Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 220 2016 SOUTHEASTERN NATURALIST 15(2):220–234 Fungi Associated with Solenopsis invicta Buren (Red Imported Fire Ant, Hymenoptera: Formicidae) from Mounds in Mississippi Sandra Woolfolk1,2, C. Elizabeth Stokes1, Clarence Watson3, Gerald Baker1, Richard Brown1, and Richard Baird1,* Abstract - In 2004, we determined baseline data on fungal-community assemblages from Solenopsis invicta (Red Imported Fire Ant) mounds in 3 counties (Hinds, Leake, and Madison) within the Natchez Trace Parkway, MS. We assayed mound soil, plant debris within the mounds, and ants obtained from mounds on 3 sampling dates (March, July, and November). We processed samples based on standard microbiological protocols, and used traditional morphological and molecular techniques to identify fungal taxa. We documented a total of 1445 isolates consisting of 50 fungal taxa and calculated a diversity index value (H') of 3.11 across all substrates, which was indicative of a variable fungal community within the mounds. The taxa with the highest percent isolation frequencies included Hypocrea lixii (12.8%), Fusarium sp. 1 (12.3%), Fusarium equiseti (7.9%), Purpureocillium lilacinum (= Paecilomyces lilacinus) (6.5%), Fusarium oxysporum 2 (5.8%), and Mortierella alpina (5.4%). We isolated 2 common parasitic (entomopathogenic) fungi, Purpureocillium lilacinum and Metarhizium anisopliae var. anisopliae (9.4%), from mound soil, plant debris, and ant external tissues. Hypocrea lixii, the teleomorphic reproductive stage of Trichoderma harzianum, is noted as a natural biological control of some soil-borne microbes, possibly limiting important natural entomopathogenic activity within the mounds. Species richness and diversity values from mound soils across locations were significantly greater (P ≤ 0.05) than those from the plant debris and ant body-tissue substrates. Species richness values between locations were similar. Species richness of samples collected in November (47) was significantly greater (P ≤ 0.05) than that of the March (41) and July (39) samples. Community coefficient values ranged from 0.79 to 0.87 between substrates, 0.85 to 0.91 between locations, and 0.85 to 0.86 between sampling dates, indicating that taxa were similar. Introduction Introduction and movement of imported fire ants (IFA) from South America, Solenopsis richteri Forel (Black Imported Fire Ant (BIFA) and Solenopsis invicta Buren (Red Imported Fire Ant (RIFA), to the US has been previously summarized (Lard et al. 2002, Tschinkel 2005). The ants’ deleterious impacts affect humans, livestock, crops, native fauna, invertebrates, and even machinery and electrical equipment (Lofgren 1986, Vander Meer et al. 1990, Vanderwoude et al. 2000). Invasion by these species can cause elimination or displacement of other exotic and 1BCH-EPP Department, Box 9655, Mississippi State University, Mississippi State, MS 39762. 2Current address - Valent BioSciences Osage, 214 350th Street, 603 N 3rd Street, PO. Box 147, Osage, IA 50461. 3University of Arkansas Division of Agriculture, 2404 North University Avenue, Little Rock, AR 72207. *Corresponding author - rbaird@plantpath.msstate.edu. Manuscript Editor: Scott Markwith Southeastern Naturalist 221 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 native ant species (Streett et al. 2006). BIFA and RIFA were first introduced into the US around 1918 and 1930, respectively, through or near the port in Mobile, AL (Buren et al 1974, Vinson and Sorensen 1986). The impacts and spread of RIFA out of Alabama have been dramatic, and their rapid spread in the US was due to a lack of the natural enemies and competitors that limit population growth in the species’ native range (Whitcomb 1980). Previous research has shown that generalist and entomopathogenic fungi occur in IFA mounds, but none of these studies was conducted at newly established locations (~6 months old), and they were primarily limited to generic-level names from either RIFA or BIFA ant mounds (Baird et al. 2007, Beckham et al. 1982, Jouvenaz et al. 1977, Zettler et al. 2002), leaving identification of generalist and entomopathogenic fungi unclear. Zettler et al. (2002) reported that RIFA mounds had greater fungal abundance than non-mound soils—over 19 times more colony-forming units—but with lower species richness and diversity. In that study, 2 fungal species made up ~75% of colonies isolated, but it was unclear if the fungal-population differences in mounds were due to ant mediation. In South America, fungi are reported to control RIFA, limiting the ants’ spread and destructiveness. Potential fungal biological-control agents include Beauveria bassiana (Bals.-Criv.) Vuill. and Metarhizium anisopliae (Metschn.) Sorokin (Meyling and Eilenberg 2007). A mortality rate of 90% was observed when BIFA was exposed to B. bassiana (Broome 1974). Stimac et al. (1993) conducted an investigation on the effects of a Brazilian strain of this fungus on IFA colonies and discovered that B. bassiana provided some control of the treated colonies. Beauveria bassiana has been formulated as baits and tested against IFA (Barr and Drees 2003, Barr et al. 2003, Patterson et al. 1993, Williams et al. 2003), resulting in development of commercial biopesticides by several companies, including Safe- Science (Boston, MA) and Troy Biosciences (Phoenix, AZ) (Williams et al. 2003). In tests with M. anisopliae, 100% mortality of 15 IFA queens occurred after 5 d (Sanches-Peña 1992). In a 4-y study in Argentina, the microsporidium Kneallhazia solanapsae Knell, Allen, and Hazard, an intracellular obligate parasite, reduced BIFA colonies from 162 to 28 (Briano et al. 1995), and this microsporidian resulted in the death of colonies of polygyne RIFA colonies in Florida over a 2-y period (Oi and Williams 2002). The endoparasitic fungus Myrmecomyces annellisae was described from fire-ant species in the US and Argentina (Jouvenaz and Kimbrough 1991). It is not known whether any of these entomopathogenic or endoparastic fungi with the potential for biological control are native to the southern US. Current economic and health concerns due to spread of RIFA into the southern US (Vanderwoude et al. 2000) increase the importance of identifying native populations of generalist or host-specific entomopathogenic fungi in this region. This work would also provide baseline information leading to possible control strategies. Dispersal of IFA species in Mississippi, especially RIFA, was very rapid, as demonstrated by field-survey monitoring of the ants’ spread into Mississippi and other southern states. The documented distribution of IFA in Mississippi includes RIFA in the southern half of the state and extending northward in western counties, Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 222 BIFA in northern counties bordering on Tennessee through north central counties, and BIFA x RIFA hybrid imported fire ants (HIFA) in a band between the 2 species (MacGown 2014, Streett 2006, Vander Meer and Lofgren 1985). By November 2003, it was first reported that portions of Mississippi were being colonized by BIFA and RIFA along the Natchez Trace Parkway (J.T Vogt , USDA/ARS, Stoneville, MS; unpubl. data). The Natchez Trace Parkway corridor is an existing north–south transect through Mississippi where fire-ant studies could be conducted with a goal of obtaining critical field-data on IFA and native ant species; therefore, a multistate-USDA/ARS project was established along the parkway. As part of this larger project, our objective was to conduct a survey of entomopathogenic microbes of IFA mounds along Natchez Trace Parkway. RIFA’s demonstrated dispersal and colonization potential suggests the importance of determining associated natural soil-microbial communities of this pest to gather baseline community-data that might provide insight into the interactions of RIFA and potential biological control agents in mounds. Materials and Methods In 2004, we sampled RIFA mounds in 3 counties along the Natchez Trace Parkway in Mississippi: Hinds County (mile markers 83–87), Madison County (mile markers 102–122), and Leake County (mile markers 129–138). In each county, we collected soil from 5 randomly selected RIFA mounds during each of 3 sampling dates in March, July, and November (Woolfolk et al. 2016). For each collection, we placed in a plastic bag 2 L of soil and other debris collected from the lower third on the north side of a RIFA mound. We obtained 500-g subsamples from the bags and stored them at 4 ºC until processed for microbial isolations. In addition, to confirm RIFA identities in each mound where we collected soil samples, we preserved 20 ants per mound (Triplehorn and Johnson 2004) for identification based on morphological characters and chemical analyses (Baird et al. 2007). We followed methods described in Baird et al. (2007) and Woolfolk et al. (2016) for substrate preparation for fungal isolations with the exceptions noted below. We used soil, plant debris within the soil samples, and internal and external ant tissues as our substrates. We plated all substrates onto Sabouraud’s dextrose agar yeast (SDAY; Difco®) medium for fungal isolations (Baird et al. 2007, Goettel and Inglis 1997) amended with 300 mg/L streptomycin sulfate (Sigma Corporation, Houston, TX) and 100 mg/L chlortetracycline (Sigma Corporation) to inhibit bacteria. Fungal isolations from ant bodies (external tissues) To obtain fungal isolates from external (cuticular surface) ant tissues, we took external swabs from 4 worker ants from each of the 5 replicate mounds from each collection date and county and plated the collected material onto SDAY medium, i.e., 5 mounds per county x 3 counties x 4 replicate ant samples per mound x 3 dates = 180 ant samples. We cultured all plates at 25 °C for a minimum of 96 h, after which we subcultured the isolates and stored the samples at -80 °C in 15% glycerol solution for identification. Southeastern Naturalist 223 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 Fungal isolations from ant bodies (internal tissues) We followed the same sampling pattern as described above for the external tissues. For the internal ant-tissue assay, we used the same 4 ants per mound from which we had made swabs for external tissue samples. We surface-sterilized the ants in 10% ETOH for 10 sec and placed each ant into an individual sterile microcentrifuge tube containing chilled buffer-Tween (50 mM phosphate buffer containing 0.01% Tween-80). We used a micropestle to grind and homogenize the samples. We diluted homogenates using a tenfold dilution series and spread 100- μl aliquots evenly onto 4 replicate SDAY-medium plates. We employed the same methods for isolation of fungi as those used for the external tissues. We subcultured fungal isolates with the same macroscopic and microscopic morphological features for identification and placed them into morphological groupings for further processing. We made microscopic confirmations of the fungal isolates using standard mycological characters given various references for anamorphic-forming fungi (Barnett and Hunter 1998, Barron 1968, Domsch et al. 1980, Ellis 1971) or by their teleomorphic states. We stored a minimum of 2–5 representative isolates per group in 15% glycerol at -80 °C for permanent culture collection and further testing. Identifications To further confirm our identifications, we grew 4 randomly selected fungi per morphological grouping on SDAY plates for a minimum of 7 d at room temperature, subcultured them onto the general growth-medium potato dextrose broth (PDB, Difco®), and grew them for 14 d at room temperature. We extracted and amplified genomic DNA using sequence primers ITS 1 and ITS 4 (Baird et al. 2014). Using the blastn program of the Basic Local Alignment Search Tool (BLAST), we compared all ITS sequence data to the GenBank database (National Center for Biotechnology Information, NCBI) to determine identities. We considered all fungal ITS-sequence data having matches with 80% coverage or higher, and a minimum of 97% homology to be the same species (Hughes et al. 2009). We confirmed Fusarium spp. isolates to genus using microscopic characterizations of asexual reproductive structures and spores. To further identify the isolates to species, we sequenced all isolates using species-specific primers for gene regions encoding b-tubulin (T1: forward primer, and T2: reverse primer; O’Donnell and Cigelnik 1997) and a-elongation factor (E1: forward primer, and E2: reverse primer; O’Donnell et al. 1998). We BLAST-queried all isolate sequences using the FUSARIUM-ID library at Penn State University (http://isolate.fusariumdb.org). Statistical analyses Using established methods, we based the biodiversity indices of fungi from RIFA and their mounds on the isolation frequencies of the fungi (Baird et al. 2007, Inglis and Cohen 2004, Woolfolk and Inglis 2004). We statistically analyzed the following biodiversity indices: species or taxon richness (n), Shannon-Weaver species diversity index (H'), coefficient of community (CC, i.e., Sørensen coefficient), and species or taxon evenness (E) (Price 1997, Stephenson 1989, Stephenson et al. Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 224 2004). We also calculated relative frequencies of fungal occurrence. We subjected relative frequency and biodiversity index values to one-way analysis of variance (ANOVA) using the general linear models procedure (Proc GLM) of Statistical Analyses System software (SAS Institute 1999) to detect differences among isolates, substrates, locations, and sampling dates, as appropriate. We compared means with Tukey’s HSD test to allow for multiple comparisons among treatment means. We set P < 0.05 to indicate significance for all tests. Results We identified a total of 50 fungal taxa from 1445 isolates from mound soil, ant tissues, and plant debris within the RIFA mounds (Table 1). Fusarium was the most commonly identified genus; our samples included 7 distinct species of that Table 1. Mean percent isolation frequency of fungal taxa identified from Red Imported Fire Ant mounds from 3 locations (Hinds, Madison, and Leake counties) along Natchez Trace Parkway, MS. NCBI = National Center of Biotechnology Institute accession number (based on the closest match of fungal taxa in GenBank database, n/a = not applicable); Ext. = external ant tissue (cuticular surface, external body regions); Int. = internal ant tissue (tissue from internal body regions); Soil = ant-mound soil, Plant = mound plant debris, and Total = overall total %. [Table continued on following page.] Total % by substrateA Fungal taxa NCBI Ext. Int. Soil Plant Total Aspergillus flavipes (Bainier & R. Sartory) Thom & KF624764 7.2 1.7 5.0 6.1 5.0 Church strain UWFP 1022 Aspergillus flavus Link KF624765 3.9 5.0 10.0 13.9 less than 1.0 Aspergillus niger Tiegh KF624766 0.0 0.0 2.2 10.0 3.1 Aspergillus nomius Kurtzman, B.W. Horn & Hesselt KF624767 0.0 0.6 0.0 0.0 less than 1.0 Aspergillus nomius Kurtzman, B.W. Horn & Hesselt KF624768 0.0 0.0 5.0 0.0 less than 1.0 Aspergillus parasiticus Speare KF624769 1.1 0.0 4.4 2.8 2.1 Aspergillus terreus Thom KF624770 1.1 0.0 2.8 1.1 1.3 Aspergillus terreus Thom KF624771 1.1 0.0 6.7 1.7 2.4 Aspergillus tubingensis Mosseray strain 3.4342 KF624772 0.0 1.7 8.9 4.4 3.8 Aspergillus versicolor (Vuill.) Tirab. KF624773 0.0 0.0 1.7 0.0 less than 1.0 Bionectria ochroleuca (Schwein.) Schroers & Samuels KF624774 1.1 0.0 3.3 6.1 2.6 Ceratocystis sp. KF624775 1.1 0.0 8.9 2.2 3.1 Cochiobolus kusanoi (Y. Nisik.) Drechsler ex Dastur KF624776 7.2 1.1 10.6 0.0 4.7 Curvularia sp. Boedijn KF624777 0.6 2.2 3.9 3.9 2.6 Fusarium acuminatum Ellis & Everh. KF624784 0.0 0.0 0.6 0.6 less than 1.0 Fusarium equiseti 1 (Corda) Sacc. KF624787 0.0 0.0 3.3 1.7 1.3 Fusarium equiseti 2 KF624789 3.3 2.2 13.9 12.2 7.9 Fusarium oxysporum f. sp. phaseoli J.B. Kendr. & KF624779 0.0 0.0 1.1 0.6 less than 1.0 W.C. Snyder Fusarium oxysporum 1 E.F. Sm. & Swingle KF624780 0.0 0.0 1.7 1.1 less than 1.0 Fusarium oxysporum 2 KF624781 5.6 0.0 15.6 3.3 5.8 Fusarium oxysporum 3 KF624782 3.3 1.1 8.9 5.0 4.6 Fusarium oxysporum 4 KF624783 2.2 0.0 2.8 0.6 1.4 Fusarium solani (Mart.) Sacc. KF624788 0.0 1.1 1.1 2.2 1.1 Fusarium sporotrichioides Sherb. KF624790 0.0 0.0 0.6 7.8 2.1 Fusarium sp. 1 Link KF624785 3.3 0.0 22.2 9.4 12.3 Fusarium sp. 2 KF624786 5.6 0.0 16.7 25.0 8.2 Southeastern Naturalist 225 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 genus and 5 subspecies or form species of F. oxysporum. Overall, we isolated 14 different taxa of Fusarium spp., with 5 isolates identified as Gibberella spp., the sexual reproductive stage. The 6 fungi with the greatest percent isolation frequencies included Hypocrea lixii (12.8%), Fusarium sp. 1 (12.3%), Fusarium equiseti (7.9%), Purpureocillium lilacinum (6.5%), Fusarium oxysporum 2 (5.8%), and Mortierella alpina (5.4%). Purpureocillium lilacinum (known previously as Paeciliomyces lilacinus Thom) and Metarhizium anisopliae var. anisopliae (2.9%), both entomopathogenic fungi, occurred on the ants’ cuticular surface, mound soil, and plant-debris substrates. We isolated Metarhizium taii (less than 1.0%)—considered to be an entomopathogen—from mound soil. We isolated 15 fungal taxa at frequencies of less than 1.0%, including unknown species, and isolation frequencies of all other fungal taxa ranged from 1.0% to 5.6%. Table 1, continued. Total % by substrateA Fungal taxa NCBI Ext. Int. Soil Plant Total Gibberella moniliformis Wineland KF624791 0.0 0.0 1.1 1.1 less than 1.0 Gibberella zeae (Schwein.) Petch KF624778 0.0 0.0 0.6 0.6 less than 1.0 Hypocrea lixii Pat. KF624792 4.4 3.9 23.9 18.9 12.8 Lecythophora sp. Nannf. KF624793 0.0 0.0 5.6 0.0 less than 1.0 Lophiostoma sp. Ces. & De Not. KF624794 0.0 0.0 0.0 0.0 1.4 Metarhizium anisopliae var. anisopliae (Metschn.) KF624795 1.7 0.0 5.0 5.0 2.9 Sorokīn Metarhizium taii Z.Q. Liang & A.Y. Liu KF624796 0.0 0.0 0.6 0.0 less than 1.0 Microsphaeropsis arundinis (S. Ahmad) B. Sutton KF624797 1.1 0.0 0.6 2.2 1.0 Mortierella alpina Peyronel KF624798 9.4 0.0 11.7 0.6 5.4 Neosartorya fischeri (Wehmer) Malloch & Cain KF624799 0.0 0.6 2.8 0.0 less than 1.0 Penicillium citrinum Link KF624801 0.6 0.0 9.4 2.8 3.2 Penicillium cairnsense Houbraken, Frisvad & Samson KF624802 0.6 0.0 1.1 0.6 less than 1.0 Penicillium euglaucum J.F.H. Beyma KF624803 0.0 0.0 4.4 0.0 1.1 Penicillium granulatum Bainier KF624804 0.0 0.0 2.2 0.0 less than 1.0 Penicillium pulvillorum Turfitt strain KF624805 1.7 0.0 7.2 0.6 2.4 Penicillium spinulosum Thom KF624806 3.4 1.1 2.3 6.1 2.9 Pseudallescheria boydii (Shear) McGinnis, A.A. Padhye KF624807 0.6 0.0 13.3 2.8 4.3 & Ajello Purpureocillium lilacinum Thom) Luangsaard, KF624800 1.7 0.0 12.2 12.2 6.5 Houbraken, Hywel-Jones, & Samson Trichoderma harzianum Rifai KF624808 0.6 0.0 1.7 2.8 1.3 Trichoderma spirale Bissett KF624809 1.1 0.0 15.0 2.2 4.6 Zygomycete sp. KF624810 3.3 0.0 5.0 13.3 5.4 Unknown n/a 1.1 0.0 1.7 0.6 less than 1.0 LSD (0.05) 4.5 2.9 9.4 6.3 3.3 AMean percent isolation from soil mounds and plant debris is based on the percent occurrences of fungal species isolated from 3 sampling dates (March, July, and November 2004), 3 locations (Hinds, Madison, and Leake counties)/sampling date, 5 active mounds/location/sampling date, 4 replicates/ mound/location/sampling date: Mean percent ÷ 180 (= 5 mounds x 3 locations x 3 sampling dates x 4 replicates) x 100. Overall mean total percentages of total fungal isolates = (total mean percent isolation from all substrates ÷ 4) x 100. For ant body tissues, the formula includes external and internal tissue samples: Mean percent ÷ 180 (= 5 mounds x 3 locations x 3 sampling dates x 4 replicates) x 100. Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 226 The absolute frequencies of substrate samples (180 of each type) with no fungi isolated from ant interior, ant exterior, mound soil, and plant debris were 72, 51, 18, and 79, respectively. Substrate samples without fungi from Hinds and Madison counties across all tissue types, 92 and 84, respectively, were higher than Leake County at 44. When we statistically analyzed insect tissues separately by internal and external tissues, we detected no significant differences for fungi isolated between location, substrate, date, or combinations. Therefore, we pooled the data for all analyses to assess species richness, diversity, evenness, and coefficent of community determinations. Analyses by substrate, location, and sampling date showed that the highest species-richness values for fungi were from mound soil (50), Hinds County (46) and samples collected during November (47) (Table 2). When we compared sampling dates, species-richness values were significantly greater in November (47) than in March (41) and July (39). We observed no significant differences between locations and individual substrate types between locations (Table 2). However, we noted different trends in species richness when we compared substrate types within a given location. For example, Hinds County richness values were significantly different between all 3 substrates: mound soil (39 species), plant debris (27), and ant-body (19) (Table 2). We observed the same trend for the Madison County: species richness for mound soil (37) was significantly greater than plant debris (30) and pooled ant-body isolates (27). For each location, species richness in mound-soil samples was significantly or at least numerically greater compared to plant debris and ant-body-isolate samples. Overall fungal-species diversity was 3.11 (H'). A value of 0 represents a fungal community with a single species; higher numbers represent communities with a greater diversity of species. When we compared diversity values for substrate data for all locations, significantly greater diversity levels were present in mound soils than the other 2 substrates between locations (Table 3). However, when analyzed Table 2. Species richness (n) of all fungal taxa isolated from Red Imported Fire Ants and mounds along Natchez Trace Parkway, MS in 2004. Within-column values with the same lowercase letter are not significantly different (P ≤ 0.05). Across-row mean values with the same uppercase letter are not significantly different. Means were compared with Tukey’s HSD test. Substrate n Location n Sampling date n Soil mound 50a Hinds 46a March 41ab Plant debris 41b Leake 42a July 39b Ant tissue 36c Madison 45a November 47a LSD (P ≤ 0.05) 5 6 6 Location Soil mound Plant debris Ant tissue LSD (P ≤ 0.05) Hinds 39a(A) 27a(B) 19a(C) (9) Leake 35a(A) 30a(AB) 22a(B) (9) Madison 37a(A) 30a(B) 27a(B) (6) LSD (P ≤ 0.05) 9B 9 10 Southeastern Naturalist 227 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 separately for each location, this pattern did not hold for Leake County, where the diversity found for plant debris, while lower than for soil mound, was not significantly less. For all counties, the diversity in ant-tissue substrate was significantly lower than in soil-mound substrate, though not significantly lower than in plant debris for Leake and Madison counties. However, we observed no significant trends when we tested for interactions between locations and each substrate type. When we compared diversity values for each substrate by location, values for mound soils were significantly greater than values for plant debris and ant bodies except for Leake County, where diversity in soil mound and plant debris were not significantly different. An evenness value of 0 indicates that 1 species dominated a location or substrate, and a value of 1.0 signifies that all species had similar population levels. Evenness values were similar among all substrates, with the highest values from mound soil (0.88) (Table 4). Madison County evenness (0.90) was significantly greater than Hinds but similar to Leake (0.88). By sampling date, March had the highest value (0.88) and was significantly greater than November (0.84), but similar to July (0.87). The actual differences were slight, and all parameters compared had high relative abundance. In general, the values varied by only 0.04 between the dates. Our results indicated that there were no consistent trends in the mean number of occurrences of fungal taxa based on interactions between substrate and location. Evenness between location and substrate type was similar among all comparisons. A coefficient of community (CC) value of 0 indicates no shared species among sites or substrates and a value of 1.0 means that all species are shared among sites or substrates. The overall CC value for fungi was 0.86, which indicated that most fungal taxa observed in the study were common among the different samples. When we compared substrates, values ranged from 0.79 to 0.92; the highest from mound soil–plant debris at 0.87 (Table 5). Among locations, CC values were highest for Hinds–Madison and Hinds–Leake comparisons, 0.92 and 0.91, respectively, and values for all sampling dates were similar. Table 3. Species diversity (H') of all fungal taxa isolated from Red Imported Fire Ants and mounds along Natchez Trace Parkway, MS in 2004. Within-column values with the same lowercase letter are not significantly different (P ≤ 0.05). Across-row mean values with the same uppercase letter are not significantly different. Means were compared with Tukey’s HSD test. Substrate H' Location H' Sampling date H' Soil mound 3.43a Hinds 3.31ab March 3.26a Plant debris 3.13b Leake 3.27b July 3.20a Ant tissue 3.10b Madison 3.41a November 3.25a LSD (P ≤ 0.05) 0.27 0.10 0.09 Location Soil mound Plant debris Ant tissue LSD (P ≤ 0.05) Hinds 3.14a(A) 2.82a(B) 2.55a(C) (0.30) Leake 3.09a(A) 2.97a(AB) 2.69a(B) (0.48) Madison 3.28a(A) 3.00a(B) 2.88a(B) (0.27) LSD (P ≤ 0.05) 0.25 0.28 0.49 Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 228 CC values for fungi were numerically higher for the soil mound–plant debris comparison (0.87) than for the soil mound–ant tissue or plant debris–ant tissue comparisons (0.79 and 0.81, respectively). At all 3 sites, mound soil–plant debris had the highest CC values—0.77, 0.75, and 0.64, from Leake, Madison, and Hinds, respectively (Table 5). Overall, the CC values by location and substrate interactions were moderate to low: Hinds plant debris–pooled ant-tissue isolate CC = 0.48, indicating approximately 50% of fungal taxa were present in both substrate tissues; for soil mound and ant body tissues the CC = 0.55; and soil mound and plant debris comparisons were only somewhat higher, with a CC = 0.64). Table 5. Coefficient of community (CC) of all fungal taxa isolated from Red Imported Fire Ants and mounds along Natchez Trace Parkway, MS. Substrates CC Locations CC Sampling dates CC Soil mound–plant debris 0.87 Hinds–Leake 0.91 March–July 0.85 Soil mound–ant tissue 0.79 Hinds–Madison 0.92 March–November 0.86 Plant debris–ant tissue 0.81 Leake–Madison 0.85 July–November 0.86 Location Substrates CC Hinds Soil mound–plant debris 0.64 Hinds Soil mound–ant tissue 0.55 Hinds Plant debris–ant tissue 0.48 Leake Soil mound–plant debris 0.77 Leake Soil mound–ant tissue 0.67 Leake Plant debris–ant tissue 0.62 Madison Soil mound–plant debris 0.75 Madison Soil mound–ant tissue 0.75 Madison Plant debris–ant tissue 0.67 Table 4. Species evenness (E) of all fungal taxa isolated from Red Imported Fire Ants and mounds along Natchez Trace Parkway, MS in 2004. Within-column values with the same lowercase letter are not significantly different. Within-column values with the same letter are not significantly different (P ≤ 0.05). Across-row mean values with the same uppercase letter are not significantly different. Means were compared with Tukey’s HSD test. Substrate E Location E Sampling date E Soil mound 0.88a Hinds 0.86b March 0.88a Plant debris 0.84a Leake 0.88ab July 0.87ab Ant tissue 0.86a Madison 0.90a November 0.84b LSD (P ≤ 0.05) 0.07 0.04 0.02 Location Soil mound Plant debris Ant tissue LSD (P ≤ 0.05) Hinds 0.86b(A) 0.86a(A 0.87a(A) (0.04) Leake 0.87ab(A) 0.87a(A) 0.87a(A) (0.07) Madison 0.91a(A) 0.88a(A) 0.87a(A) (0.04) LSD (P ≤ 0.05) 0.05 0.04 0.09 Southeastern Naturalist 229 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 Discussion The primary purpose of the current study was to determine the baseline fungal community in RIFA mounds with particular emphasis on potential biological control agents. Of the total 50 taxa of fungi we identified from subsamples of soil-mound substrates, we isolated 6 common fungal species at a significantly greater level than we observed in plant debris or ant-tissue samples. Hypocrea lixii (anamorph: T. harzianum), the most commonly isolated species, occurred on all substrates but was absent from internal ant tissues. We believe that H. lixii is a common soil inhabitant; this and other anamorphic species of Trichoderma are known associates of soil, roots, and above-ground parts of plants (Baird et al. 2007). Harmon (2000) reported that this fungal species is an important cellulose-degrading saprophyte (Harmon 2000), and Zhao et al. (2013) found that H. lixii survived as an endophyte of different plant species including Cajanus cajan (L.) Millsp. (Pigeon Pea). The anamorphic stage reported in the literature has been widely reported as a biological control for soil-borne generalist or pathogenic fungi of agricultural ecosystems (Chaverri and Samuels 2003). It is possible that RIFA directly or indirectly maintain a population of H. lixii as a protective mechanism to prevent entomopathogenic fungi such as Metarhizium or Beauveria spp. from establishing and impacting a colony. However, H. lixii in RIFA mounds is probably acting more as a generalist that consumes carbon sources and indirectly affects entomopathogenic microbe survival and growth; this relationship has not been tested. Preliminary in vivo studies using isolates from the sampling described herein showed that T. harzianum were antagonistic, and inhibited the RIFA from reaching a food source until the ants formed soil or plant-debris bridges to bypass the fungal colony (S. Woolfolk, unpubl. data). The other fungal species isolated during our study could be classified as generalists, and some are reported to have entomopathogenic potential. The most common genus of fungi isolated was Fusarium. The 14 species identified during this study have been documented from many habitats but in particular are pathogens of agricultural ecosystems (Booth 1971). Some Fusarium species can survive saprophytically on plant debris in soil or may be parasites of many plant species with no known impacts or benefits to RIFA in mounds. Their direct involvement as insect biocontrol agents has not been reported. Paecilomyces lilacinum, which had the 3rd-highest isolation frequency in this study, can survive as a saprophyte, entomopathogen, and/or is nematophagous (Atkins et al. 2005). This species, which has been widely tested for control of nematode species (Gunasekera et al. 2000), closely aligned phylogenetically to Trichoderma (sexual stage = Hypocrea) and Gliocladium, which have been tested and formulated for biocontrol of fungi (Inglis and Tigano-Milani 2006). In preliminary studies in our laboratory, artificial RIFA colonies died within 48 h when we introduced P. lilacinum into the in vivo established colonies (S. Woolfolk, unpubl. data). Two previous studies were conducted to determine fungal-species richness and abundance in RIFA and BIFA/RIFA imported ants from Mississippi (Baird et al. 2007) and RIFA microbial baseline data from South Carolina (Zettler et al. 2002). Some of the taxa observed in the current RIFA investigation were isolated Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 230 previously in those studies. In particular, Fusarium spp., which are common in nonant- infested soils, occurred in abundance—11 species were identified in the South Carolina study (Zettler et al. 2002) and 10 in Mississippi ant-mound soils (Baird et al. 2007). We observed the common soil-borne fungus F. oxysporum across all sampling dates, but some were strains or form species known to be pathogens of economically important agricultural and horticultural plants (Booth 1971). Many other Fusarium spp. appeared to be the same as those identified during past experiments, but even with sequence data from the current study, we were unable to identify several of them beyond the generic level. Other fungi we observed in the RIFA mounds included 5 Trichoderma spp. that were also observed in South Carolina (Baird et al. 2007) and 2 that we recorded in the Mississippi study (Zettler et al. 2002). None of the isolates from those past investigations were ever verified as T. harzianum (H. lixii). Baird et al. (2007) isolated the entomopathogenic species B. bassiana at 6.7% frequency from BIFA/RIFA hybrid-ant bodies, mound soils, and mound-soil plant-debris, but we did not detect it during the current study nor did Zettler et al. (2002) from RIFA mounds in South Carolina. These results indicate that B. bassiana might be associated with BIFA/RIFA hybrid-ant habitats or geographical locations, rather than RIFA; further studies are necessary to confirm that possibility. Another reason for variations in fungal-population levels is that RIFA produces an exudate containing alkaloids shown to reduce condia germination of Metarhizium or Beauveria spp. and P. lilacinum, though hyphal growth was unaffected (Beattie et al. 1985, Storey et al. 1991). The substance might prevent new colonies of these pathogenic fungi from overtaking the mounds, thus limiting their ability to follow the movement of the ants to new locations. Furthermore, Zettler et al. (2002) indicated that fungal diversity in RIFA mounds might be affected by various environmental conditions such as temperature, precipitation, pH, or other factors. In a study of ant-nest soil properties at colonies of 9 non-RIFA ant species, pH tended to shift towards neutral, plant-debris accumulations were high, and tissue decomposition resulted in greater nutrient (nitrogen and phosphorus) availability in the mounds compared to soils where ants were absent (Frouz and Jilkova 2008). These researchers also showed that increased soil porosity caused by ant formation of corridors or galleries directly impacted temperatures and available moisture, which both affected microbial activity. Total species richness was higher in samples from mound soil than those from plant debris and pooled ant tissues. These values were highest in November at the end of the growing season—a time of year when fungal populations are most diverse if environmental conditions were conducive for growth. This pattern was also observed in a study by Baird et al. (2007) of BIFA/RIFA hybrid mounds. Samples from mound soils in the 2007 study had much greater levels of species richness of fungal taxa across sampling dates than those detected from ant-body substrates. A thorough review of the literature and reasons for these variations in species richness from different substrates is discussed in Zettler et al. (2002) and Baird et al. (2007). Southeastern Naturalist 231 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 Species-richness interactions between sampling location and substrate were similar among the 3 locations, but we observed differences between substrates by location. However, we noted no environmental differences between those sites. Similar to species richness, abundance (total isolations) was also significantly greater from soil mounds than from the other 2 substrates, with no apparent trends by sampling date; Baird et al. (2007) obtained a similar result. Zettler et al. (2002) double-compared species richness, evenness, and diversity values for the substrates to non-mound soils and found that RIFA may regulate microbial occurrences in mounds or select fungi being utilized by RIFA, but could not explain the phenomenon. No other trends were noted based on isolation data from the earlier RIFA investigations. Species richness was highest in November whereas diversity was similar across sampling dates. In this study, the consistent fungal diversity across months may be an indication of RIFA involvement in the mound. Coefficient of community values were similar in all comparisons of sampling dates, locations, and substrates. Baird et al (2007) noted that CC values were different between the first and last sampling dates, but that study extended only from October 2002 through January 2003—a period when temperatures are generally low and microbial activity is reduced. Although the July average temperature was 10 °C warmer than averages on the other 2 sampling dates, neither temperature nor rainfall showed any relationship with community structure. Average monthly rainfall from January to June was 160 cm (300-cm maximum in June), and the average monthly rainfall from July through December was 36 cm (120-m as the maximum in October) (Western Regional Climate Center 2015). As stated previously, the consistent presence and purposes of the fungal communities in RIFA mounds remain uncertain. Taxa observed in this and past studies varied, with greater differences in fungal taxa among substrates in the current investigation. Differing from past research, we used molecular sequence data to support the macroscopic and microscopic identifications for all fungi. We identified several potential entomopathogenic fungi from the different mound substrates, but their role in that habitat remains unclear. Additional research should be conducted on the insect pathogenicity potential of the associated entomopathogenic fungi cultured in this study and examining the direct or indirect impact H. lixii (T. harzianum) has on potential survival, colonization, and disease levels in RIFA mounds. Acknowledgments The authors are grateful for a USDA-ARS Specific Cooperative Agreement that provided the largest portion of financial support for this study under Project Numbers 6402-22320- 00300D, Sigma Xi for its Grants-In-Aid of Research in 2004, and to the National Park Service for providing a permit to collect samples from Natchez Trace Parkway area in Mississippi. We thank Jian Chen (Biological Control of Pests Research Unit, USDA-ARS, Stoneville, MS) for identification efforts. We acknowledge Mississippi State University (MAFES publication number 12459) for providing field and laboratory research facilities and supplies. The research presented in this paper represents a portion of S. Woolfolk’s Ph.D. from Mississippi State University. The first- and last-listed authors contributed equally to this paper. Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 232 Literature Cited Atkins, S.D., I.M. Clark, S. Pande, P.R. Hirsch, and B.R. Kerry. 2005. The use of real-time PCR and species-specific primers for the identification and monitoring of Paecilomyces lilacinus. FEMS Microbiology Ecology 51:257–264. Baird, R.E., S. Woolfolk, and C.E. Watson. 2007. Survey of bacterial and fungal associates of black/hybrid imported fire ants from mounds in Mississippi. Southeastern Naturalist 6:615–632. Baird, R., A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson. 2014. Rhododendron decline in Great Smoky Mountains and surrounding areas: Intensive site study of biotic and abiotic parameters associated with the decline. Eastern Biologist. 13:1–29. Barnett, H.L., and B.B. Hunter. 1998. Illustrated Genera of Imperfect Fungi. 4th Edition. APS Press, St. Paul, MN. 218 pp. Barr, C.L., and B.M. Drees. 2003. Acceptance and efficacy of Beauveria bassiana formulations to laboratory colonies of the Red Imported Fire Ant (Hymenoptera: Formicidae). Available online at http://fireanttamu.edu/researchlarr/category/non-chemica1l90– 91Pg44/90-91Pg44.pdf. Accessed January 2004. Barr, C.L., R.L. Best, L. Lennon, D. Mott, and S. Russell. 2003. Final report on testing of TB 131-1, a Beauveria bassiana product manufactured by Troy Biosciences for the control of individual colonies of Red Imported Fire Ants. Available online at http:// fireanttamu.edu/research/arr/category/individual/97-01Pg13/97-01Pg13.pdf. Accessed January 2004. Barron, G.L. 1968. Genera of Hyphomycetes from Soil. Robert E. Krieger Publishing Company, Malabar, FL. 364 pp. Beattie, A.J., C.L. Turnbull, T. Hough, and R. Bruce Knox. 1985. The vulnerability of pollen and fungal spores to ant secretions: Evidence and some evolutionary implications. American Journal of Botany 72:606–614. Beckham, R.D., S.L. Bilimoria, and D.P. Bartell. 1982. A survey of microorganisms associated with ants in western Texas. Southwestern Entomologist 7:225–229. Booth, C. 1971. The Genus Fusarium. Commonwealth Mycological Institute, Kew, Surrey, UK. 237 pp. Briano, J., D. Jouvenaz, D. Wojcik, H. Cordo, and R. Patterson. 1995. Protozoan and fungal diseases in Solenopsis richteri and S. quinquecuspis (Hymenoptera: Formicidae) in Bueno Aires Province, Argentina. Florida Entomologist 78:531–536. Broome, J.R 1974. Microbial control of the imported fire ant Solenopsis richteri. Ph.D. Dissertation. Mississippi State University, Mississippi State, MS. Buren, W.F., G.E. Allen, F.E. Whitcomb, F.E. Lennartz, and R.N. Williams. 1974. Zoogeography of the imported fire ants. Journal of the New York Entomological Society 82:113–124. Chaverri, P., and G.J. Samuels. 2003. Hypocrea/Trichoderma (Ascomycota, Hypocreales, Hypocreaceae): Species with green ascospores. Studies in Mycology 48:1–119. Domsch, K.H., W. Gams, and T.H. Anderson. 1980. Compendium of Soil Fungi (Vol.1). Academic Press, London, UK. 859 pp. Ellis, M.B. 1971. Dematiaceous Hyphomycetes. Commonwealth Mycological Institute, Kew, UK. 608 pp. Frouz, J., and V. Jilkova. 2008. The effect of ants on soil properties and processes (Hymenoptera: Formicidae). Myrmecological News 11:191–199. Southeastern Naturalist 233 S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 Goettel, M.S., and G.D. Inglis. 1997. Fungi: Hyphomycetes. Pp. 213–249, In L.A. Lacey (Ed.). Manual of Techniques in Insect Pathology. Academic Press, Inc., San Diego, CA. 409 pp. Gunasekera, T.S., R.J. Holland, M.R. Gillings, D.A. Briscoe, D.C. Neethling, K.L. Williams, and K.M.H. Nevalainen. 2000 Phenotypic and genetic characterization of Paecilomyces lilacinus strains with biocontrol activity against root-knot nematodes. Canadian Journal of Microbiology 46:775–783. Harmon, G.E. 2000. The myths and dogmas of biocontrol: Changes in perceptions derived from research on Trichoderma harzianum strain T-22. Plant Disease 84:377–393. Hughes, K.W., R.H. Petersen, and E.B. Lickey. 2009. Using heterozygosity to estimate a percentage DNA sequence similarity for environmental species’ delimitation across basidiomycete fungi. New Phytologist 182:795–798. Inglis, G.D., and A.C. Cohen. 2004. Influence of antimicrobial agents on the spoilage of a meat-based entomophage diet. Journal of Economic Entomology 97:235–250. Inglis P., and M. Tigano-Milani. 2006. Identification and taxonomy of some entomopathogenic Paecilomyces spp. (Ascomycota) isolates using rDNA-ITS sequences. Genetics and Molecular Biology 29:132–136. Jounvenaz, D.P., and J. Kimbrough. 1991. Myrmecomyces annellisae gen. nov., sp. nov. (Deuteromycotina: Hypomycetes), an endoparasitivc fungus of fire ants, Solenopsis spp. (Hymenoptera: Formicidae). Mycological Research 95:1395–1401. Jouvenaz, D.P., G.E. Allen, W.A. Banks, and D.P. Wojcik. 1977. A survey for the pathogens of fire ants, Solenopsis spp., in the southeastern United States. Florida Entomologist 60:275–279. Lard, C., D. Willis, V. Salin, and S. Robison. 2002. Economic assessment of fire ants on Texas urban and agricultural sectors. Southwestern Entomologist 25:123–137. Lofgren, C.S. 1986. History of imported fire ants in the United States. Pp. 36–47, In C.S. Lofgren and R.K. Vander Meer (Eds.). Fire Ants and Leaf-cutting Ants: Biology and Management. Westview Press, Boulder, CO. 435 pp. MacGown, J. 2014. Ants (Formicidae) of the southeastern United States. Available online at http://mississippientomologicalmuseum.org.msstate.edu/Researchtaxapages/Formicidaehome. html#.VNkDEIWR_go. Accessed 11 December 2014. Meyling, N.V., and J. Eilenberg. 2007. Ecology of the entomopathogenic fungi Beauveria bassiana and Metarhizium anisophiae in temperate agroecosystems: Potential for conservation biological control. Biological Control 43:145–155. O’Donnell, K., and E. Cigelnik. 1997. Two divergent intragenomic rDNA ITS2 types within a monophyletic lineage of the fungus Fusarium are nonorthologous. Molecular Phylogeny and Evolution 7:103–116. O’Donnell, K., E. Cigelnik, and H. I. Nirenberg. 1998. Molecular systematics and phylogeography of the Gibberella fujikuroi species complex. Mycologia 90:465–493. Oi, D.H., and D.F. Williams. 2002. Impact of Thelohania solenopsae (Microsporidia: Thelohaniida) on polygyne colonies of Red Imported Fire Ants (Hymenoptera: Formicidae). Journal of Economic Entomology 95:558–562. Patterson, R.S., J.A. Briano, K.B. Wildey, and W.H. Robinson. 1993. Potential of three biological control agents for suppression of Solenopsis invicta, the Red Imported Fire Ant. Pp. 35–43, In K.B. Wildey (Ed.). Proceedings of the 1st International Conference on Insect Pests in the Urban Environment. ICIPUE Cambridge, UK. 498 pp. Price, P.W. 1997. Insect Ecology. John Wiley & Sons, Inc., New York, NY. 874 pp. Sanchez-Peña, R. 1992. Entomopathogenic fungi against Red Imported Fire Ant Solenopsis invicta (Hymenoptera: Formicidae). M.Sc. Thesis. Texas Tech University, Lubbock, TX. Southeastern Naturalist S. Woolfolk, C.E. Stokes, C. Watson, R. Brown, and R. Baird 2016 Vol. 15, No. 2 234 SAS Institute. 1999. SAS/STAT, V8. Cary, NC. Stephenson, S.L. 1989. Distribution and ecology of myxomycetes in temperate forests. II. Patterns of occurrence on bark surface of living trees, leaf litter, and dung. Mycologia 81:608–621. Stephenson, S.L., M. Schnittler, and C. Lado. 2004. Ecological characterization of tropical myxomycete assemblage: Maquipucuna Cloud Forest Reserve, Ecuador. Mycologia 96:488–497. Stimac, J.L., R.M. Pereira, S.B. Alves, and L.A. Wood. 1993. Beauveria bassiana (Balsamo) Vuillemin (Deuteromycetes) applied to laboratory colonies of Solenopsis invicta Buren (Hymenoptera: Formicidae) in soil. Journal of Economic Entomology 86:348–352. Storey, G.K., R.K. Vander Meer, D.G. Boucias, and C.W. McCoy. 1991. Effects of Fire Ant (Solenopsis invicta) venom alkaloids on the in vitro germination and development of select entomogenous fungi. Journal of Invertebrate Pathology 58:88–95. Streett, D.A., T.B. Freeland Jr., and R.K. Vander Meer. 2006. Survey of imported fire ant populations in Mississippi. Florida Entomologist 89:91–92. Triplehorn, C.A., and N.F. Johnson. 2004. Borror and DeLong’s Introduction to the Study of Insects. 7th Edition. Thomson Brooks/Cole, Belmont, CA. 864 pp. Tschinkel, W.R. 2005. The Fire Ants. Harvard University Press, Cambridge, MA. 747 pp. Vander Meer, R., and C. Lofgren. 1985. Biochemical evidence for hybridization in fire ants. Florida Entomologist 68:501–506. Vander Meer, R.K., K. Jaffe, and A. Cedeno. 1990. Applied Myrmecology: A World Perspective. Westview Press, Boulder, CO. 741 pp. Vanderwoude, C., L.A. Lobry de Bruyn, and A.P.N. House. 2000. Response of an openforest ant community to invasion by the introduced ant, Pheidole megacephala. AustEcology 25:253–259. Vinson, S.B., and A.A. Sorensen. 1986. Imported fire ants: Life history and impact. Texas Department of Agriculture, Austin, TX. 28 pp. Western Regional Climate Center. 2015. RAWS USA climate archives. Available online at http://wwwraws.dri.edu. Accessed February 2015. Whitcomb, W.H. 1980. Expedition into the Pantanal. Proceedings of Tall Timbers Conference of Ecological Animal Control for Habitat Management 7:113–122. Williams, D.F., D.H. Oi, S.D. Porter, R.M. Pereira, and J.A. Briano. 2003. Biological control of imported fire ants (Hymenoptera: Formicidae). American Entomologist 49:150–163. Woolfolk, S.W., and G.D. Inglis. 2004. Microorganisms associated with field-collected Chrysoperla rufilabris (Neuroptera: Chrysopidae) adults with emphasis on yeast symbionts. Biological Control 29:155–168. Woolfolk, S.W., R.E. Baird, and C.E. Watson. 2016. Survey of bacteria associated with Red Imported Fire Ants from mounds in Mississippi. Southeastern Naturalist 15(1):83–101. Zettler, J.A., T.M. McInnis, C.R. Allen, and T.P. Spira. 2002. Biodiversity of fungi in Red Imported Fire Ant (Hymenoptera: Formicidae) mounds. Annals of the Entomological Society of America 95:487–491. Zhao J., C. Li, W. Wang, C. Zhao, M. Luo, F. Mu, Y. Fu, Y. Zu, and M. Yao. 2013. Hypocrea lixii, novel endophytic fungi producing anticancer agent cajanol, isolated from Pigeon Pea (Cajanus cajan [L.] Millsp.). Journal of Applied Microbiology 115:102–113.