Checklist of thallus-forming Laboulbeniomycetes from Belgium and the Netherlands, including Hesperomyces halyziae and Laboulbenia quarantenae spp. nov.

Abstract In this paper we present an updated checklist of thallus-forming Laboulbeniomycetes (Ascomycota, Pezizomycotina), that is, the orders Herpomycetales and Laboulbeniales, from Belgium and the Netherlands. Two species are newly described based on morphology, molecular data (ITS, LSU ribosomal DNA) and ecology (host association). These are Hesperomyces halyziae on Halyzia sedecimguttata (Coleoptera, Coccinellidae) from both countries and Laboulbenia quarantenae on Bembidion biguttatum (Coleoptera, Carabidae) from Belgium. In addition, nine new country records are presented. For Belgium: Laboulbenia aubryi on Amara aranea (Coleoptera, Carabidae) and Rhachomyces spinosus on Syntomus foveatus (Coleoptera, Carabidae). For the Netherlands: Chitonomyces melanurus on Laccophilus minutus (Coleoptera, Dytiscidae), Euphoriomyces agathidii on Agathidium laevigatum (Coleoptera, Leiodidae), Laboulbenia fasciculata on Omophron limbatum (Coleoptera, Carabidae), Laboulbenia metableti on Syntomus foveatus and S. truncatellus (Coleoptera, Carabidae), Laboulbenia pseudomasei on Pterostichus melanarius (Coleoptera, Carabidae), Rhachomyces canariensis on Trechus obtusus (Coleoptera, Carabidae), and Stigmatomyces hydrelliae on Hydrellia albilabris (Diptera, Ephydridae). Finally, an identification key to 140 species of thallus-forming Laboulbeniomycetes in Belgium and the Netherlands is provided. Based on the combined data, we are able to identify mutual gaps that need to be filled as well as weigh the impact of chosen strategies (fieldwork, museum collections) and techniques in these neighboring countries. The aim of this work is to serve as a reference for studying Laboulbeniomycetes fungi in Europe.


Introduction
Herpomycetales and Laboulbeniales are two orders within the class Laboulbeniomycetes (Ascomycota, Pezizomycotina), consisting of arthropod-associated biotrophs. Both orders are unique among related fungi in that they do not form hyphae; instead, thalli are produced by mitotic divisions from a two-celled ascospore. Herpomycetales was recently described and includes a single genus, Herpomyces Thaxt., with 27 described species-all associated with cockroaches (Blattodea) (Haelewaters et al. 2019b;Gutierrez et al. 2020). The Laboulbeniales order, on the other hand, successfully radiated on a wide range of hosts. Representatives of this order can be found in three arthropod subphyla, including mites and harvestmen (in subphylum Chelicerata), millipedes (in subphylum Myriapoda), and many orders of true insects (in subphylum Hexapoda). The vast majority of about 2,325 described species (Kirk 2019) are known from beetles (order Coleoptera), hence the common name once introduced for the group, "beetle hangers" (Cooke 1892). The early taxonomic history of these fungi is fraught with confusion (Blackwell et al. 2020), but the incorporation of sequence data has led to a conclusive placement of these fungi within Ascomycota (Blackwell 1994;Weir and Blackwell 2001;Schoch et al. 2009).
Early studies on Laboulbeniales (including Herpomyces at that time) in Belgium and the Netherlands are scarce. In Belgium, Collart (1945Collart ( , 1947 and Rammeloo (1986) made noteworthy contributions, followed by multiple publications by De Kesel and colleagues (1989-present). The Laboulbeniomycetes from Belgium were for the first time summarized by De Kesel and Rammeloo (1992), who reported 1 species of Herpomyces and 47 species of Laboulbeniales. De Kesel et al. (2020) provided an updated -and illustrated -Catalogue of the Laboulbeniomycetes of Belgium, with a total of 115 species (3 Herpomycetales, 112 Laboulbeniales) from 222 host species. For more details regarding the study of Herpomycetales and Laboulbeniales in Belgium, we refer to De Kesel and Rammeloo (1992) and De Kesel et al. (2020). In the Netherlands, thus far, no effort has been made to publish a checklist.
The study of Laboulbeniales in the Netherlands started during a meeting of the Dutch Entomological Society in 1906, triggered by a question from Dr. Johannes P. Lotsy, then director of the "Rijksherbarium" (Leiden). In response, Prof. Dr. De Meijere remembered that he once observed an infected Drosophila funebris (Fabricius, 1787) fly, collected at the ARTIS Amsterdam Royal Zoo in 1904, but had not thought it worthy of mention at the time. Recent infected material of D. funebris from nature reserve De Kaaistoep has thus far always been associated with Stigmatomyces entomophilus (Peck) Thaxt. (Haelewaters et al. 2015b) and hence it is likely that S. entomophilus represents the very first report of Laboulbeniales from the Netherlands. The first published account was a developmental study of Stigmatomyces baeri H. Karst. by Boedijn (1923). The fungus was found on an atypical host -Fannia canicularis (Linnaeus, 1761); this fly is the only reported host for Fanniomyces ceratophorus (Whisler) T. Majewski, which is morphologically different from Boedijn's (1923) drawings. We agree with Thaxter (1931) that the fungus was probably correctly identified by Boedijn, but perhaps the host was not.
Next, in the 1930s, only two species of Laboulbeniales were reported in the Netherlands: Laboulbenia cristata Thaxt. from Paederus riparius (Linnaeus, 1758) (Kossen 1936(Kossen , 1938 and Laboulbenia flagellata Peyr. from Platynus spp. (Zaneveld 1938). It was not until Abraham Middelhoek (1906Middelhoek ( -1968) that the number of reported species of Laboulbeniales in the Netherlands would increase by 25 (Middelhoek 1941(Middelhoek , 1942(Middelhoek , 1943a(Middelhoek , b, c, d, 1945(Middelhoek , 1947a(Middelhoek , b, 1949. Middelhoek was first an artist who, among other things, made stained glass windows. Only after World War II, he studied biology and raised an interest in fungi, particularly the Laboulbeniales. After Middelhoek, Laboulbeniales were forgotten about in the Netherlands except for a single paper by Meijer (1975), who proposed to use Laboulbeniales fungi as "biological tags" to trace migration patterns. Since 2012, Haelewaters and colleagues have published several papers dealing with Laboulbeniales in the Netherlands, which together have more than doubled the number of reported species in this country (De Kesel and Gerstmans 2012;Haelewaters 2012Haelewaters , 2013Haelewaters et al. 2012aHaelewaters et al. , b, 2014Haelewaters et al. , 2015aHaelewaters et al. , b, 2020De Kesel et al. 2013;Haelewaters and De Kesel 2013;Haelewaters 2014, 2019;Haelewaters and van Wielink 2016). To date, 79 species of Laboulbeniales are reported from the Netherlands.
In this contribution we compile all available data from Belgium and the Netherlands. Keeping in mind that both countries show some geographical differences, especially due to specific soils and increasing altitude in the southern part of Belgium, we think a combined checklist makes sense at this point. This is mainly because the sampling effort for Laboulbeniomycetes in the southern part of Belgium has been much lower compared to the northern and central areas of the country (De Kesel et al. 2020). As a result, the bulk of Belgian and Dutch records come from biogeographically comparable regions. The here presented checklist is useful to illustrate where mutual gaps need to be filled and what the impact has been of the chosen strategies (fieldwork, museum collections) and trapping techniques. In combination with the recently published Belgian catalogue (De Kesel et al. 2020) presenting illustrations and identification keys to 115 taxa, this checklist will serve as a reference for mycologists, students, and scholars studying Laboulbeniomycetes fungi. In addition, this work is an appropriate starting point for an updated checklist of thallus-forming Laboulbeniomycetes from Europe-an ongoing project that needs to be updated, three decades after the massive undertaking of Santamaría et al. (1991).

Specimen collection and morphological study
Insects were collected in Belgium and the Netherlands using pitfall traps and on an illuminated white screen at night. Specimens were preserved in 96-99% ethanol until they were screened for presence of thalli of Laboulbeniomycetes at 20-50× magnification. Thalli were removed from the host at the foot and mounted in Amann solution following the methods in De Kesel et al. (2020). Drawings and measurements were made using a BX51 light microscope (Olympus, Tokyo, Japan) with drawing tube, digital camera, and AnalySIS software (Soft Imaging System GmbH, Münster, Germany); or an an Olympus BH2 bright field compound microscope with SC30 camera and cellSens 1.18 imaging software.
Infected hosts found in Belgium and the Netherlands are preserved at Meise Botanic Garden (BR) and the Brabant Museum of Nature, Tilburg (NNKN), respectively. Microscope slides of Laboulbeniales are deposited at BR, FH, GENT, and NMBT (Thiers continuously updated).

DNA extraction, PCR amplification, sequencing
Three thalli of Laboulbenia quarantenae sp. nov. were used for DNA isolation using the REPLI-g Single Cell Kit (Qiagen, Stanford, California) with modifications (Haelewaters et al. 2019b). The DNA extract was stored at -20 °C until PCR amplification. Recent studies found that even though the internal transcribed spacer (ITS) region is a good marker for species delimitation in Laboulbeniomycetes, it is difficult to amplify in this group. Instead, the large subunit (LSU) of the ribosomal RNA gene has been put forward as a secondary barcode because it is easy to amplify and provides high discriminative resolution at species-level (e.g., Haelewaters et al. 2018;Sundberg et al. 2018b;Walker et al. 2018;Liu et al. 2020). The partial LSU was amplified using primers LIC15R (Miadlikowska et al. 2002) and LR6 (Vilgalys and Hester 1990). Sequencing was outsourced to Macrogen Europe (Amsterdam, the Netherlands) with the same PCR primers and an additional reverse primer, LR3 (Vilgalys and Hester 1990). Resulting forward and both reverse sequence reads were assembled and edited with Sequencher version 5.2.3 (Gene Codes Corporation, Ann Arbor, Michigan).
For Hesperomyces halyziae, molecular work had been done previously . DNA was extracted using the Extract-N-Amp Plant PCR Kit (Sigma-Aldrich, St. Louis, Missouri) (methods in Haelewaters et al. 2015c). Seven thalli were placed in a 1.5 mL tube with 40 µL of Extraction Solution and sterilized sand. The tube was then placed in a FastPrep FP120 Cell Disrupter (Thermo Fisher Scientific,Waltham,Massachusetts) to mechanically crush fungal material at 5.5 m/s for 20 sec, and then on a heating block to incubate at 95 °C for 10 min. Finally, a total of 120 µL Dilution Solution was added to the mixture. Because we needed to define "H. virescens sensu stricto", additional extractions from single Hesperomyces thalli removed from Chilocorus stigma (Say, 1835) were performed using the REPLI-g Single Cell Kit with modifications. Amplification of the ITS was done using primers ITS1f (Gardes and Bruns 1993) and ITS4 (White et al. 1990) as well as Hesperomyces-specific primers ITShespL and ITShespR (Haelewaters et al. 2019b). Purification and sequencing (same primers) of these PCR products were outsourced to Genewiz (Plainfield, New Jersey).

Phylogenetic analyses
Methods for both datasets -ITS for Hesperomyces, LSU for Laboulbenia -were largely identical. Sequences were downloaded from NCBI GenBank (https://www.ncbi.nlm. nih.gov/genbank/) and supplemented with sequences that were generated during this study. Sequences were aligned using MUSCLE version 3.7 (Edgar 2004), which is available on the CIPRES Science Gateway V. 3.3 (Miller et al. 2010). After alignment of the ITS dataset, partial SSU and partial LSU were removed by looking for the motifs 5'-ATCATTA-3' (3' end of SSU) and 5'-TGACCT-3' (5' start of LSU), and deleting downstream and upstream sequence data, respectively (Baral et al. 2018). For the LSU dataset, we unsuccessfully searched for the 5'-TGACCT-3' motif. We then looked for the motif following 5'-TGACCT-3' in a Hesperomyces sequence (GenBank acc. no. MG757513), which is 5'-CGGAT-3', found this motif in the Laboulbenia dataset, and then realized that the 5' start of LSU in Laboulbenia includes one nucleotide substitution compared to the conventional motif: 5'-TGGCCT-3'. We deleted the downstream sequence data to remove partial ITS. Next, ambiguously aligned regions and uninformative positions were removed using the command line version of trimAl v1.2 (Capella-Gutiérrez et al. 2009) with gap threshold = 0.6 and minimal coverage = 0.5. Models of nucleotide substitution were selected by considering the Akaike Information Criterion corrected for small samples (AICc) with ModelFinder Plus (Kalyaanamoorthy et al. 2017). Maximum likelihood (ML) was inferred for each dataset under the selected model with IQ-TREE (Nguyen et al. 2015;Chernomor et al. 2016). Ultrafast bootstrap (BS) analysis with 1000 replicates estimated branch support in the ML trees (Hoang et al. 2018).
Bayesian analyses were done using a Markov chain Monte Carlo (MCMC) coalescent approach implemented in BEAST 1.8.4 (Drummond et al. 2012), with a strict clock assuming a constant rate of evolution across the tree, a Yule Speciation tree prior (Yule 1925;Gernhard 2008), and the nucleotide substitution model as selected by jModelTest 2.1 (Darriba et al. 2012) under the AICc criterion. For each dataset, four runs were performed from a random starting tree for 10 million generations with a sampling frequency of 1000. All settings were entered in BEAUti 1.8.4 to generate an XML file, which was run in BEAST on the CIPRES Science Gateway (Miller et al. 2010

Checklist
For the checklist of thallus-forming Laboulbeniomycetes from Belgium and the Netherlands, we used De Kesel et al. (2020) for Belgium and all available published papers (since 1938 up to 2020) for the Netherlands. Laboulbeniomycetes and their hosts are listed alphabetically, starting with Herpomycetales, followed by Laboulbeniales. Fungal species are numbered throughout (1-140), authority and reference to the protologue are presented. For each fungus, hosts are presented alphabetically, with classification (order, family) and country in which the association has been reported: "Be" for Belgium, "Nl" for the Netherlands. No detailed collection information is shown except for new country records. In several instances, taxonomic notes are provided. Hosts are according to Vorst (2010) and Beccaloni et al. (2014). Names of fungi correspond to Index Fungorum (2020).

Results
The ITS dataset consisted of 31 Hesperomyces sequences (Table 1) and 724 characters, of which 462 were constant and 198 were parsimony-informative. The selected nucleotide substitution model under AICc was TVM+F+G4 (-lnL = 2790.545, ModelFinder Plus) and TVM+G (-lnL = 2786.8769, jModelTest 2). The Hesperomyces virescens sensu lato  clade has maximum support from both ML and Bayesian analyses (Figure 1). Each of the nine clades within H. virescens s.l. consists of isolates from thalli removed from a single host species, except for the Adalia clade, which includes isolates from both A. bipunctata and A. decempunctata. One of the clades consists of isolates from Chilocorus stigma, the host on which H. virescens was originally described (Thaxter 1891). This clade, representative of Hesperomyces virescens sensu stricto, receives maximum support. The single isolate of Hesperomyces halyziae, from Halyzia sedecimguttata, is placed as sister to H. virescens s.l. from Harmonia axyridis (Pallas, 1773) (pp = 0.8).
The LSU dataset consisted of 24 Laboulbenia sequences (Table 2) and 682 characters, of which 558 were constant and 63 were parsimony-informative. The selected  Description. Thallus 335-453 µm long from foot to perithecial apex; colored yellow except for a somewhat darker region right above the foot. Cell I obtriangular, 2.0-2.5× longer than broad, broadening distally, with very oblique septum I-II. Cell II longer than broad, 23-28 × 16-21 µm, subtrapezoidal in section. Cell III always smaller than cell II, 14-20 × 14-19 µm, with inflated dorsal cell wall. Primary appendage consisting of 4 superposed cells, 61-67 µm long; in the same axis as cells I and III, separated from the latter by the constricted primary septum; its basal cell somewhat longer than broad, longer than each of the remaining cells of the appendage; second to fourth cells carrying a single antherium externally, the fourth cell also carrying a second upwardly directed antherium. Antheridia flask-shaped, Figure 2. Maximum clade creditability tree of Laboulbenia isolates reconstructed from an LSU dataset, with L. bruchii as outgroup. The topology is the result of Bayesian inference performed with BEAST. For each node, ML BS (≥ 65) and Bayesian pp (≥ 0.7) are presented above/below the branch leading to that node. Isolates are color-coded by host; L. quarantenae sp. nov. is highlighted with gray shading. with slightly (dorsally and/or basally) curved efferent necks, the upper antheridium carrying at its dorsal side a pointed process, which represents the original ascospore apex. Cell VI with subparallel margins to broadening distally, 33-70 × 23-33 µm. Perithecium 194-291 × 62-86 µm (not including basal cells), symmetric or with the anterior margin convex and the posterior one almost straight or concave; broadest near the upper third, then gradually tapering towards the apex; apex complex with 2 short lower lobes, 2 upper (terminal) lobes, and 2 prominent lips surrounding the ostiole; lower lobes tapering to a rounded tip, the ventral lobe outwardly directed; terminal lobes unicellular, elongated, 29-42 µm in length, curved upwards and outwardly; ostiole with two lips, 25-29 µm in length, one lip triangular, the other slightly shorter, blunt or rounded, basally carrying the remainder of the trichogyne. Ascospores 70-85 µm long, with conspicuous slime sheath only surrounding the larger cell.  van Wielink 2016, Haelewaters et al. 2017) and H. virescens sensu lato (De Kesel et al. 2020). One unverified record is available from France (Justamond 2019).
Notes. Supported by multi-locus phylogenetic analyses and sequence-based species delimitation methods, Haelewaters et al. (2018) showed that H. virescens Thaxt. is a complex of multiple species, segregated by host. The authors proposed to "restrict H. virescens sensu stricto to those thalli found on Chilocorus stigma, the host species on which the fungus was originally described" (Thaxter 1891). Here, we included two isolates from C. stigma (Say, 1835), and found the clade representative of H. virescens sensu stricto. Based on this analysis and previous work , we can start describing the individual clades as distinct species. A monographic work with formal descriptions for the seven other species within H. virescens s.l. is in preparation, but in the light of this checklist we decided to describe H. halyziae, which was only known from a single collection in the Netherlands until we recently collected it in Belgium (Mar.-Apr. 2019). Haelewaters and van Wielink (2016) reported an infected specimen of Halyzia sedecimguttata from nature reserve De Kaaistoep in the Netherlands. In 1997-2015, 476 individuals of H. sedecimguttata were collected on a lighted white sheet and screened for presence of Laboulbeniales, only resulting in one individual (parasite prevalence 0.2%). In Belgium, a population of infected H. sedecimguttata was found at the Meise Botanic Garden. Specimens were collected in spring 2019 while they were leaving their overwintering place-deep cracks in the woodwork of a small forest chapel. Screening of 46 specimens of H. sedecimguttata revealed nine infected ones (parasite prevalence 19.5%). This ladybird species seems to overwinter singly or in small congregations in narrow overwintering places, including in leaf litter, under foliage on stone walls, on trunks and branches (Majerus and Williams 1989). This congregation behavior is beneficial for transmission of the fungus and is also observed in Harmonia axyridis (Haelewaters et al. 2017).
Morphologically, H. halyziae is very similar to what we have thus far accepted as H. virescens. Within the Kingdom Fungi, there is an incredible diversity that cannot be perceived through morphology. Cryptic species are being uncovered in Agaricomycetes (e.g., Stefani et al. 2014;Sánchez-García et al. 2016), Lecanoromycetes (e.g., Singh et al. 2015), Leotiomyces (e.g., Grünig et al. 2008), Pucciniomycetes (Bennett et al. 2011), Ustilaginomycetes (e.g., Li et al. 2017), and other major clades. And while the Laboulbeniales has been the subject of a large-scale study to estimate the global species richness of the group (Weir and Hammond 1997), cryptic diversity was not part of the equation. In other words, the number of estimated species of Laboulbeniales, between 15,000 and 75,000, is likely to be corrected to include cryptic species. We note that the recognition of H. halyziae is only possible through molecular data and host associa-tion. Our current understanding is that, within this species complex, there is a strict parasite-host association, with one parasite found only on one host. We think that this host specificity exists at the genus level, given the Adalia clade (Figure 1), which includes isolates from thalli removed from two host species within the same genus. Diagnosis. Morphologically similar to Laboulbenia vulgaris Peyr., but the insertion cell is attached to the lower fifth of the posterior margin of the perithecial wall and the outer appendage is composed of 4-6(-8) branches resulting from successive dichotomies starting at the suprabasal cell, which is poorly pigmented or nearly hyaline.  Carabidae), ADK6448 (BR), slide BR5020212163329V (1 mature thallus, prothorax). Isotypes: ibid., slides BR5020212162292V (2 mature thalli, right mesofemur), BR5020212161264V (6 mature thalli, right protibia), BR5020212166412V (5 immature thalli, mesothorax), BR5020212165385V (1 mature thallus, right protibia), and BR5020212164357V (1 mature thallus, right mesofemur). Paratype: Belgium, Province Vlaams-Brabant, Meise, Domein van Bouchout, 50.92745N, 4.323917E, 32 m a.s.l., 30 Apr. 2020, leg. A. De Kesel, rivulet-associated grassland, on B. (P.) biguttatum, ADK6523 (BR), slide BR5020195033527V (2 mature thalli, mesosternum).

Laboulbenia quarantenae
Etymology. From quarantena, which was used in 14 th -15 th century Venetian language for a forty-day isolation period. The new species was described during the 2020 quarantine period imposed to curb the spread of the COVID-19 virus.
Description. Thallus 300-465 µm long from foot to perithecial tip; colored hyaline at the lower receptacular cells (I and II) and the inner appendage, otherwise pigmented light to dark brown; especially the upper receptacular cells (III, IV and V), cell VI, and the perithecium darkening with age. Cell I elongated, usually straight, 56-107 × 22-33 µm; sometimes bent and then wider at the upper end. Cell II slender, mostly with parallel margins, longer than cell I, 73-160 × 29-40 µm, anterior margin shorter than posterior. Cells III and VI side by side, with septum II-III always much shorter than septum II-VI. Cell III with a narrow base, 29-43 µm long, widening upwards and then 22-29 µm wide at the apex. Cell VI more or less rectangu- from Bembidion biguttatum, specimen ADK6448: A mature thallus from prothorax, slide BR5020212163329V, holotype B mature thallus from prothorax with less pigmented perithecium C mature thallus from the right mesofemur D-F mature thalli from the right protibia G immature thallus from the prothorax H mature thallus from the right mesofemur I ascospores J-K laboulbenia vulgaris Peyr: J mature thallus from prothorax of Bembidion tetracolum, specimen ADK5557 K mature thallus from mesothorax of Ocys harpaloides, specimen ADK6353. One of the diagnostic characteristics of the new species-the positioning of the insertion cell-is shown in a mature thallus of L. quarantenae (E) and one of L. vulgaris (J). Scale bar: 100 µm. lar, 30-34 × 23-30 µm. Cell IV more or less rectangular, slightly broader than long, 20-32 × 25-30 µm. Cell V small, triangular, situated in the inner-upper corner of cell IV, 9-14 × 7-14 µm, as pigmented as surrounding cells. Insertion cell brownish black, flattened, barely marking a constriction on the posterior margin of the thallus, attached to the lower fifth of the posterior margin of the perithecial wall, 18-25 µm wide and 90-128 µm from the perithecial tip. Inner appendage hyaline, composed of 2-4(-6) short branches, rarely exceeding the perithecial tip, 88-150 µm long, resulting from successive dichotomies starting at the basal cell, the latter 9-14 × 6-12 µm. Antheridia short, flask-shaped, few in number, usually on the young inner appendage and arising laterally from its suprabasal cell. Outer appendage up to 250-335 µm long, extending beyond the perithecial tip, often entirely light brown, composed of 4-6(-8) branches, resulting from successive dichotomies starting at the suprabasal cell; the basal cell longer than broad, 23-32 × 15-21 µm, almost entirely hyaline. Perithecium ellipsoid, venter only very slightly asymmetrical, anterior and posterior margins almost equally convex, 109-157 × 43-64 µm, length/width ratio 1.9-2.5, widest in the middle; perithecial tip asymmetrical, with prominent and rounded posterior margin; preostiolar spots black, in older thalli merging into a pre-apical ring, always with distinctly paler zone under the posterior spot. Ascospores two-celled, hyaline, 59-65 × 4. Notes. Morphologically, L. quarantenae mostly resembles L. vulgaris Peyr., but it differs from it by the very low position of the insertion cell (regardless of the origin of the thallus), the successive dichotomous branching of the outer appendage, the poorly pigmented to nearly hyaline basal cell of the outer appendage, and the slender habitus. Although these characters may vary to some extent, eventually resulting in specimens that are morphologically close to L. vulgaris, our LSU phylogeny ( Figure 2) shows that sequences of typical L. vulgaris obtained from Carabidae known to host L. vulgaris-Bembidion tetracolum Say, 1823 and Ocys harpaloides (Audinet-Serville, 1821) (Santamaría et al. 1991;Majewski 1994;Haelewaters et al. 2019a;De Kesel et al. 2020)-fall in a monophyletic clade separated from L. quarantenae. The two isolates of L. vulgaris from B. tetracolum were collected in Belgium (isolate E10T2) and Latvia (isolate E11T6), from populations that are 1,550 km apart, but they were placed together among isolates from O. harpaloides (all from Belgium). Laboulbenia quarantenae, on the other hand, was collected between <1 and 21 km distance from where hosts of L. vulgaris were collected.
Phylogenetically, L. quarantenae may be more closely related to L. flagellata than to L. vulgaris. Laboulb quarantenae and L. flagellata (sensu lato) were retrieved as sister taxa in our phylogeny, although no statistical support was retrieved for this sister relationship. Whereas species boundaries are evident based on our phylogeny, it goes with-out saying that both taxon sampling and sequence data need to be greatly expanded upon to resolve relationships among species of Laboulbenia. The new species is apparently very rare and was never found in combination with L. vulgaris, the more common parasite from Bembidion biguttatum in Belgium (De Kesel 1998;De Kesel et al. 2020).
In Europe, many species of Laboulbenia have been reported on Bembidion Latreille, 1802 (Santamaría et al. 1991). Of those, L. pedicellata Thaxt. and L. vulgaris Peyr. are among the most reported ones. Bembidion biguttatum belongs to subfamily Trechinae. To our knowledge, this species is infected by either L. murmanica Huldén (S. Santamaría pers. comm.), L. pedicellata (Scheloske 1969;Majewski 1994), or L. vulgaris (Majewski 1994;De Kesel et al. 2020). Based on the position of its insertion cell as well as the morphology of both the outer appendage and the androstichum (cells II, IV, and V), L. quarantenae is fundamentally different from these three species. The outer appendage of L. quarantenae is reminiscent of the one from L. flagellata, which, however, is a more robust species reported from 80 genera of Carabidae belonging to Anthiinae, Brachininae, Elaphrinae, Harpalinae, Loricerinae, Nebriinae, and Patrobinae (but not Trechinae) (Santamaría et al. 1991;Santamaría 1998;Haelewaters et al. 2019a).
Bembidion biguttatum, the host for L. quarantenae, belongs to the subgenus Philochtus. Representatives of Laboulbenia reported from Bembidion subgenus Philochtus are few and include two species only: L. pedicellata and L. vulgaris. Two thalli of Laboulbenia "sp. similar to L. vulgaris" from Bembidion bruxellense Wesmael, 1835[as B. rupestre (Linnaeus, 1767 are illustrated in Majewski (1994: Pl. 53, Figs 1, 2). Their morphology comes close to L. quarantenae but cell V is much larger and the insertion cell is not situated low enough along the posterior margin of the perithecial wall. Also L. parvula is reported on subgenus Philochtus in Santamaría et al. (1991), but this species is much smaller (180-190 µm total length) compared to L. quarantenae, it has a deeply pigmented basal cell of the outer appendage, the inner and outer appendage each carry 4-8 very slender branches, and its perithecial tip is rather squarish.
As we explore patterns of speciation of taxa in both Herpomycetales and Laboulbeniales using integrative taxonomy, we can start linking some of these patterns to morphological or life history traits. One candidate trait is the haustorium-a rhizoidal structure that penetrates the host's integument to make contact with the haemocoel, increasing surface area for nutrient uptake and providing holdfast. We hypothesize that -due to the invasive nature of their haustorium -Herpomycetales and haustorial Laboulbeniales, such as species of Hesperomyces, maintain close interactions with their hosts, possibly involving adaptations to the hosts' defense systems and leading to escape-and-radiate coevolution (Ehrlich and Raven 1964). These developments result in an evolutionary arms race, with specialization and leading to speciation (One Host One Parasite model, Figure 1). While all 27 species of Herpomyces form multiple haustoria, not all Laboulbeniales penetrate their host. Recently, Tragust et al. (2016) presented evidence for four species of Laboulbeniales to be superficially attached to their host, and also L. flagellata and L. vulgaris do not seem to perforate their hosts. There are no strict developmental barriers for non-penetrating species and their ascospores may develop on multiple arthropods given that they co-occur in a given microhabitat, resulting in parasite species with more than one host (e.g., L. vulgaris in Figure 2), in contrast to the host-specific species of Hesperomyces. Undoubtedly, other factors come into play; more studies of speciation and species limits, specificity, host shifting, and transmission patterns are needed to test said hypothesis.

New species and new records
In this paper, we describe two new species of Laboulbeniales based on the combination of molecular data, morphology, and ecology (host association). These are Hesperomyces halyziae on Halyzia sedecimguttata in Belgium and the Netherlands, and Laboulbenia quarantenae on Bembidion biguttatum in Belgium. Additionally, Laboulbenia aubryi and Rhachomyces spinosus are newly reported from Belgium. Seven previously described species of Laboulbeniales are reported for the first time from the Netherlands: Chitonomyces melanurus, Euphoriomyces agathidii, Laboulbenia fasciculata, Laboulbenia metableti, Laboulbenia pseudomasei, Rhachomyces canariensis, and Stigmatomyces hydrelliae.
The report of L. aubryi from Belgium is only the third one from Europe. Laboulbenia aubryi was thus far only recorded from India, Nepal, Poland, and Spain (type). Reported hosts are species in Amara Bonelli, 1810 (= Bradytus Stephens, 1827, = Leironotus Ganglbauer, 1892) (Santamaría et al. 1991;Santamaría 1998;Majewski 1999), a diverse genus that is only exceptionally reported with Laboulbeniales (Santamaría et al. 1991). Scheloske (1969) mentioned L. flagellata on Amara plebeja (Gyllenhal, 1810), but considered it an accidental host ("Zufallswirt"). Moreover, based on its simple outer appendage, L. aubryi can easily be separated from L. flagellata. The closest related species, morphologically speaking, is L. argutoris Cépède & F. Picard, but L. aubryi can be separated from it by the insertion cell that is free from the perithecium wall and by the structure of its inner appendage (Santamaría 1998).
Rhachomyces spinosus was recently described from Spain (Santamaría et al. 2020). The most characteristic feature of this species is the spinous process on the second cell of the primary appendage, absent in similar species R. lavagnei (F. Picard) W. Rossi and R. sciakyi W. Rossi. The reported host for R. spinosus in both Belgium and Spain is Syntomus foveatus (Coleoptera, Carabidae). Rhachomyces lavagnei is found on Microlestes spp. and R. sciakyi on Pseudomesolestes sp. All these hosts are placed in the subtribe Dromiusina (Harpalinae, Lebiini); it is possible that these species of Rhachomyces have a high degree of host specificity, which will only come to light as more material will be collected.
Chitonomyces melanurus is easily recognized from other congeneric species by the apically hooked, dark brown to blackish basal cell of its primary appendage. Nine species of Chitonomyces Peyr. occur in Europe, all of them occupying a specific position of the host integument. Chitonomyces melanurus grows almost exclusively on the upper margin of the left elytron of Laccophilus Leach, 1815 water beetles (Coleoptera, Dytiscidae). It has thus far has been reported in Europe from Austria (type), Belgium, Croatia, Finland, France, Germany, Hungary, Italy, Poland, Spain, Ukraine, United Kingdom; also found in Asia and Africa (Bánhegyi 1960;Huldén 1983;Santamaría et al. 1991;Majewski 1994;De Kesel and Werbrouck 2008;Rossi 2018).
The Dutch report of E. agathidii is found on Agathidium laevigatum, the host species from which the type was described (Maire 1920). Euphoriomyces agathidii is thus far found on members of Agathidium Panzer, 1796, Amphicyllis Erichson, 1845, and Cyrtusa Erichson, 1842 (Coleoptera, Leiodidae) in Bulgaria, Germany, Italy, Morocco (type), Poland, South Korea, Spain, and Sweden (Huldén 1983;Majewski 1994;Lee et al. 2007;Rossi et al. 2018). Our material is consistent with E. agathidii, with two mature perithecia at one side and a third, immature perithecium at the other side of the receptacular axis.
Laboulbenia fasciculata is recognized by the receptacular cell V, which proliferates upwards in a series of 4-8 superposed cells V' gradually decreasing in size. Each of these cells V' gives rise to a small trapezoidal cell that carries an appendage consisting of cells separated by dark and constricted septa. This species is very widespread, with reports across Europe, in Africa, Asia, and North and South America. Hosts are members of Carabidae, often Chlaenius Bonelli, 1810 (subfamily Harpalinae) and Patrobus Dejean, 1821 (subfamily Trechinae), but also several other genera in subfamilies Cicindelinae, Brachininae, Harpalinae, Nebriinae, Omophroninae, Patrobinae, and Trechinae (Santamaría et al. 1991). The reports on Omophron spp. are sometimes considered a form of L. fasciculata but this is not accepted by all (Spegazzini 1914;Majewski 1994;but Santamaría 1998).
The status of L. metableti as a separate species has been disputed. Formally synonymized with L. notiophili by Rossi andSantamaría (2006), De Kesel et al. (2020) reinstated L. metableti as a separate species based on characteristics of the appendage system. This species has a European distribution, with reports in Andorra, Austria, Belgium, Finland, Germany (type), Hungary, Italy, Poland, Russia, and the United Kingdom (reviewed in Rossi and Santamaría 2006). Hosts are species of Syntomus Hope, 1838 (= Metabletus Schmidt-Goebel, 1846) (Coleoptera, Carabidae, Harpalinae, Lebiini). We propose using molecular characters to resolve the debate given the taxonomic confusion of species of Laboulbenia on European hosts in the Lebiini tribe: L. baetica Balazuc, L. blanchardii Cépède, L. cymindicola Speg., L. metableti, L. notiophili, and L. pulchella Speg.
Laboulbenia pseudomasei is recognized by cell V that has an internal convex margin and is separated from the perithecium (Villarreal et al. 2010). Cell V sometimes proliferates into a simple or divided branch that grows upwards between the perithecium and insertion cell (Majewski 1994;Santamaría 1998). Rossi and Weir (1997) illustrated that L. pseudomasei can be morphologically highly variable even on a single host insect. Also in the newly reported material from the Netherlands, L. pseudomasei was variable, with the thallus from the right elytron without proliferation of cell V, and the thallus from the prosternum with proliferation of cell V. The geographic distribution of L. pseudomasei is problematic; many old records are unillustrated and the specimens are not preserved (Rossi and Weir 1997).
Rhachomyces canariensis was described from Tenerife (Thaxter 1900) and has since been reported from several countries in Europe and North Africa, Madeira, and the Canary Islands, always associated with species of Trechus Clairville, 1806 (Coleoptera, Carabidae) (Arndt and Santamaría 2004). Majewski (1994) noted the variability of this species and Tavares (1985) suggested material from large geographic distances to the type locality be segregated into a separate taxon.
The only species of Laboulbeniales found on Hydrellia Robineau-Desvoidy, 1830 flies (Diptera, Ephydridae) is S. hydrelliae. Thaxter (1901) described it from Kittery Point in Maine, USA (Thaxter 1901) and it has since then been reported in Finland, France, Italy, Poland, Portugal, Russia, the United Kingdom (Santamaría and Rossi 1993, Weir and Rossi 1995), and New Zealand (Hughes et al. 2004). The new report from the Netherlands is the first one on the European continent in 25 years. Stigmatomyces hydrelliae is recognized by its straight appendage with sterile basal cell and stout antheridia, the spiralled cell walls of the perithecium, and the rounded perithecial apex with one of the lip cells forming a slender, bluntly pointed projection. Hughes et al. (2004) noted that S. hydrelliae thalli from New Zealand are different in their perithecial wall cells not being spiralled and lacking apical projections at the perithecial apex.

Checklist
The current list of thallus-forming Laboulbeniomycetes from Belgium and the Netherlands includes 140 species. Sixty-three species have been found in both countries. A total of 118 species are found in Belgium, and 85 species in the Netherlands. Of the 140 species in the checklist, 55 have not (yet) been reported from the Netherlands, and 22 species have not (yet) been reported from Belgium. Laboulbeniales research in both Belgium and the Netherlands has also resulted in the discovery of new taxa; over the years, 16 species have been described based on material from Belgium and/or the Netherlands (Table 3). It is remarkable that we keep finding undescribed species in two of the most urbanized countries in the world. The reason for this can be found in the fact that Laboulbeniomycetes are severely understudied; only a handful of researchers work on these fungi. In addition, some of the most recently described species are the result of previously unavailable molecular data, long-term study of humid habitats, and focus on unexplored niches.
Finally, undersampled habitats have been cited repeatedly as one of the main sources to find undescribed fungi (e.g., Blackwell 2011, Hawksworth and Lücking 2017, Wijayawardene et al. 2020. This is especially true for the Laboulbeniomycetes. Sampling from dung, fresh and brackish water, animal nests, caves, carcasses, and rotting plant debris has greatly contributed to discoveries in this field of research, not only adding to numbers of described species but also building on our understanding of the ecology of these minute fungi. For example, Pfliegler et al. (2016) sampled ants and their associates from ant nests and, for the first time since its description (Cavara 1899), R. wasmannii was observed on hosts other than Myrmica, including inquiline mites and a fly larva. A survey of Laboulbeniales from coprophilic beetles on Galloway dung in Belgium resulted in two reports of species that until then had only been found in Poland, thus representing a large geographical range expansion (De Kesel 2010). And signal crayfish traps in nature reserve 'De Kaaistoep' have thus far revealed an undescribed species of Diphymyces (De Kesel and Haelewaters 2019) and more material is awaiting detailed study.