem leaves. Turkish J Biol 36(2): 161–169.
- 33. Godfrey LD, Fuson KJ, Wood JP, Wright SD (1997) Physiological and yield responses of cotton to mid-season cotton aphid infestations in California. Proc Beltwide Cotton Conferences 1048–1051.
- 34. King EG, Phillips JR, Head RB (1987) 40th Annual conference report on cotton insect research and control. In: JM Brown and DA Richter Proc Beltwide Cotton Prod.
- 35. Godfrey LD, Rosenheim JA, Goodell PB (2000) Cotton aphid emerges as major pest in SVJ cotton. California Agriculture 54(6): 32–34.
- 36. O'Brien PJ, Hardee DD, Grafton-Caldwell EE (1990) Screening of Aphis gossypii for insecticide tolerance. Insecticide and Acaracide Tests 15: 254–255.
- 37. Grafton-Caldwell EE (1991) Geographical and temporal variation in response to insecticides in various life stages of Aphis gossypii (Homoptera:Aphididae) infesting cotton in California. J Econ Entomol 84: 741–749.
- 38. Kerns DL, Gaylor MJ (1992) Insecticide resistance in field populations of the cotton aphid (Homoptera:Aphididae). J Econ Entomol 85: 7–8.
- 39. Ek-Ramos MJ, Zhou W, Valencia CU, Antwi JB, Sword GA (2013) Spatial and temporal variation in fungal endophyte communities isolated form cultivated cotton (Gossypium hirsutum). PLOS ONE 8: e66049.
- 40. Atkins SD, Clark IM, Pande S, Hirsch PR, Kerry BR (2004) The use of real time PCR and species specific primers for the identification and monitoring of Paecilomyces lilacinus. FEMS Microbiol Ecol 51: 257–264.
- 41. Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull 19: 11–15.
- 42. Martinuz A, Schouten A, Menjivar RD, Sikora RA (2012) Effectiveness of systemic resistance toward Aphis gossypii (Aphididae) as induced by combined applications of the endophytes Fusarium oxysporum Fo162 and Rhizobium etli G12. Biological Control 62: 206–212.
- 43. Akello J, Sikora R (2012) Systemic acropedal influence of endophyte seed treatment on Acyrthosipon pisum and Aphis fabae offspring development and reproductive fitness. Biological Control 61: 215–221.
- 44. Reddy N, Ali Khan AP, Devi UK, Sharma HC, Reineke A (2009) Treatment of millet crop plant (Sorghum bicolor) with the entomopathogen fungus (Beauveria bassiana) to combat infestation by the stem borer, Chilo partellus Swinhoe (Lepidoptera:Pyralidae). J of Asia-Pacific Entomol 12: 221–226.
- 45. Brownbridge M, Reay SD, Nelson TL, Glare TR (2012) Persistence of Beauveria bassiana (Ascomycota:Hypocreales) as an endophyte following inoculation of radiata pine seed and seedlings. Biological Control 62(3): 194–200.
- 46. Raps A (1998) Vidal S (1998) Indirect effects of an unspecialized endophytic fungus on specialized plant-herbivorous insect interactions. Oecologia 114: 541–547.
- 47. McGee PA (2002) Reduced growth and deterrence from feeding of the insect pest Helicoverpa armiguera associated with fungal endophytes from cotton. Australian J Exp Ag 42(7): 995–999.
- 48. Vicari M, Hatcher PE, Ayres PG (2002) Combined effect of foliar and mycorrhizal endophytes on an insect herbivore. Ecology 83: 2452–2464.
- 49. Price PW, Denno RF, Eubanks MD, Finke DL, Kaplan I (2011) Plant and herbivore interactions. In: Insect Ecology. pp. 121–127. Cambridge University Press.
- 50. Munawar F, Bhat MY, Ashaq M (2011) Combined application of Paecilomyces lilacinus and carbosulfan for management of Meloidogyne incognita and Rotylenchulus reniformis. Ann Plant Protection Sciences 19(1): 168–173.
- 51. Kiewnick S (2011) Effect of Meloidogyne incognita inoculum density and application rate of Paecilomyces lilacinus strain 251 on biocontrol efficacy and colonization of egg masses analyzed by real-time quantitative PCR. Phytopathology 101(1): 105–12.
- 52. Chaudhary KK, Kaul RK (2012) Compatibility of Pausteria penetrans with fungal parasite Paecilomyces lilacinus against root knot nematode on chilli: Capsicum annuum. South Asian J Exp Biol 1(1): 36–42.
- 53. Clay K, Marks S, Cheplick GP (1993) Effects of insect herbivory and fungal endophyte infection on competitive interactions among grasses. Ecology 74: 1767–77.
- 54. Clay K (1996) Interactions among fungal endophytes, grasses and herbivores. Res Popul Ecol 38(2): 191–201.
- 55. Carriere Y, Bouchard A, Bourassa S, Brodeur J (1998) Effect of endophyte incidence in perennial ryegrass on distribution, host choice, and performance of the hairy chinch bug (Hemiptera: Lygaeidae). J Econ Entomol 91: 324–328.
- 56. Gindin G, Barash I, Harari N, Raccah B (1994) Effect of endotoxic compounds isolated from Verticillium lecanii on the sweetpotato whitefly, Bemisia tabaci. Phytoparasitica 22: 189–196.
- 57. Wang L, Huang J, You M, Guan X, Liu B (2007) Toxicity and feeding deterrence of crude toxin extracts of Lecanicillium (Verticillium) lecanii (Hyphomycetes) against sweet potato whitefly, Bemisia tabaci (Homoptera: Aleyrodidae). Pest Manag Sci 63: 381–387.
- 58. Ball OJP, Barker GM, Prestidge RA, Lauren DR (1997a) Distribution and accumulation of the alkaloid peramine in Neotyphodium lolii infected perennial ryegrass. J Chem Ecol 23: 1419–1434.
- 59. Ball OJP, Miles CO, Prestidge RA (1997b) Ergopeptine alkaloids and Neotyphodium lolii mediated resistance in perennial ryegrass against adult Heteronychus arator (Coleoptera: Scarabaeidae). J Econ Entomol 90: 1382–1391.
- 60. Latch GCM (1993) Physiological interactions of endophytic fungi and their hosts: biotic stress tolerance imparted to grasses by endophytes. Agriculture, Ecosystems and Environment 44: 143–156.
- 61. Bush LP, Wilkinson HH, Schardl CL (1997) Bioprotective alkaloids of grass-fungal endophyte symbioses. Plant Physiol 114(1): 1–7.
- 62. Dugassa-Gobena D, Raps A, Vidal S (1996) Einuû von Acremonium strictum auf den Sterolhaushalt von Panzen: ein moÈ glicher Faktor zum veraÈ nderten Verhalten von Herbivoren. Mitt Biol Bundesanst 321: 299.
- 63. Saikkonen K, Wali PR, Helander M (2010) Genetic compatibility determines endophyte grass combinations. PLOS ONE 5(6): e11395.
- 64. Saari S, Helander M, Faeth SH (2010) The effects of endophytes on seed production and seed predation of tall fescue and meadow fescue. Microb Ecol 60: 928–934.
- 65. Rasmussen S, Parksons AJ, Fraser K, Xue H, Newman JA (2008) Metabolic profiles of Lolium perenne are differentially affected by Nitrogen supply, Carbohydrate content and fungal endophyte infection. Plant Physiol 146(3): 1440–1453.
- 66. Saikkonen K, Lehtonen P, Helander M, Koricheva J, Faeth SH (2006) Model systems in ecology: dissecting the endophyte-grass literature. Trends Plant Sci 11: 428–433.
- 67. Young C, Wilkinson H, (2010) Epichloe endophytes: models of an ecological strategy. In Cellular and Molecular Biology of Filamentous Fungi. ASM Press, Whashington DC. Pp 660–671.
- 68. Ownley BH, Griffin MR, Klingeman WE, Gwinn KD, Moulton JK, et al. (2008) Beauveria bassiana: endophytic colonization and plant disease control. J Invertebr Pathol 3: 267–270.
- 69. Ownley BH, Gwinn KD, Vega FE (2010) Endophytic fungal entomopathogens with activity against plant pathogens: ecology and evolution. BioControl 55: 113–128.
- 70. Vega FE, Posada F, Aime MC, Pava-Ripolli M, Infante F, et al. (2008) Entomopathogenic fungal endophytes. Biological Control 46: 72–82.
- 71. Raps A, Vidal S (1998) Indirect effects of an unspecialized endophytic fungus on specialized plant – herbivorous insect interactions. Oecologia 114: 541–547.