Bullet m S.R.B.E.. K B VE., 138 (2002) 20-28 A contribution to the distribution and ecology of Asilid in the sandy regions of Flanders (Diptera ) with a focus on the paucity in the Flemish coastal dunes

1 1 2 Dries BONTE , Wouter DEKON1NCK & Patrick 0ROOTAERT

1 University of Ghent, Dep. Biology, Lab of Ecology, Zoogeography and Nature Conservation, K.L. Ledeganckstraat 35, B-9000 Ghent (e-mail : [email protected]). 2 Royal Belgian Institute of Natural Sciences, Dep. Entomology, Vautierstraat 29, 1000 Brussels.

Summary

In this contribution we discuss the distribution and ecology of some Asilid flies, captured in sandy regions of Flanders, within the framework of some large invertebrate survey projects conducted at the Royal Belgian Institute of Natural Sciences and the University of Ghcnt. A total of 14 species were found, from which several are extremely rare: Antipalus varipes, Asilus crabroniformis, arthriticus and Pamponerus germanicus. Their distribution and ecology is discussed. We also propose a Red List status for the captured species. Keywords : Asilidae, distribution, phenology, ecology, dunes.

Introduction region. Their results indicated that Asilid flies, despite of their well developed flying capacity, ln our temperate regions, Asilid flies are in are poor colonisers since they were almost their adult life stage active predators, which completely absent from recently cleared forests, catch their prey mainly during flight. Almost all but potentially suitable inner dunes. ln a way our Belgian species are typical for undisturbed they can be seen as the counter actors of areas and can thus be used as bio-indicators in Therevid flies from which is shown that they are semi-natural open and thermophile habitats. typical pioneer species (GROOTAERT et al., 2001). ln , Asilid flies are mainly found in the Pleistocene sandy regions of inner-Flanders, In this contribution we give an overview of the coastal dunes, and the area of Samber-Meuse new records on the distribution of Asilid flies in in the South of the country. VERLINDEN (1982) sandy regions in Flanders. We especially focus could identify six large biogeographical districts, on the paucity of the Asilid fauna in the Flemish based on the collections at the Royal Belgian coastal dunes and differences in their phenology. Institute of Natural sciences. TOMASOVIC (1998) 1 he ecology of some species is finally discussed. summarised the distribution and the evolution of Material and Methods Asilid flies in Belgium and found that most of the species (probably with the exception of The data, presented here could be gathered flava and socius) are de­ within the framework of some large survey pro­ clining in our region. jects, coordinated at the Royal Belgian Institute of Natural Sciences, department Entomology and Although a basic knowledge is present on the the University of Ghent, Lab of Animal Ecology. distribution of Asilid flies in our country, almost Parts of the results are already published by no ecological or applied ecological research has TOMASOVIC & DEKONINCK (2000). Additional & been done till now. Only TOMASOVIC DEKO­ data were obtained from inventory projects in NTNCK (2000) investigated the presence of Asilid Eastern Flanders (projects AMINAL-Eastern flies in the inner dunes of Eastern Flanders in Flanders, evaluation of the invertebrate fauna of function of the ground use. They concluded that some humid inland dunes and on the road verges the species composition could differ strongly in Waasmunster - DEKONJNCK & GROOTAERT, between analogue sandy areas within the same

20 Table I. Overview of the sampled localities, the habitat characterisation, sampling methodology and sampling year.

Number Locality habitat Sampling Year methodolOI!Y 1-7 De Panne coastal dunes (grassland, ma.tTam dunes, white pitfalls 2000 grey dunes, dune slack) 8-9 DcPanne coastal dunes, grassland white pitfalls 2000 10-11 Ghyvelde-Adinkerke coastal dunes, grassland, grey dune white pitfalls 1999 12-14 Ursel grassland white pantraps 2002 15-18 Ileidebos, Moerbeke heathland, grassland white pantraps 2001 19 Serskamp grassland white pantraps 2001 20-21 Molsbergen, Lokeren river dune, grey dunes white pantraps 2002 22 Opgrimbie, Kikbeek heath! and hand catches 2000 23-27 Hageven, Neerpelt heathland white pantraps 2001 28-31 Teut, Zophoven heathland white pantraps 2001 32-37 Road verges E 17, Waasmunster heathland white pantraps 2000 38 Schobbejakshoogte, Brugge heathland, grey dune hand catches 1995 39 Honegem-Reebroek woodland white pantraps 2002 40-42 Stropers, Stekene grassland white pantraps 2002

Table 2. Total number of the captured species per locality.

.,. QC) N "" ~ ...... r- ..... r- ::::: ""~.~ r- 0\ ...... N ~ ~ QC) ~ Species I I I I 0\ I N I I 0\ ,. I ~ ..... N QC) ~ ~ Q s ....0 =~ ~ Q,l ..... QC) Q N trl Q ~ N --g." Locality ...... N N N ~ .,. z= ..5l" ~"'

Antipalus varipes (Meig.) 0 0 0 0 12 0 0 0 0 0 4 0 0 0 2 16 Asilus crabroniformis (L.) 0 0 0 0 0 0 0 0 6 0 0 0 0 0 1 6 atricapilla Meig. 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 1 Meig. 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 1 L. 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 1 Dysmachus trigonus Meig. 22 8 61 1 39 0 5 1 87 36 23 2 0 0 11 285 ruji.barbis (Meig.) 0 0 0 0 0 0 0 0 3 0 0 0 0 0 1 3 cinctus (Fab.) 0 0 0 0 96 1 26 0 287 13 11 0 0 20 7 454 Machimus athriticus Zeller 0 0 0 0 0 0 0 0 5 0 0 0 0 0 1 5 Machimus atricapillus Fallen 1 0 1 6 330 0 1 0 11 25 5 2 0 10 7 392 Machimus cingulatus Fabr. 0 0 0 2 8 0 7 0 182 2 1 2 0 0 8 204 (Leow.) 0 0 0 0 1 0 0 0 0 0 11 0 0 0 2 12 Pamponerus germanicus (L.) 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 Phi/onicus a/biceps (Meig.) 78 0 12 0 57 0 62 0 56 8 0 2 1 3 9 279

2001), the coastal dunes (Project AMINAL­ monitoring of dune management) and Limburg (VLINA-project-evaluation ofNature restoration on former arable fields). All were inventoried with white pan or pitfall traps. Addi­ tional data are included from net sampling in the Kikbeek-area in Limburg and the Schobbejaks­ hoogte in Brugge (ZWAENEPOEL, pers. comm.). An overview of the sampled places and the sampling methodology is given in Table 1. The sites are visualised on Figure 1. For the most common species (Dysmachus Fig. l. Localisation of the different sampled areas. trigonus, Lasiopogon cinctus, Machimus atrica-

21 pillus, M cingulatus and Philonicus a/biceps), munster. The sex-ratio mm/ff was 0.18, indi­ we analysed habitat preference by non-linear cating that especially females were caught. The regression of the species' abundance per plot (if species can be found during the months July and sampled with three white pan-traps or five white August (Fig. 3). pitfall traps, which had in total similar trapping 0 ~------, surface) with the recorded habitat parameters Antipa/us varipes 'distance to woodland, surface of bare sand and 5 vegetation height'. We also calculated the sex­ ratio rnmlff and compared it with an equal sex­ I! 2 ratio via X-tests. .! 3 ~ Finally, we attempt to define the species' Red z list status in Flanders, based on their known distribution, habitat preference and trend (TOMA­ SOVIC, 1998)

Results Period General results Fig. 3. Phenology ofAntipalus varipes in Waasmunster (Road verges El7). A total of 1660 individuals were identified, belonging to 14 species (Table 2). D. trigonus, L. Asilus crabroniformis cinctus, M. atricapillus, M. cingulatus and Philo­ nicus a/biceps were both in abundance and This is the first documented record of this number of localities the most common species. coprophytic species since 1963 (TOMASOVIC. Rare captured species were Asilus crabroni­ 1998). During the recent sampling campaign, this formis, Dioctria cothurnata, Dioctria oelandica, species was only found in the llageven-reserve, a Machimus arthriticus and Pamponerus germa­ nature reserve with large areas of short grass- and nicus. heathland, grazed by Galloway-cattle. A total of 5 females and 1 male were captured in these Ecology and distribution habitats, during July-September (Fig. 4).

3 Alltipalus varipes I As!lus crabronlformis A. varipes was only recorded in the inland dunes in Eastern Flanders, where well-developed hcathland was present (Fig. 2). In Belgium, this species is very rare (TOMASOVIC, 1998). TOMA­ j E SOVIC & DEKONICK (2000) found the species in z" Sint Martens Latem, Wetteren and Moerbeke. Our study reveals the presence of a population along the road verges of the E 17 in Waas- ..,.__,_.,___.,[J • ·+· Period Fig. 4. Phenology ofAsilus crabroniformis in Neerpelt (IIageven).

Dioctria atricapilla, D. cotllumata and D. oelandica Only one individual of each species was

Dilt. T-vk: & o.iloNIKII caught during the different inventories. One male " ""- .,, ...... ,._...... , D. atricapilla was found in Moerbeke in the first half of July (Fig. 5). This is probably the only ---J...... - species which can be found in large numbers Fig. 2. The occurrence ofAntipalus varipes in the sam­ outside the sandy regions of Flanders and the pled areas. ------22 ' ·•• ·

Dtul~oii(.&Oii:o-* o..t-&0....-* .. ... - ~ -­ ...- .--,1 ...... , ... -- --o-...- - o-. """""" Fig. 5. The occurrence of Dioctria atricapilla in the Fig. 6. The occurrence of Dysmachus trigonus in the sampled areas. sampled areas. 30 ,----- (VAN VEEN, 1996). One male of D. Dysmachus trigonus cothurnata was found during the same period in 25 r- femol.. l the Ilageven-reserve. These species are fairly ~ males common on sandy soils, but doesn't prefer typi­ 20 cal oligotrophic vegetation (VAN DER Goor, t! " 1985; TOMASOVIC, 1998). They can both be e 1s found in woodlands, dense grasslands in both z"' 10 humid and dry situations (VAN VEEN, 1996). Because of their leaf-hunting behaviour, these species are probably difficult to survey with pantraps.

Period D. oelandica is the rarest species of the three (TOMASOVIC, 1998), for which the number of Fig. 7a. Phenology of Dysmachus trigonus in the Flemish coastal dunes. localities is strongly declining. A male was found ~ r------, in the first half of July in the Hageven. The Dysmachus tngonus species is typical for wooded heathland areas (VAN DER GOOT, 1985; VAN VEEN, 1996).

Dysmachus trigonus 111 i 20 This is probably the most common Asilid 15 in Flanders, which can be found on open sandy grounds. In the coastal dunes, this species is also 10 very common (Fig. 6) Our analysis of the species' habitat preference indicates that the species doesn't has any prefe­ Perlode rences for height of the grass-vegetation or the amount of bare sand. The absence of the species Fig. 7b. Phenology ofDysmachus trigonus in the inland in a recently optimiscd habitat in Uitbergen dunes of Flanders. (TOMASOVIC & DEKONINK, 2000), indicates the species' larval development. "I he sex-ratio however that the species is a bad coloniser. The mrnlff is 1.28 and not significant!} different from species can be found from the beginning of May one (X2-test, p>0.05). till September. Interesting is that the species occurs a half a month later in the coastal dunes, in comparison with inland data (Fig. 7a,b & This species is also rare in our country and TOMASOVIC & DEKONICK, 2000). The peak in the phenology is also a month later in the coastal declining (TOMASOVIC, 1998). We only caught dunes, indicating a possible effect of lower two males and one female in Neerpelt (Fig. 8) in temperatures during the spring and summer on the second half of June and the first half of July. TOMASOVIC & DEKONINCK (2000) fo und howe-

23 ~ver two populations in Eastern Flanders. The 110 LtJSJOpogon clnctus l species is rare and appears to be bound to open 140 sandy patches in heathland areas. V AN VEEN 120 r.:-::-J ( 1996) also mentions the species from forest 100 ~ edges on sandy soils. 10

eo

40

20 ..__...,.._...... -.- M ~ ~ ~ ~ N ~ ~ " ~ ~

Penod Fig. I 0. The phenology of Lasiopogon cinctus in the inland dunes of Flanders.

120 o,.,.. r-.."~14: a. O.lte~t~t'ICit 1\ fiOIMwt• • 111'•••1'11 Hlrwii•UI r.l

Fig. 8. The occWTence of Eutolmus rufibarbis in the sampled areas.

Lasiopogon cinctus

TOMASOVIC (1998) considered this species as 20 rare in Belgium. Our surveys indicated that this species is however rather common in the inves­ 0 -' 1000 1200 tigated areas, but certainly absent in the Flemish Distance to woodland (m) coastal dunes (Fig. 9). The sex ratio was 1.52 L. cinctus 2 Fig. 11. The abundance of in function of the and significantly different from one (X -test, distance to woodland. p<0.05), indicating that more males were caught. L. cinctus is active during the spring period (Fig. 10). The species is typical for forest edges in sandy regions, since it reacts significantly to the distant to the forest edge (Fig. 11) and to the LC(R~.41) amount of bare sand (Fig. 12). • •

PA(~.77)

10 100 Coverage bare aand (%) Fig. 12. The abundance of L. cinctus (LC) and P. a/bi­ ceps (PA) in function of coverage of bare sand.

l)loUITOiliiNO'I'tc&DIHIONtck Machimus artllriticus 1\ NlfoWIII .A IIN•MI'I -...... ltull"'tt This species has never been found in Flanders. ...l - - Only three records are known from Belgium: two Fig. 9. The occWTence of Lasiopogon cinctus in the from the 19"' century (Brussels and Carlsbourg­ sampled areas. Luxembourg) and one from Liege in 1970. We found four females and one male in a bare sandy sod cut field in the Hageven-rescrve in June and the beginning of July (Fig. 13). In the Nether­ lands the species is also very rare and restricted

24 1.00 Mach1mus arthrrticus Mach1mus atncapillus fomolt•J 1<10 L~C-:1- molu 100 §:"]•

j 10 i E E z" 110 z" 40

<10

0 ----..-- ...--.--_.,_--.111~.... W ~ ~ ~ ~ N ~ ~ ~ ~ ~ Period Period Fig. 13. The phenology of Machimus arthriticus in Fig. 15. The phenology of Machimus atricapillus in Neerpelt (Hageven): inland dunes of Flanders. to the large heathland areas if the Veluwe 01AN Machimus cingulatus VEEN, 1996). M cingu/atus is also widely distributed in Belgium (TOMASOVIC, 1998), but is absent from Machimus atricapillus the coastal dunes (Fig. 16; see also VERLINDEI\, This species is the only member of the genus 1982). The species needs more specific habitat occurring in the coastal dunes, although at very requirements than M. atricapillus : maximal low abundance (Fig. 14). M atricapillus shows abundances appear at low vegetation height (Fig. the same distribution as D. trigonus and is 18). common on open sandy soils, without specific Its sex-ration was not significantly different habitat requirements (no reaction on vegetation from one (1.19). M cingulatus is earlier active height, amount of bare sand, presence of Calluna ' vulgaris). The sex-ratio was 0.74 (significantly different from one), indicating that we, contrary I to V AN VEEN ( 1996), caught more females with ·•· white pan- and pitfall traps. The species can be found from the end of July till the end of September with a peak in the last half of August (Fig. 15). Because of its preference for heath­ lands and dunes, we consider the species to be vulnerable in Flanders. DIU lOIMIO\rlc & ~rW!c.k ...1\ ,..MttcMII\•.. .,. --tl ...... t·­- Fig. 16. The occurrence of Machimus cingulatus in the sampled areas.

N 10 Machlmus clngulatus ·•• · ~ 1---,·-]

20

10 O...... t.- ,.....,..,0¥1t... _ l Dftolwllc_ .0 ...... ~ ~ ~ ~ ~ ~ N ~ ~ ~ N -- Period o- Fig. 14. The occurrence of Machimus atricapillus in Fig. 17. The phenology of Machimus cingulatus in the sampled areas. inland dunes of Flanders. 25 t20 deciduous woodland. Before 1950, the species

tOO was found in Eastern Flanders, but seems now to • be extinct in this province.

!! Philonicus albiceps ! 110 MC ~R'"O.K) 0 JJ" TOMASOVIC ( 1998) classifies this species as " rather common and stable in Flanders. We found the species at nine places. In the coastal dunes, it 20 • is the dominant species (Fig. 20). The sex ratio 0 mm/ff is 1.36, significantly different from one. 0 tO 20 30 The species has specific habitat requirements: its Height veget.ttion (an) abundance increases with the amount of bare Fig. 18. The abundance of M. cingula/us (MC) and P. sand (Fig. 12), but it also needs a moderately a/biceps (PA) in function of the height of the high grassland vegetation (Fig. 18), indicating its vegetation. preference for sandy soils with the presence of grass tussocks. Its phenology is also delayed in than the previous, with a peak in the second half the coastal dunes (Fig. 21 a,b), again indicating of August (Fig. 17). that lower temperatures slow down the larval development. Because of its preference for open Neoitamus cyanurtts (dynamic) sandy soils in Flanders, we consider During our surveys, we only found one male the species as vulnerable in Flanders. in a dry woodland (edges) in Moerbeke and three males and eight females in the wooded heathland of the E17-road verges in Waasmunster in the period half June-July (Fig. 19). This preference for dry deciduous woodlands on sandy soils is confirmed by Dutch data, presented by V AN DER GOOT ( 1985) and V AN VEEN ( 1996). TOMASOVIC (1998) considers this species as common and stable in Belgium.

DIDT.-..-&~-­-.-­ I) nail ...... ·-. ,... .-.; --o ....

Fig. 20. The occurrence of Philonicus a/biceps in the sampled areas.

30 -r------, Phtlontcus a/biceps OIMT-....-&~... 11 -­IMIIIowntl- --• .,.ttfll o- 20 1 Fig. 19. The occunence of Neoitamus cyanurus in the E t5 sampled areas. z" 10 Pamponerus germanicus We captured one male of this species in the 0 ' .'u l' beginning of June in the valley of the Kikbeek, ~ ~ ~ ~ ~ N N ~ ~ ~ N in birch forested heathland. P germanicus is Period strongly declining in Belgium (TOMASOVIC, I 1998) and bound to heathland and edg-es_ o_f_d_ry--F-ig_._21a. The phenology ofPhilonicus albiceps in the L Flemish coastal dunes. 26 lO ,. Ph1/onicus a/biceps Detailed analysis ofthe habitat preference Interesting is the significant reaction towards the amount of bare sandy soil of Lasiopogon cinctus and Philonicus a/biceps. The latter species prefers sandy places with high vegetation tussocks, a typical habitat in the coastal dunes, where it is the only species reaching high abun­ dances. L. cinctus appears to be a typical species of woodland edges on sandy soils, confirming 0 - '-Y--Y--"1.,_..._._. 4-11 !W !WI ~ 1-11 7~ NI Il-l the habitat categorisation of V AN VEEN ( 1996)

Period and V AN DER Goor (1985). Machimus ciHgu­ latus requires the same macrohabitat as M. atri­ Fig. 21 b. The phenology of Philonicus a/biceps in the capillus, but seems more sensitive to grass inland dunes of Flanders. encroachment, since it prefers short vegetation. Conclusion Coastal versus inland dunes Rare species and vulnerability in Flanders In general, we can conclude that the Asilid fauna in the Flemish coastal dunes is poor. In In this contribution we discuss the distribution comparison to the Netherlands, species like Anti­ and ecology of some Asilid Dies, captured in palus varipes, Eutolmus rufibarbis, Lasiopogon sandy regions of Flanders. A total of 14 species cinctus, Asilus crabroniformis and Pamponerus were found. Of these, some extremely rare spe­ germanicus are absent. VERLINDEN (1982) states cies were recaptured since a long period: Anti­ that extinction events due to shrub encroachment palus varipes, Asilus crabroniformis, Machimus without recolonisation probably lay at the basis arthriticus and Pamponerus germanicus. of this low diversity since suitable habitat is Twelve species are highly characteristic for largely present. This is in our opinion only a part sandy soils and can in the future thus be con­ of the explanation and probabl; only true for sidered as Red List species in Flanders because A. crabroniformis, from which old records arc their habitat is rare. An intensive sampling present from the coastal dunes. The retrieval of campaign remains however necessary for the local farmers (and their cattle) after the Second confirmation of their assumed Red List status. As World War has probably fastened this species' mentioned before, four species are extremely rare extinction. The other species all prefer more in Flanders and show a strong declining trend forested dunes. Since this habitat was almost (TOMASOVIC, 1998): A. varipes, A. crabroni­ completely absent till the seventies, suitable formis, M arthriticus and P. germanicus. For habitat was historically absent, so species were these species we propose the status of 'critically never present. Now, dune forest has developed endangered'. Dioctria oelandica and E. rufibar­ but the accompanying species have probabl; not brs arc more common than the former species, been able to colonise these habitats because of show a declining trend, but remain rare in their isolated distribution in inland dunes. The Flanders. For these species, we propose the establishment of dune woodland should thus status of 'endangered'. With the exception of enhance the Asilid richness in the future, if Dioctria atricapilla and D. cothurnata, all the colonization from source populations can occur. other discussed species arc bound to sandy The coastal climate can also influence the pre­ grounds, but can still be found in numerous sites sence of Asilid flies, since certainly the pheno­ in large numbers. Because of their habitat logy of the two common species is retarded with preference (and the rarity of that habitat: approximately one month, indicating lower oligotrophic grasslands, dry deciduous wood­ temperatures during the larval development. fhis lands on sandy soils, dunes) we propose them as factor cannot influence the absence of the 'vulnerable' in our region. The two mentioned previous mentioned species, since they all occur Dioctria-specics are common in widespread in the Holland dunes 01AN VEEN, 1996), where habitats and are probably not threatened in the climate is even colder than along the Flemishj Flanders. coast.

27 Acknowledgement Belgique et des regions limitrophes (9). Rarefac­ tion en occidenta1e et porte disparu en The data were obtained within some conser­ Belgiquc : Asilus crabroniformis LINNE, 1758 vation studies, founded by the Flemish administration (Diptera, Brachycera, Asilidae). BulleUn et An­ responsible for the cnvironrocnt and nature conser­ nales de la Societe royale beige d'Entomologie, vation (A.MINAL). 131(2}: 245-248. TOMASOVIC G., 1998. - Evolution de la faune des Literature Asilidae (Dipteres Brachyccrcs) de Belgique au cours de ce dernier siecle. Notes faunistiques de DEKONTNCK W., BONTE D. & GROOTAERT P. (eds), Gembloux 35: 7-19. 2000. - Onderzoek naar de hcrstclmogelijkheden TOMASOVIC G. & DEKONINCK W., 2000. - Donnees t.b.v het bchoud van de specifieke entomofauna sur la faunc de Asilidae de la Flandre Oricntale - van de landduinen in Oost-Vlaanderen. Rapport Notes faunistiques de Gembloux 41 : 23-32. Ent.2000.05, onderzoeksopdracht AMINAUNN VAN DER GOOT V.S., 1985.- De Snavelvliegen (Rha­ 98, Urtiversiteit Gent & K.B.l.N., 211 pp + giorudae), Roofvliegen (Asilidae) en aanverwante bijlagen. families van Noordwest-Europa. Wetenschap­ DEKONINCK W. & GROOTAERT P., 2001.- Onderzoek pe/ijke Mededelingen van de Konink/ijke Neder­ naar de faunistische waarde van de autosnelweg te landse Natuurhistorische Vereniging, 177, 66 pp. Waasmunster. Rapport ENT.200 1.02, studie­ VAN VEEN M.P., 1996. - De Roofvliegen van opdracht voor het ministerie van de Vlaamse Nedcrland. Wetenschappelijke Mededelingen van Gemeenschap, 115 pp. de Konink/ijke Neder/andse Natuurhistorische GROOTAERT P., DEKONINCK W. & BON"lE D., 2001. - Vereniging, 216: 120 pp. Therevidae (Diptera) in the East-Flemish inland VERLINDEN L., 1982. - The Asilidac (Oiptcra) of dunes, pioneers on sandy soils? - Bulletin van de Belgiwn and their distribution in this country. Belgische Entomo/ogische Vereniging 137: 32-35. Bulletin et Annales de la Societe royale beige TOMASOVIC G., 1995. - Notes sur les Asilidae de d'Entomologie, 119:177-185.

2o