Introduction

Ambrosia beetles are small wood boring insects that reside in the true weevil family Curculionidae. Approximately 3500 species have been described in two sub-families (Platypodinae and Scolytinae), all of which have obligate associations with filamentous fungi (Six 2012; Jordal 2015). These fungi are primarily Ascomycota that are cultivated along the gallery walls and serve as the primary food source of the beetles and their growing broods (Batra 1966; Massoumi-Alamouti et al. 2009). While developing in the brood galleries, the beetles collect the spores of the fungal symbionts and store them within specially evolved structures known as mycangia (Klepzig and Six 2004). This not only maintains the association of the fungal symbionts within and between different generations of the beetles, but also provides the fungi with a consistent means of dispersal and introduction into a relatively competition-free environment in which they proliferate (Six 2012). In return, the fungi concentrate available nutrients from the host into a form usable for their beetle partners (De Fine Licht and Biedermann 2012).

The symbiotic fungi of bark and ambrosia beetles represent a polyphyletic assemblage of filamentous fungal genera, which have evolved convergent morphological traits that favour insect dispersal (Cassar and Blackwell 1996; Zipfel et al. 2006). Most fungal symbionts of ambrosia beetles reside in the orders Ophiostomatales and Microascales, although some species of Hypocreales and Basidiomycota have also been discovered (Kolařík and Kirkendall 2010; Kasson et al. 2013, 2016; Machingambi et al. 2014; Lynn et al. 2020). In the Ophiostomatales, there are three genera regarded as primary ambrosia beetle symbionts. These include Affroraffaelea (Bateman et al. 2017), Aureovirgo (van der Linde et al. 2016), and Raffaelea sensu lato (Dreaden et al. 2014). Most of the remaining genera in this order associate with bark beetles although some species, are associates of other arthropods such as mites or they occur in non-insect niches such as soil (De Beer and Wingfield 2013). Additionally, the family includes species such as Hawksworthiomyces lignivorus, which was originally isolated from decaying telephone poles (De Meyer et al. 2008), as well as a small number of species in the Sporothrix schenckii clade that are opportunistic human and animal pathogens (López-Romero et al. 2011).

It has been argued for a relatively long time that obligate insect-fungus mutualisms, such as the ambrosia symbioses, represented a one-on-one relationship (Hubbard 1896; Talbot 1977; Cook and Rasplus 2003). This hypothesis appeared to hold true for ambrosia beetles as they were typically found associated with a single, dominant primary symbiont. However, Batra (1966, 1967) opposed this view and a few recent studies have shown that ambrosia beetles can also associate with multiple secondary (or auxiliary) fungal species (Kolařík and Kirkendall 2010; Carrillo et al. 2019). In many cases, these auxiliary associates reside in the same orders as the primary symbionts, and in some cases the primary symbiont of one beetle species may serve as an auxiliary species of another (Batra 1966). However, unlike the primary fungal symbionts, the roles of auxiliary species remain unclear although various hypotheses have been proposed. These include (1) serving as a nutritional source during brood development and succession (Freeman et al. 2016); (2) enabling a beetle to adapt to a new host or environment (Carillo et al. 2014) and (3); increasing beetle fitness by reducing host tree defences and allowing colonization by the primary symbiont (Saucedo et al. 2018).

In many Southern hemisphere countries, including South Africa, ambrosia beetles and their associated fungi are poorly known. This is attributed to the fact that most of these beetles are regarded as harmless, secondary pests infesting stressed or dying trees (Huclr et al. 2017). However, with increasing globalization and the introduction of invasive pests and their associated pathogenic fungi, interest regarding ambrosia beetles in their native ranges and their potential to become economically significant has increased (Liebhold et al. 1995; Ploetz et al. 2013; Hulcr et al. 2017). This elevated interest, as well as the importance of these insects and their fungal associates, prompted the present study to investigate the diversity ophiostomatalean fungi associated with some commonly encountered ambrosia beetles in a native forest area of South Africa.

Materials and methods

Collection of beetles and isolation of fungi

Beetle specimens were obtained from direct field sampling. Field collections were carried out at two locations in Tzaneen, Limpopo Province, South Africa (23° 42′ 29.491″ S 30° 5′ 57.638″ E and 23° 44′ 29.491″ S 30° 11′ 15.417″) using a modified Bambara beetle trap (Hulcr and McCoy 2015). The modified trap contained a wire mesh insert (gap size of 0.5 mm) between the entrance and collection zone, preventing the insects from making contact with the ethanol lure. Traps were set out in the late afternoon with 90% ethanol and left over-night, after which live beetles were collected early the following morning. Beetles were placed on the surfaces of 65 mm Petri dishes and allowed to walk over the agar. These Petri dishes contained malt extract agar (MEA: 2% malt extract and 2% Difco® agar, Biolab, Midrand, South Africa) amended with streptomycin (0.04%, Sigma-Aldrich, Missouri, United States) to control growth of bacteria and cycloheximide (0.03%, Sigma-Aldrich) that is selective for species in the Ophiostomatales. The beetles were removed from the plates after 24 h and transferred into individual cryotubes containing 90% ethanol and stored at − 20 °C for species-level identification.

Petri dishes were inspected regularly for fungal growth and ophiostomatalean isolates were purified by transferring hyphal tips to new MEA plates. Multiple isolates with culture morphologies resembling those of the ophiostomatalean fungi were obtained and used for morphological and DNA sequence-based characterisation. Pure cultures are maintained in the culture collection (CMW) of the Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, South Africa (Table 1) and representative isolates of novel taxa were also deposited in the culture collection (CBS) of the Westerdijk Fungal Biodiversity Institute, Utrecht, The Netherlands.

Table 1 Isolates obtained in this study
Full size table

Identification of beetles

Beetle specimens were examined using an automated Zeiss Discovery V12 dissection microscope (Zeiss, Oberkocken, Germany). Based on overall morphology, beetles were first sorted into groups before they were identified to species level using the key of Rabaglia et al. (2006). The ventral, lateral and dorsal aspects of specimens were examined, and photographic images were captured using a Zeiss Axiocam IcC5 (Zeiss, Oberkocken, Germany). Focus-stacked photographs were produced for the dorsal aspects using Helicon Focus v. 5 (HeliconSoft, Kharkiv, Ukraine) with up to 30 different images.

One of the beetle specimens from which fungal isolates were obtained had obscure morphological characteristics and the specimen was subjected to PCR amplification and sequencing of the ribosomal large subunit (28S) gene region. DNA was extracted using the Macherey Nagel NucleoSpin Tissue Kit (Macherey–Nagel, Dueren, Germany) from the dissected head of the beetle. DNA extraction was performed following the manufacturer’s protocols, except for the final elution volume that was reduced to 60 µl. PCR amplification of the partial ribosomal large subunit (28S) was done using the primers 3665 and 4068 (Belshaw and Quicke 1997; Cognato 2013) in 25 µL reaction volumes as described by Cognato (2013). PCR products were treated with ExoSAP-IT™ PCR Product Clean-up Reagent (ThermoFisher Scientific, Massachusetts, United States). Sequencing reactions were carried out in both the forward and reverse directions using the same primers used in PCR using the BigDye® Terminator v3.1 cycle sequencing kit (ThermoFisher Scientific) with an annealing temperature of 55 °C. Sequencing PCR products were precipitated using the sodium acetate/ethanol method and submitted to the DNA sanger sequencing facility based at the University of Pretoria for analyses on ABI PRISM®3500 Genetic Analyzer (Applied Biosystems, California, United States). The software package Sequence Scanner v. 1.0 (https://sequence-scanner-software.software.informer.com/) was used for quality assessment and editing of the obtained sequencing reads. The consensus sequence was used in a BLASTN search against NCBI GenBank nr/nt database (www.ncbi.nlm.nih.gov/Genbank) to confirm putative morphological identification. The newly obtained sequence has been deposited in NCBI GenBank with the accession number MT355516.

Identification of fungal isolates

The obtained fungal isolates were grouped based on culture morphology and DNA extraction was carried out using lyophilized mycelium of representative isolates following the method of De Beer et al. (2014). PCR amplification was carried out for the partial 28S ribosomal large subunit (LSU) using the primers LR0R and LR5 (Vilgalys and Hester 1990), the internal transcribed spacer region (ITS) using primers ITS1F (Gardes and Bruns 1993) and ITS4 (White et al. 1990) and the partial Beta-tubulin (β-tubulin) gene using the primers T10 (O’Donnel and Cigelnik 1997) and Bt2B (Glass and Donaldson 1995). PCR reactions were carried out as described by De Beer et al. (2014) in 25 µL reaction volumes and the annealing temperature was set at 55 °C. PCR products were treated with ExoSAP-IT™ PCR Product Clean-up Reagent (ThermoFisher Scientific). Sequencing reactions, precipitation and quality assessment was carried out as described above. Reactions were performed in both the forward and reverse direction using the same primers used in PCR and an annealing temperature of 55 °C was used for all three gene regions. Newly obtained sequences have been deposited in NCBI GenBank with the accession numbers provided in Table 1.

Obtained consensus sequences were used in a BLASTN search against NCBI GenBank nr/nt database for putative identification to genus level. Based on these identifications, various datasets were prepared and analysed for each genus separately. Datasets were prepared using MEGA v. 7.0.26 (Kumar et al. 2016) and alignments were done using the online version of MAFFT v. 7 (Katoh and Standley 2013) with default settings. Due to high variability within the three ITS datasets, alignment was subjected to Gblocks v. 0.91b (Castresana 2000) analysis with the less stringent options to remove ambiguous aligned positions before being used for phylogenetic analyses.

Maximum parsimony analyses were performed using MEGA v. 7.0.26. Phylogenies were generated using Subtree-Pruning-Regrafting algorithm starting with 10 random initial trees. Alignment gaps and missing data were included. Confidence levels for the nodes were tested using 1000 bootstrap replicates.

Maximum likelihood analyses were performed using the software raxmlGUI v. 2 (Silvestro and Michalak 2012; Elder et al. 2019) following the General Time Reversible + GAMMA (GTR + G) nucleotide substitution model. Ten random ML searches followed by 1000 bootstrap replicates were performed.

Bayesian inference analyses were performed using MrBayes v. 3.2.5 (Ronquist and Huelsenbeck 2003). Phylogenies were generated following the Markov Chain Monte Carlo (MCMC) method using the GTR + G model. Ten parallel runs, each with four independent MCMC chains were conducted. Trees were sampled every 1000 generations for 10 million generations. Trees sampled in the burn-in phase (25% of trees sampled) were discarded and posterior probabilities were calculated from the remaining trees. Multi-locus phylogenies were constructed for Ceratocystiopsis and Raffaelea using the same methods described above on the combined LSU, ITS and βT datasets.

Morphological observations


Microscopic structures—Fungal isolates were examined using a Zeiss AxioScop 2 compound microscope with an affixed Zeiss Axiocam 105 color camera and a Swift M3602-3DGL light microscope with a built-in 3-megapixel digital camera. Specimens of actively growing cultures were mounted in 80% lactic acid (ThermoFisher Scientific) and examined using bright field and differential interference contrast microscopy. Fifty to 100 measurements were made for all characteristic morphological structures using Zen Blue v. 2.6 (Zeiss). Measurements were taken for conidia, conidiophores, conidiogenous cells, and some additional characteristics when present. Values are presented as minimum—maximum.


Colony growth and morphology—A growth study was conducted to determine optimal growth conditions for two putative new species. Two isolates of each lineage were used to determine growth rate and culture morphology. Agar plugs of 5 mm in diameter were excised from the edge of an actively growing culture and placed, mycelium side down, onto the centre of a 65 mm 2% MEA plate. Plates were incubated in the dark at a temperature range from 15 to 30 °C at 5 °C intervals. Three replicates were carried out for each isolate at each temperature. Isolates were maintained for 10 days after which mean colony diameter was determined. Colony colours were determined using the colour charts of Rayner (1970).

Results

Identification of fungi

The collected beetle specimens yielded 38 isolates with morphologies resembling those of ophiostomatalean fungi. These included five distinct morphological groups and of these, 27 isolates (Table 1) were selected for DNA extraction, PCR amplification and sequencing of the LSU region.

BLAST searches using the LSU sequences revealed that all sequenced isolates were members of the Ophiostomatales. Based on the preliminary taxonomic groupings identified using BLAST, an LSU dataset (LSU-Ophiostomatales) was prepared and analysed to identify generic placement of the isolates. Phylogenetic analyses of the LSU-Ophiostomatales dataset (Fig. 1) including representative isolates of all genera of the Ophiostomatales, showed that the isolates considered in this study resided in in three different genera, Ceratocystiopsis, Ophiostoma, and Raffaelea, and represented five separate lineages (lineages A–E, Fig. 1).

Fig. 1
figure 1

RAxML phylogram derived from analysis of LSU data including all major groups in the Ophiostomatales. Bold branches indicate posterior probabilities ≥ 95%. Bootstrap values ≥ 75% for maximum likelihood and maximum parsimony analyses are indicated at nodes as ML/MP. G = sequence extracted from available genome. T = Ex-Type isolate

Full size image

Based on the genus and species complex assignments obtained from analyses of the LSU dataset (Fig. 1), seven genus level datasets were prepared and analysed (Tables 2, 3). The number of characters of each aligned dataset, numbers of parsimony informative characters, and additional information regarding phylogenetic analyses are presented in Table 3.

Table 2 Isolates used in phylogenetic analyses and their sequence accession numbers
Full size table
Table 3 Number of characters strains and substitutional models used in phylogenetic analyses
Full size table

Ophiostoma—Based on the analyses of the LSU-Ophiostomatales dataset, five isolates grouped as a single lineage (A) in the O. pluriannulatum complex. Phylogenetic analyses of the ITS-Oph dataset (Fig. 2), that included a subset of Ophiostoma species focussed on the O. pluriannulatum complex, identified this species as O. palustre.

Fig. 2
figure 2

RAxML phylogram derived from analysis of ITS-Ophiostoma data including taxa from the genus Ophiostoma specifically focussing on the O. pluriannulatum complex. Bold branches indicate posterior probabilities ≥ 95%. Bootstrap values ≥ 75% for maximum likelihood and maximum parsimony analyses are indicated at nodes as ML/MP. G = sequence extracted from available genome. T = Ex-Type isolate

Full size image

Ceratocystiopsis—Two isolates resided in the genus Ceratocystiopsis based on the analyses of the LSU-Ophiostomatales dataset. These isolates formed a distinct clade with high statistical support together with sequences of three undescribed isolates available from GenBank (lineage C). Phylogenetic analyses of the genus level datasets (LSU, ITS, and β-tubulin) (Fig. 3a-c) all supported this grouping. No ITS and β-tubulin sequence data were available for these undescribed isolates, but based on their grouping and sequence conservation in the LSU analysis, they likely belong to the same species as the isolates recovered from the current study.

Fig. 3
figure 3

RAxML phylogenies derived from analyses of various Ceratocystiopsis datasets including all major taxa described in the genus. a Phylogram derived from the LSU analyses; b Phylogram derived from the ITS analyses; c Phylogram derived from the β-tubulin analyses. Bold branches indicate posterior probabilities ≥ 95%. Bootstrap values ≥ 75% for maximum likelihood and maximum parsimony analyses are indicated at nodes as ML/MP. G = sequence extracted from available genome. T = Ex-Type isolate

Full size image

Raffaelea—Twenty isolates, representing three morphological groups, grouped in three separate lineages within Raffaelea sensu lato (Fig. 1). Ten isolates grouped in the Raffaelea sulphurea complex (lineage B) and were identified as R. sulphurea based on phylogenetic analyses of LSU and ITS sequence data (Figs. 1 and 4a and b). The remaining ten isolates grouped as two distinct lineages (D and E) within Raffaelea sensu stricto (Figs. 1, 4). Five isolates were identified as R. arxii based on phylogenetic analyses of LSU and ITS sequence data (Figs. 1, 4a and b). The five remaining isolates (lineage D) together with an additional isolate labelled as Raffaelea sp. PL1001 (Eskalen & McDonald 2011; Dreaden et al. 2014), for which LSU and β-tubulin sequences are available in GenBank, formed a distinct lineage separate from all other species in the genus, but closely related to R. cyclorhipidia (Fig. 4b and c). An additional isolate labelled as Raffaelea sp. Hulcr7507 (Simmons et al. 2016), for which ITS and β-tubulin sequences are available on GenBank, grouped peripheral to this clade. Based on phylogenetic placement and ITS and β-tubulin sequence conservation this isolate most likely belongs to the same species as Raffaelea sp. PL1001 and the five isolates recovered from the current study.

Fig. 4
figure 4

RAxML phylogenies derived from analyses of various Raffaelea datasets including all major taxa described in the genus. a Phylogram derived from the LSU analyses; b Phylogram derived from the ITS analyses; c Phylogram derived from the β-tubulin analyses. Bold branches indicate posterior probabilities ≥ 95%. Bootstrap values ≥ 75% for maximum likelihood and maximum parsimony analyses are indicated at nodes as ML/MP. G = sequence extracted from available genome. T = Ex-Type isolate

Full size image

Identification of beetles

The fungal isolates obtained in this study originated from 16 beetle specimens. Based on their morphological characters, these beetles were identified as four different species (Fig. 5). Three of these species resided in the Xyleborini (Scolytinae), namely Xyleborinus (Xbi.) saxesenii; Xylosandrus (Xs.) crassiusculus; and Xyleborus (Xbo.) affinis (Fig. 5b-d). The fourth was a species of Bostrichidae (Fig. 5a). Isolates used in this study originated from thirteen Xbi. saxesenii specimens, and a single specimen each of Xs. crassiusculus, Xbo. affinis and the unknown Bostrichid. Due to some obscured morphological characters for the Xbo. affinis specimen, DNA sequencing of the ribosomal large subunit was carried out and results positively confirmed our identification.

Fig. 5
figure 5

Beetle species from which fungal isolates were obtained; a A unknown species of Bostrichidae beetle; b Xyleborinus saxesenii; c Xylosandrus crassiusculus; d Xyleborus affinis. Scale = 1 mm

Full size image

Beetle-fungus associations

Thirty-eight isolates with ophiostomatalean-like morphologies were isolated from 16 living beetle specimens. After initial separation based on morphology, 27 representative isolates were selected for further DNA sequence-based characterisation. Five of these fungal isolates were identified as O. palustre (Figs. 1, 2—lineage A) and all five isolates originated from a single beetle specimen in the Bostrichidae. Twenty-one isolates recovered from Xbi. saxesenii specimens were putatively identified as R. sulphurea based on morphology and this was confirmed by DNA sequencing of ten representative isolates originating from different beetles (Figs. 1,4—lineage B). Another five isolates also obtained from Xbi. saxesenii specimens were identified as a new species of Raffaelea (Figs. 1, 4—lineage D). Five isolates identified as R. arxii (Figs. 1, 4—lineage E) were obtained from a single beetle specimen identified as Xbo. affinis. Two isolates were obtained from a single specimen of Xs. crassiusculus and were identified as a new species of Ceratocystiopsis (Figs. 1, 3—lineage C).

Taxonomy

Ceratocystiopsis lunata W.J. Nel sp. nov. Figure 6.

Fig. 6
figure 6

Morphological characteristics of asexual structures of Ceratocystiopsis lunata sp. nov. a–c Conidiogenous cells giving rise to oblong conidia; d Conidia; e–g Conidiogenous cells giving rise to falcate conidia; g arrows Presence of denticles; h Pure culture gown on MEA in the dark for 12 days. Scale = 10 µm

Full size image
MycoBank MB838616

Etymology: Name reflects the crescent shaped conidia.


Description: Conidiophores mononematous, macronematous, arising from vegetative hyphae, simple, upright, straight, curved or undulate, 5–97 µm long (avg. 26.1 ± 15.8 µm). Conidiogenous cells integrated, hyaline, blastic, sometimes denticulate, 3–42 × 1.5–3.1 µm (avg. 13.7 ± 5.5 × 2.1 ± 0.3 µm). Conidia hyaline, aseptate, two types falcate to crescent shaped, no sheath 4.9–9.5 × 1.3–2.8 µm (avg. 6.6 ± 0.8 × 1.9 ± 0.3 µm) and oblong with the upper part swollen, apex round, tapering toward base, base truncated, 2–6 × 1.2–3.1 µm (avg. 4.5 ± 0.6 × 2.1 ± 0.4 µm), yeast-like budding observed in fresh culture.


Specimens examined: South Africa, Limpopo, Tzaneen, isolated from a living Xs. crassciusculus beetle, W. J. Nel. 21 September 2019, holotype (PREM 63099, living culture ex-holotype CMW 55897 = CBS 147171).


Additional specimens: South Africa, Limpopo, Tzaneen, isolated from a living Xs. crassciusculus beetle, W. J. Nel. 21 September 2019, paratype (PREM 63100, living culture CMW 55898 = CBS 147172).


Cultures: Moderate growth rate on 2% MEA in dark. Grows best at 30 °C reaching an average of 48.8 mm (± 0.9 mm) in 10 d. Colony growth circular with smooth margins, both abundant aerial and submerged mycelia present, flat, whitish to creamy in colour.


Notes: Ceratocystiopsis lunata can be distinguished from other species of Ceratocystiopsis based on its conidial morphology. Ceratocystiopsis lunata produces both falcate and oblong conidia whereas other species of Ceratocystiopsis typically only produce oblong conidia (De Beer and Wingfield 2013).

Raffaelea promiscua W.J. Nel sp. nov. Figure  7 .

Fig. 7
figure 7

Morphological characteristics of asexual structures of Raffaelea promiscua sp. nov. a–e Conidiogenous cells giving rise to conidia; a and b arrows indicate conidiogenesis taking place; f Conidia; g Pure culture gown on MEA in the dark for 12 days. Scale: a–c, f = 10 µm; d–e  = 20 µm

Full size image
MycoBank MB838615

Etymology: Name refers to the promiscuous (Promiscuum L.) nature of the species that is associated with different ambrosia beetles.


Description: Conidiophores mononematous, macronematous, arising from vegetative hyphae, mostly simple, occasionally branched, upright, straight, curved or undulate, tapering towards apex, 14–195 µm long, reduced to conidiogenous cell. Conidiogenous cells integrated, hyaline to lightly pigmented, cylindrical, or peg-like, tapering towards apex, blastic, 5–56 × 1–3.9 µm (avg. 21.9 × 2.4 µm). Conidia hyaline, aseptate, majority oblong with the upper part swollen, apex round, tapering toward base, base truncated, 2–9 × 1–4 µm (avg. 5.1 × 2.5 µm), yeast-like budding observed in fresh culture.


Specimens examined: South Africa, Limpopo, Tzaneen, isolated from a living Xbi. saxesenii beetle, W. J. Nel. 21 September 2019, holotype (PREM 63101, living culture CMW 55899 = CBS 147173).


Additional specimens: South Africa, Limpopo, Tzaneen, isolated from a living Xbi. saxesenii beetle, W. J. Nel. 21 September 2019, paratype (PREM 63102, living culture CMW 55900 = CBS 147174). South Africa, Limpopo, Tzaneen, isolated from a living Xbi. saxesenii beetle, W. J. Nel. 21 September 2019 (living culture CMW 55901 = CBS 147175). South Africa, Limpopo, Tzaneen, isolated from a living Xbi. saxesenii beetle, W. J. Nel. 21 September 2019 (living culture CMW 55902 = CBS 147176).


Cultures: Slow growing on 2% MEA in dark. Grows best at 25 °C reaching and average of 34 mm (± 1.9 mm) in 10 d. Colonies circular with smooth margins, mycelia mostly submerged, aerial hyphae present, flat, initially whitish turning brownish olive to dark greyish brown with age starting at the centre of the colony. Yeast-like growth often present at inoculation site from initial growth of colony.


Notes: Raffaelea promiscua can be distinguished from its sister taxon R. cyclorhipidia (Simmons et al. 2016) by its smaller conidia and the fact that these are predominantly oblong with enlarged apices, and are 5.1 × 2.5 µm on average, whereas those of R. cyclorhipidia are elliptical to elongate and 7.3 × 3.5 µm on average. Raffaelea promiscua colonies are smooth with aerial hyphae whereas R. cyclorhipidia has a tough and wrinkled appearance.

Discussion

A total of 38 ophiostomatalean fungal isolates were obtained from 16 adult beetles representing Xs. crassiusculus, Xbi. saxesenii, Xbo. affinis, and an unidentified species of Bostrichidae. The fungi were identified as five distinct species in the Ophiostomatales, one of which was recorded from South Africa for the first time and two represented novel species described here as C. lunata and R. promiscua.

An unusual association of O. palustre with a species of Bostrichidae emerged from this study. Although some small species of Bostrichidae can easily be confused with species of Scolytinae (Ivie 2002), these beetles are not known to associate with fungi, preferentially infesting wood with a low moisture content (Creffield 1996; Ivie 2002). However, some species of Ophiostoma, including many species in the O. pluriannulatum complex such as O. palustre (Osorio et al. 2016), are associated with wounds on trees. It is consequently possible that the bostricid beetle accidentally picked up spores of O. palustre colonizing the wounded tissue induced by its gallery. Many ophiostomatalean fungi also associate with phoretic mites vectored between hosts by various beetles (Hofstetter et al. 2013) and it is also possible that the fungus originated from a mite carried by the beetle.


Xyleborinus saxesenii is an ambrosia beetle with a cosmopolitan distribution and is among the most common ambrosia beetles found globally. This species was originally described from Germany (Ratzeburg 1837) and has been detected on every continent except Antarctica (Wood and Bright 1992). Considerable research has been conducted on the fungal associates of Xbi. saxesenii, with its primary fungal symbiont Raffaelea sulphurea first described by Batra (1967). Subsequently, a second dominant mycangial symbiont of this insect, Fusicolla acetilerea, was detected by Biedermann et al. (2013). Xbi. saxesenii is also associated with a number of less dominant fungi including species of Paecilomyces, Cladosporium, Ramularia, and Aureobasidium (Biedermann et al. 2013; Malacrinò et al. 2017).


Xyleborinus saxesenii was first reported in South Africa by Schedl (Schedl 1975) where it was collected in the Western Cape and Kwazulu-Natal provinces, but its fungal associates were not considered. In the present study, 26 fungal isolates were obtained from specimens of Xbi. saxesenii. Of these, 21 were identified as R. sulphurea and this represents the first report of the fungus from South Africa. The remaining five fungal isolates grouped together with an undescribed Raffaelea sp. isolated from an unknown ambrosia beetle recorded as PL1001 by Eskalen and McDonald (2011). Subsequent to its first report, this undescribed fungus has been identified in several studies associated with various Xyleborini including Xbo. bispinatus, Xbo. volvulus and Xbi. saxesenii (Cruz et al. 2018, 2019; Saucedo-Carabez et al. 2018). Based on their distinct morphology and phylogenetic grouping separating them from their closest relative, R. cylcorhipidia, the isolates from the present study, including Raffaelea sp. PL1001, were described as R. promiscua.


Xylosandrus crassiusculus is a commonly encountered, cosmopolitain ambrosia beetle. Yet relatively little is known regarding the fungi associated with this beetle. Its primary fungal symbiont, Ambrosiella roeperi, was described relatively recently (Harrington et al. 2014), which is surprising given that the beetle was described as long ago as 1866. Recent pyrosequencing of the mycangial community of Xs. crassiusculus revealed that this niche is dominated by Ambrosiella, but other species including Ceratocystis, Fusarium, Cladosporium and various yeasts were also shown to be present (Kostovic et al. 2015).

In this study, two isolates of Ceratocystiopsis were obtained from a living Xs. crassiusculus specimen. Species of Ceratocystiopsis are common symbionts of scolytine bark beetles, which are close relatives of the ambrosia beetles. Previous studies investigating the fungal symbionts of ambrosia beetles in the Platypodinae showed that some species of Ceratocystiopsis have a promiscuous relationship with these insects (Inácio et al. 2012; Li et al. 2018). Li et al. (2018) concluded that their Ceratocystiopsis sp. 2 obtained from Euplatypus compositus, E. parelellus and Oxoplatypus quadridentatus, was the same species as Ophiostoma sp. X obtained from Platypus cylindrus in a study by Inácio et al. (2012). Phylogenetic analysis of the LSU region of the Ceratocytiopsis isolates obtained from Xs. crassiusculusin the present study, grouped these isolates along with those obtained by Li et al. (2018) and they were described here as C. lunata. This is the first time that C. lunata has been obtained from an ambrosia beetle in the Scolytinae. Our results, therefore, suggest that it could be an auxiliary symbiont of multiple species of ambrosia beetles.


Xyleborus affinis is a pan topical species of ambrosia beetle, native to tropical America (Rabgalia et al. 2006). This species has been introduced into Europe, Asia, Australia and Africa, where its presence was later recorded in South Africa in the early 1980’s (Schedl 1982; Rabgalia et al. 2006). Although the beetle was described more than a decade ago, its primary fungal symbiont is unknown. However, community pyrosequencing of the mycangia from 38 Xbo. affinis beetles showed that their fungal communities are highly diverse, including species from the Ophiostomatales, Microascales and a large variety of yeasts (Kostovic et al. 2015).

In this study, five isolates of Raffaelea arxii were obtained from a living specimen of Xbo. affinis. Raffaelea arxii was first described from South Africa as the primary fungal symbiont of Xbo. torquatus (= Xbo. volvulus) by Scott and Du Toit (1970). Aside from being the primary fungal symbiont of Xbo. torquatus, previous studies have found this species to be vectored by numerous other Xyleborus spp. including Xbo. affinis (Campbell et al. 2016; Saucedo-Carabez et al. 2018). Our findings provide additional support for R. arxii being the primary symbiont of both Xbo. volvulus and Xbo. affinis, as has previously been hypothesized (Saucedo et al. 2016). However, because R. arxii was obtained only from the body surfaces of the beetle specimen and not the mycangia, its symbiotic relationship with Xbo. affinis could not be deduced from this study.

Conclusions

Ambrosia beetles associate with various fungal symbionts that act as their source of nutrition, aid in their development and play an important part in their adaptive success. In this study, five species of ophiostomatalean fungi were obtained from four species of wood-boring beetle, three species of scolytine ambrosia beetle and one species of bostrichid beetle. Based on morphological characters and DNA sequence data, two new species of ophiostomatalean fungi were described and one is reported from the country for the first time. This study, like many others investigating the ophiostomatalean fungi in South Africa has led to the discovery and description of new species, suggesting that there are more novel species to be discovered. However, with very few studies focussed on investigating the diversity of ophiostomatalean fungi associated with ambrosia beetles in South Africa, this appears to be a niche that warrants further investigation.