Fera Science Ltd. provides diagnostic support for the plant health service in England and Wales. The Plant Health and Seeds Inspectorate (PHSI) submit samples that they supect may be regulated or non-native to Fera for identification. Apart from gall midges (Cecidomyiidae) and darkwing fungus gnats (Sciaridae) (common pests in greenhouses), other members of the Sciaroidea, like the true fungus gnats (Mycetophilidae) and predatory fungus gnats (Keroplatidae) (such as Proceroplatus trinidadensis, Lane 1960, Fig. 1B), are rare interceptions. The two main reasons for this are: 1) There are no regulated species of fungus gnats and 2) fungus gnats are generally sporophagous or mycetophagous as the name suggests and, thus, PHSI inspectors are less likely to submit them. This assumption, while generally true, leads to specimens only being identified as "Mycetophilidae" or "Keroplatidae" or specimens are frequently left at generic level. In the last 10 years, Sciophila corlutea Chandler & Blasco-Zumeta, 2001 (adventive on Prunus persica ex. Spain, Fig. 1C) has been intercepted, as well as an unidentifiable female of a Leia species on Zingiber ex. China (Fig. 1A).

Figure 1.  

Very few gnats are intercepted on produce into and out of the United Kingdom with few also detected on plants grown in nurseries. Above are examples of some of the species that have been confirmed. A Leia sp. intercepted on Zingiber from China. B Proceroplatus trinidadensis Lane, 1960 intercepted on Monstera deliciosa that originated from the Netherlands. C Sciophila corlutea intercepted on Prunus persica from Spain (likely adventitious).

Aside from these, one species that has been encountered more than others in recent times is the Neotropical fungus gnat Sciophila fractinervis (Edwards, 1940) (Fig. 2A, B), described from the south of Brazil and not currently recorded elsewhere in South America (Chandler 2022).

Figure 2.  

Adult specimens of Sciophila fractinervis. A Gravid female Sciophila fractinervis (atypical form lacking R₂₊₃ vein, further reinforcing the instability of this character); B A typical male Sciophila fractinervis (R₂₊₃ vein present creating a radial cell).

The species was first recorded in Britain over 10 years ago by Peter Chandler from specimens recovered from a glasshouse in Warwickshire and apparently associated with Lisianthus (Eustoma grandiflorum) and potentially Christmas cacti (Epiphyllum) (Chandler 2010). Interestingly, around the same time as Chandler’s confirmation, Fera Science Ltd. received two adults and two larvae of suspected S. fractinervis from Warwickshire, except only a single adult female was reared with an overall habitus resembling that of S. fractinervis. At the time, it was noted that the individual resembled “the British species S. interrupta in the form of the posterior fork on the wing” (basal abbreviation of the posterior fork, CuA1 Fig. 2A, B), a character also shared with S. cincticornis), but was “too light in colour to be that species”. Without a male at the time, confirmation was impossible. Since then, the status of S. fractinervis in the U.K. has remained largely a mystery and its biology enigmatic. A recent increase in interceptions led to the writing of this article.

1. In November 2020, the first gnat specimens started arriving in the labs. Reports of insect infestation was reported in glasshouses in Chichester, West Sussex in apparent association with a proportion of (ca. 500) Ficus benjamina, Ficus elastica and Dracaena reflexa var. angustifolia (Syn. D. marginata) plants with some plants originating from Sri Lanka and others from the Netherlands. Initial suspicions by growers and inspectors were of “Orchid worm” (various Keroplatidae known to be an issue in greenhouses in the Netherlands, such as Lyprauta spp., Proceroplatus trinidadensis) or an unidentified tortrix moth due to webbing networks erected at the base and further up the plants, especially near the leaf axils. The PHSI (Plant Health and Seed Inspectorate) submitted symptomatic plants. Dracaena plants arrived to the labs with a thin, filamentous layer of webbing strewn across the potting medium’s surface at the base of the plants, as well as the lower stem and leaf axils (Fig. 3). To the casual eye, the webbing resembled saprophytic fungi or actinomyces bacteria that sometimes colonise the surface of potting media in greenhouses. A Nematoceran larva (Fig. 4A) was seen moving about on webbing at the base of the plant and a further pupa was suspended in webbing higher up in the plant (Fig. 4B).
2. Two female gnat specimens arrived in January 2021 from glasshouses in East Riding of Yorkshire involving ca. 928 plants with an estimate of 25% of plants affected. The crop consisted of mixed herbs (Rosmarinus officinalis, Thymus vulgaris and Origanum vulgare) originating from Italy. The specimens were mixed in amongst frequently detected glasshouse insects, such as Ligurian leafhoppers (Eupteryx decemnotata Rey, 1891), Onion thrips (Thrips tabaci Lindeman, 1889), dark-wing fungus gnats (Bradysia spp.), as well as various Collembola and predatory Acarina.
3. A third interesting interception on 8 May 2021 involved numerous gnats, both males and females, being found on sticky traps in a nursery in Preston, Lancashire, U.K. (Fig. 5). The sticky traps were being used to monitor pests in and around a Fragaria crop.
4. A fourth finding on 25 May 2022 was of a Nematoceran pupa adhering to a leaf of an Impatiens New Guinea hybrid from a nursery in Middlesbrough, North Yorkshire, U.K.
5. The final finding was on 19 October, 2022 where several larvae and a live adult were found in association with Chrysanthemum from the Netherlands where larvae resided under lower leaves and axils of the plant that were draped over the soil.

Figure 3.  

Showing the fine, filamentous webbing produced from labial glands of Sciophila fractinervis larvae in and around Dracaena plants from Sri Lanka via the Netherlands.

Figure 4.  

The immature lifestages of Sciophila fractinervis. A The larva of Sciophila fractinervis with close-up of cuticular ultrastructure made up of a mesh-like network on a decaying Chrysanthemum leaf. B The suspended pupa of Sciophila fractinervis surrounded by fine, filamentous webbing on Dracaena.

Figure 5.  

Numerous Sciophila fractinervis sampled using sticky traps during standard monitoring for Bemisia tabaci in a Fragaria crop- Preston, Lancashire, U.K.

All gnat specimens were initially examined using a stereomicroscope (Leica 205C) with a Schott KL1500 LCD light source. All live larvae were examined by gently rolling them between the lid of a 150 mm × 15 mm Petri dish and the lid of a 60 mm × 15 mm Petri dish so as to view all characters required to identify the larva. Care was taken to avoid excessive pressure being applied to the specimens. Larvae were provisionally identified using a combination of Madwar (1937), Hutson et al. (1980) and Zaitzev (1982a)

Symptomatic plants (with the larvae, pupae etc.) received were enclosed in sealed containers with breathing holes (sealed with mesh) and then placed in an incubator (240L Sanyo Co2 Incubator MIR-253) subsequent to visual examination where possible so as to rear these specimens to adulthood. In the case of the Middlesbrough finding, the pupa was placed on filter paper, in a Petri dish and sealed.

For adult gnat specimens submitted on sticky traps as in the case of the Preston finding, extraction and cleaning of the glue followed Appendix 1 (Protocol for removal of adult whitefly from sticky traps) in Malumphy et al. (2010). All adult gnats, whether successfully reared or extracted from sticky traps, were then slide-mounted unless they were destined for DNA sequencing. Males were ultimately needed for species determination. Their abdomens were removed and genitalia dissected away from the abdomen with dissecting needles and placed in an embryo dish with a solution of 10% potassium hydroxide (KOH), covered with a glass lid and placed on a hotplate at 80°C for 20 minutes to allow for maceration of the soft body contents. The genitalia were then gently pressed/palpated with a micro-spatula to expel any residual soft tissue. After neutralising with glacial acetic acid (CH₃COOH), the genitalia were immersed in 70% ethanol and further tissue and cuticle were removed exposing the genitalia. Genitalia were then transferred to absolute ethanol for 5 minutes, then clove oil before being mounted on a slide in Canada Balsam with an 11 mm coverslip along with the rest of the specimen (head, wings, legs and thorax). Adults were identified to species (if male) and genus (if female) using a combination of Edwards (1940), Hutson et al. (1980), Chandler (2006), Chandler and Pijnakker (2009) and Søli (2017). Specimens identified morphologically were deposited in the FERA plant health entomological reference collection.

Several strands of silk laid down by gnat larvae in the Dutch Chrysanthemum sample were examined for the presence of fungi. Strands were plated up on sterilised Petri dishes of PDA (Potato Dextrose Agar), incubated and identified on the basis of cultural, microscopic and morphological characteristics if possible.

Molecular methods

Two adult female gnats originating from East Riding of Yorkshire were sequenced for the COI DNA barcode (Hebert et al. 2003), as females cannot be identified morphologically and also due to a clear difference in wing morphology between the specimens. The hind legs of each specimen were removed, placed in 1.5 ml Eppendorf containers and stored in a freezer at -18°C. DNA was extracted from each leg separately (4 samples in total, from 2 individuals) using the QIAGEN Blood and Tissue Kit, following the manufacturers' recommended protocol. Samples were amplified by PCR using the primers C1-JF-1718 and C1-NR-2191 (Simon et al. 1994) and MiFi Mix (Bioline, UK) PCR reagent master mix. PCR amplicons were cleaned up using ExoSAP-IT Express (ThermoFisher) and sequenced using the PCR primers by Sanger sequencing at Eurofins Genomics, Ebersberg, Germany.

Any fungi not identified using morphological means were identified via DNA sequencing in the Btub and ITS gene regions.

Larvae examined from Chichester, West Sussex exhibited all the characteristics of a Sciophila spp. larva. Under-developed antennae, well-developed maxillary palps, peripneustic in terms of spiracular layout and locomotory hooks being visible on the ventral surface of the body (presumably used to adhere to the webbing upon which it moved). It took 10 days for a single male fungus gnat to be reared from specimens on plants stored in the incubator and this was determined as Sciophila fractinervis (Edwards, 1940), a Neotropical species that is established in nurseries in the Netherlands (Chandler and Pijnakker 2009).

Specimens from East Riding of Yorkshire (both female) were confirmed as Sciophila spp. and one close to S. fractinervis with the other close to S. cincticornis (Edwards, 1940). The suspected S. cincticornis generally followed the description by Edwards (1940) and Chandler and Pijnakker (2009) with vein R₄ (now to be interpreted as R₂₊₃ following Søli (2017)) absent, flagellomeres slightly yellow on basal two fifths to half and abbreviated anterior branch of the posterior fork (CuA1) (Fig. 2A, B). The other specimen was more typical of Sciophila fractinervis with the R₂₊₃ vein present, creating a radial cell. All other instances of suspect S. fractinervis were confirmed morphologically. Several fungi were present on the isolation plates cultured from strands created by larvae intercepted amongst Chrysanthemum from the Netherlands. The majority were saprophytic with an unknown Acremonium- like fungus being the most dominant. Alternaria, Penicillium and a Mucor species were also present.

Molecular results

COI DNA sequences were generated for the two reference samples, with two independent sequences generated per sample. Final sequence lengths were between 444 and 457 base pairs long. Both specimens had the same haplotype (i.e. they share the same DNA sequence), indicating they were likely the same species. There were no COI reference sequences for Sciophila cincticornis or Sciophila fractinervis on either the BOLD (Ratnasingham and Hebert 2007) or GenBank (Benson et al. 2012) public databases. A BLAST search against the GenBank nucleotide database found close matches (up to 99% pairwise similarity, closest match MG104750.1) to Sciophila sequences that were not identified to species. Similarly, a search against the BOLD database found close matches (up to 100% pairwise similarity) to Sciophila sequences not identified to species. These were in BOLD BIN BOLD:ABV9018 (Ratnasingham and Hebert 2013) and it seems likey that this BIN either corresponds to the species S. fractinervis or contains it. Sequences will be uploaded to GenBank and BOLD [accessions to follow, contact author for details]. The unknown "Acremonium-like fungus" was sequenced and identified in the Btub region but only to genus level: a Plectosphaerella species. In ITS, one of the cultures matched 100% to two Plectosphaerella species (P. pauciseptata and P. cucumerina); the other culture also matched 100% to two species – P. plurivora and P. niemeijerarum. Unfortunately the sequencing results from these gene regions did not allow us to distinguish them further.

Not a great deal is known about the biology of the Sciophila genus nor the larval diet (Kurina 2020). The larvae of most species tend to live on, within or on the underside of the sporophore/fruiting bodies of mainly wood-associated or lignicolous fungus species where they construct webs and feed on spores (Zaitzev 1979, Zaitzev 1982a, Zaitzev 1982b Zaitzev 1982b, Falk and Chandler 2005, Chandler 2006, Ševčík 2010, Bouchard and Bouchard-Madrelle 2010, Jakovlev 2011). They may also be found in association with fungal mycelia (Zaitzev 1979, Zaitzev 1982a, Zaitzev 1982b, Chandler and Pijnakker 2009) particularly that which is found in association with deadwood (Zaitzev 1979, Zaitzev 1982a, Bechev and Koç 2006). The larvae of S. fractinervis, as in other members of the Sciophilinae, tend to be enclosed in a mucous tube or “delicate tube of mucilage” created from labial glands around the mouth (Madwar 1937, Zaitzev 1982b) which we observed often giving larvae a shiny appearance (Fig. 4A). Coupled with this, the cuticular ultrastructure of larvae appeared to have a fine mesh/reticulate network. This network likely acts as a plastron of sorts that, amongst other things, aids respiration.

Growers, in many of the instances where S. fractinervis was found, noticed webbing forming on the compost surface (Figs 3, 6) even prior to planting. It is apparent that webbing produced from labial glands situated in the heads of larvae of S. fractinervis is of great importance to their life-history. The scaffolding-like webs appear to help the larvae pupate in a dry place, while also acting as a potential barrier protecting against predation (Fig. 4B). Any sudden vibrations on the webs or "predator-like" movements towards larvae of S. fractinervis in the lab, elicited a rapid retreat response. Potting media used in horticulture are frequently enriched with microbial biostimulants that can promote mycorrhizal fungus. It is likely that the larvae of S. fractinervis potentially feed on spores stuck to their webs that originate from the potting medium itself below the webs, but also fungi residing on the plants themselves. It is evident that the surface of webs become covered in airborne spores in glasshouses, generally saprophytic species. Some Plectosphaerella species are known plant pathogens (P. cucumerina). Many are associated with soil and plant debris. There are also species isolated from living plants. Further gut analysis of larvae is needed to ascertain whether larvae are opportunists or select certain species.

Figure 6.  

Neil Helyer (Fargro Ltd.) lifting the leaf of a Primula obconica plant in a nursery in Ireland to reveal a "Sciophila-like" larva on the underside of the plant. Note the webbing again near the base of the plant.

Other members of the Sciophilinae have been known to use webbing networks to effectively snare and feed on smaller invertebrates, but such feeding behaviour was not observed here or in the account by Chandler (2010). As far as the adults are concerned in terms of diagnostics, it would appear that the presence of R₂₊₃ in Sciophila fractinervis may be an unstable character like other Sciophila spp. in the world as Chandler and Pijnakker (2009) suggest.

To conclude, it appears that S. fractinervis is here to stay in Britain with interceptions and submissions to the lab on the rise as those in industry recognise symptoms of presence, larvae and the adults. More work ought to be carried out to ascertain the larval diet in an ex-situ context; however, saprophytic fungi likely sustain the larvae in a horticultural setting.

It is uncertain as to where the above-documented occurrences of the species originated. It is most likely the Netherlands, but in some instances, plants also originated from Costa Rica and Denmark. It is likely that this enigmatic species is more widespread than first realised especially in Europe. There is evidence that it is in the Republic of Ireland. Photos (Fig. 6) were submitted at the end of 2020 to the corresponding author of a "Sciophila-like" larva and accompanying webbing on Primula obconica from a nursery between Cork and Dublin with anecdotal evidence of similar larvae being found on Poinsettia the previous year. Unfortunately from photos, species determination was not possible.

S. fractinervis does not appear to be doing any damage to the great many plant species it has been associated with under protection so far as we currently know. The webbing is viewed by the industry as "unsightly" and whether the webbing itself is facilitating any damage remains to be seen. In terms of control, Decis Protech (Bayer Crop Science UK) a deltamethrin-based insecticide has been shown to be effective at combatting older generations of S. fractinervis in glasshouses, but future generations appear to recolonise shortly after (pers. comm. Andrew Gaunt, PHSI/APHA). Alternatively, the Staphylinid biological control agent Atheta spp. which is effective against Ephydrids and Sciarids has been shown to be very effective (pers. comm. Neil Helyer, Fargro Ltd.).

This work was supported by the UK Government's Department of Environment, Food and Rural Affairs (Defra) under the Defra-Fera long term services agreement.

The authors would first like to thank Andrew Gaunt (PHSI/APHA), Neil Helyer (Fargro Ltd), Daryl Nightingale (PHSI/APHA) and Luke Lloyd for all the helpful information, photos and for submitting samples to the Fera entomology labs. Thank you to Aiga Ozilina (FERA) for isolating and identifying fungi and Dr. Chris Malumphy (FERA) for helpful edits and changes to the text. Thanks to all the FERA entomology team for the insightful discussions (Adele, Gemma, Sharon, Eleanor, Joe, Chris, Charlie, Rowan, Noel, James). Finally, thank you to several anonymous Bibionomorpha reviewers who provided invaluable suggestions and edits.

No potential conflict of interest was reported by the authors.