CPP ALES EN PU OT EES MERITARE i bane eed 4 ti KAREL di iI PAARL ith Ay utd il HU Hl UNE san 4 ih JA N (ll defi byt a CEE PAR tet DU A Bu 4 DICH NEE PER HERDER AN) ATEN il JUNE Hunt ACL ih N ln hi DAS i PIATTI ite MAN ll È i pou all AO eR i Hin I I! n 4] iit EMI, ihn LU EUX ance ray ii nah fl i Bu Sa + à hi di Be) N 14). I EHE aar da ss iy ni Sol fi bot} I hr Iran a un anal) nn Ni a M RON MAGLIA DA Ty en i eier MAR TETE dati md Visita tl | AI i hi i} ja H bj ti 1h mi à Mo ih N sal HA n TOGO Ke UNI ait 4 tonte { wig vi RCH AU pia Qi, (RE HI Li i ni ded sib satya fe ATH À sazia bed 141 40) ROMANE ive Sepa ke |) NN Ba handen fand deinde ka ds an abu Hi Hath HECHI it 8 a in Il thats UE più, HA + Ap sie Ad Ver NT NH N hel 168108) hie MENTA pen eha y al aan hl 04 dida ; Wal dont bed i ABATE 4) tee CHANT] ï RM IM WP TOA TOL RAD SI FA #20 di LR im shad bal qu MAN att te tes Al baits iba D> HAT | KEES EINE Rp 4 dort El nua DIGI INIT ped: AH Ent ton lt be ee Dt NA ts ff L DES tad, LEE i Saige (u N RME) qui Hy sir PERONI AH Art A vr iin bl M TUE OE pin GRAN PROTRAE und rhin | bi de Mij ha er ipa Ce Ph HET TOTI hj EURE ER ENEN HE ibs Dea befana MENTO Ma tee an PATENT RATE LOT EN ANNEE 149 Vite) } RIRE band mdod hen LI i fas FAHRERN. RENTE ANA aire) VRE RULLI IE bog bn ke FL SEN A Ws OMAN STATE À Might DE DANSE Sands HA hehe be EN MOTTE TEE an isha be aang. asa NE HUE PAUS MI path î AES Neb ave hebbend ien he IE TRE À TI (TER RENEE Ue Pre, | HR MEAN be WY SO Bal bat 47 (LEE DEST 7) PR EN ides etende hade Tatas TANT AN BENIN AG ih pani e, te basique ES in LEHE REN, wahr PRECI A FA ARTS 4 5. PUERTA VET i Meer E with À 4} aia} Arm Hinz! LADD pe neg Cate) AR Dé lu Li HERNE ; (LIN LMS Hecate ahi PAIE ‘ Kian nan ANA Fish os dje 4 sgr MAP CRT # ù #93, hr A Tele siPtiri baie a ih (54 va md ni W HER ERLE e FM ty er fr 14 Hr ni ia HO Fi ke 192 È È tit qe u i da da tint 4 ul, Hu, bi MR MIE) + DER ARRETE A 1:45} en Do re Ger Sun 1 Ri nd A dedi he 8 Hit hs yea. H ‘hw Went rks be fh VEN dt fal i jee ig) ‘ ER, (ea ie! zut 7 HARVARD UNIVERSITY © Library of the Museum of Comparative Zoology \ S : : Li i n ‘i x hi ee + È; ì à Pe - hia +e t'es HEA Log PA gel dy ot) tay Ata ETS LT mgmt gd Nar 1976 DEEL 119 MUS. COMP, ZOOL: i rn LIBRARY id MAR 2/3 1977 HARVARD SI, DSCHRIFI VOOR ENTOMOLOGIE UITGEGEVEN DOOR DE NEDERLANDSE ENTOMOLOGISCHE VERENIGING Tijdschrift voor Entomologie, deel 119, 1976 NEDERLANDSE ENTOMOLOGISCHE VERENIGING BESTUUR (BOARD) Voorzitteri(Chaisman) an se eee J.T. Wiebes Vice-Voorzitter (Vice-President) ........ Th. van Leeuwen Secretanis\(Secretaty)i) 2. on en R. de Jong AAL A RIA EOLIE Rijksmuseum van Natuurlijke Historie, Raamsteeg 2, Leiden Penningmeester (Treasurer) ........... H. Wiering AGES TIR AA, Doorntjes 29, Bergen Bibliothecaris (Librarian) ............. W.N. Ellis Plantage Middenlaan 64, Amsterdam Leden\(Members)- … ..3 564006 2c ee ne A. K. Minks, H. van Oorschot TIJDSCHRIFT VOOR ENTOMOLOGIE Redactie (Editorial Board) ............ P.J. van Helsdingen, R. de Jong, J. Krikken, M.A. Lieftinck Addresse ere PA A een ee Rijksmuseum van Natuurlijke Historie, Raamsteeg 2, Leiden The journal serves the publication of papers on Insecta, Myriapoda and Arachnoidea. Subscription rate: D.Fl. 150.— per year. Afleveringen 1-3 verschenen 7.VII.1976 Afleveringen 4-7 verschenen 25.X.1976 Afleveringen 8-9 verschenen 29.X11.1976 ERRATUM a P. 219: 2nd line of summary, “with (hippophorbiae x euphorbiae) 9.” 9° should read “with (hippophorbiae x hippophaes) ©. INHOUD VAN DEEL 119 Achterberg, C. van. — A preliminary key to the subfamilies of the Braconidae (Hymenoptera) Atyeo, W. T., & P. C. Peterson. — The feather mite genera Zumptia Gaud & Mouchet and Para- zumptia gen. nov. (Acarina, Analgoidea) ........................ Blokhuis, B., zie Heerdt, P. F. van. Ellis, W. N. — Autumn fauna of Collembola from Central Crete .................. Haaften, C. van, zie Heerdt, P. F. van. Heerdt, P. F. van, B. Blokhuis & C. van Haaften. — The reproductive cycle and age composi- tion of a population of Pterostichus oblongopunctatus (Fabricius) in the Netherlands (@oleopteraa@arabidae) ern. A one Jong, R. de. — Affinities between the West Palaearctic and Ethiopian butterfly faunas ..... Kruseman, G., zie Willemse, F. Lith, J. P. van. — New species and records of Indo-Australian Psenini (Hymenoptera, Spheci- dacsPemphredoninde) arene nl) Tee coc) I ee a eee Loeliger, E. A. — The enigma of Celerio hybr. Pauli Mory ...................... Meeuse, A. D. J., zie Stelleman, P. Peterson, P. C., zie Atyeo, W.T. Stelleman, P., & A. D. J. Meeuse. — Anthecological relations between reputedly anemophilous flowers and syrphid flies. I. The possible role of syrphid flies as pollinators of Plan- COLORE td SOTA AO ANTON Willemse, F., & G. Kruseman. — Orthopteroidea of Crete ...................... DEE ooo ROIO ORE CORRI We 79 217 15 123 337 DEEL 119 AFLEVERING 1 1976 PP 7569.2, TIJDSCHRIFT VOOR ENTOMOLOGIE MUS. cori a UITGEGEVEN DOOR LIBRA re ve \AUG 1 DE NEDERLANDSE ENTOMOLOGISCHE VERENIGINGY AF > È UNIVERSIT INHOUD _P. F. van HEERDT, B. BLOKHUIS and CAROLINE VAN HAAFTEN. — The reproductive cycle and age composition of a population of Prerostichus oblongopunctatus (Fabricius) in the Netherlands (Coleoptera: Carabidae), p. 1—13, fig. 1—4, pl. 1—2. | Tijdschrift voor Entomologie, deel 119, afl. 1 Gepubliceerd 7-VII-1976 THE REPRODUCTIVE CYCLE AND AGE COMPOSITION OF A POPULATION OF PTEROSTICHUS OBLONGOPUNCTATUS (FABRICIUS) IN THE NETHERLANDS (COLEOPTERA: CARABIDAE) by P. F. VAN HEERDT, B. BLOKHUIS and CAROLINE VAN HAAFTEN Zoologisch Laboratorium, Rijksuniversiteit, Utrecht With two plates and four text-figures ABSTRACT The reproduction of a Dutch population of Prerostichus oblongopunctatus (F.) is analysed. Seasonal variations in egg numbers, “corpora lutea” sizes, the relative positions of eggs in ovarioles and oviducts, and the conditions of the eggs were studied. Three age classes (corresponding with one-, two-, and three-year-old females) could be distinguished by measuring the sizes of the “corpora lutea”. The duration of the reproduction and its course could be established. The survival value of a population consisting of more than one age class is discussed. INTRODUCTION Previous observations by Den Boer (1968) and Van Dijk (1973) provided information on the heterogeneity of the reproductive cycle which, in some species of carabid beetles, may be important to survival under adverse environmental conditions. This heterogeneity, which presumably is partly genetically fixed, may contribute to spreading the risk of extinction of the population. Den Boer and Van Dijk observed that not only one-year-old females of Calathus species contributed to reproduction, but also two- and even three-year-old ones. These authors observed that in young females at the start of the development of the ovaries no “‘corpora lutea” (abbreviated c. 1.) were present and that only after several weeks the first c.l. could be found. Their results suggest that it would be possible to discriminate between females of the old and the new generation in a field population by means of the cl. in the ovaria. The present work was started to investigate the validity of these facts also for Pterostichus oblongopunctatus, a carabid beetle common in woods in the Nether- lands. The reproductive pattern has been studied by observation of the develop- ment of the ovaries and the maturation of the eggs. These features have been studied during a number of successive years in order to unravel the complex dynamics of the population concerned. 2 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 The area where the population of P. oblongopunctatus has been studied, occupies 2.4 hectares and was planted with oak (Quercus robur) about 90 years ago. The oaks are sparsely intermingled with birch (Betula spec.) and have formed a dense canopy over a well-developed shrub-stratum. Herbaceous growth is practically non-existent, but a thick layer of partly decomposed leaf-litter (5-10 cm) makes a good hiding place for a rich epedaphic fauna. MATERIAL AND METHODS Pterostichus oblongopunctatus (F.), a small bronze-black coloured beetle (9-12 mm) lives as an adult in the superficial layer of leaf-litter, preferably in broadleave« woods. The larvae usually develop in the deeper layers. Reproduction occurs in spring, the eggs are deposited in the litter. Pitfall traps (Den Boer, 1968) were used for sampling: a plastic funnel (16 cm) . fitted to a glass jar filled partially with 4% formaldehyde and dug into the soil, the rim of the funnel being flush with the soil surface. A cover (Q 25 cm) is placed 5 cm above the funnel to prevent rain from penetrating into the trap and diluting the formaldehyde. Three units of five traps each were dug in. Once a week the jars were changed for fresh ones, and the specimens of P. oblongopunctatus selected from the contents of these jars. Fresh ovaries proved to be preferable for this study to those preserved in formaldehyde, but the former were available only during 1972. The females were dissected carefully and the ovaries (Fig. 1) were taken out. The development of the ovary and the absence or presence of the corpora lutea are a criterion to the age of the female. terminal filament germarium vitellarium clod of trophocytes follicle epithelium plug follicle epithelium “corpus luteum”, calyx Fig. 1. Ovariole of Pterostichus oblongopunctatus, with indication of elements VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Prerostichus oblongopunctatus 3 RESULTS In the Netherlands, this species has one generation annually. The newly hatched adults hibernate as virgins, and copulate in spring and summer. The new generation appears in autumn. Thus the activity pattern of the population shows two distinct periods in spring and in autumn (Fig. 2). The spring activity starts in 100 90 80 70 Numbers captured 60 Pterostichus oblongopunctatus O E J 50 40 30 20 10 Fig. 2. The activity of Pterostichus oblongopunctatus in 1971 approximately the first week of April and is related to reproduction. The number of beetles caught gradually diminishes towards the end of the reproduction period, i.e. the beginning of August. The appearance in September of the new generation probably stands for the major part of the higher catches in autumn. “Spent” females (i.e. those that have oviposited) represent one-third of the specimens trapped, as their activity is low at this time. Until the end of May more males than females are caught (1971: 260 against 189). Presumably the former are more active in this period, in search for a partner. From about the end of May until the first weeks of August the females make up the majority (154 against 186). The following categories of 9 in the different stages of development were recognized (PI. 2): (a) Without eggs and without c.l. (b) With eggs, without c.l. (c) With eggs under the c.l. (d) With eggs under and above the c.l. (e) With eggs above the c.l. (f) Without eggs, with c.l. (“spent”). In the females without c.l. in April, the eggs are still above the calyx, in the ovarioles: a few © (PI. 2, punctated yellow) have eggs in the calyx, but have c.l. not yet visible. Many 9 with c.l. are bearing their eggs at this time already under the c.l. in the calyx. The females with c.l. but without eggs may have finished oviposition at this moment (“spent”). The data suggest that at the start of the reproduction season females with c.l. (Pl. 2, gray) may start earlier to reproduce than females without c.l. (i.e. those belonging to the new generation). 4 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 According to Vlijm & Van Dijk (1967) it is possible to determine the age of the females by the development of the ovary and the absence or presence of corpora lutea, as follows: First year 9: (1) an ovary in a state of early development (2) asmall, pointed ovary (3) acalyx, not yet elongated before reproduction (4) without corpora lutea (5) thin, yellowish-brown c.l. È (6) thin, but distinct, dark coloured c.l. only at the end during reproduction of the reproduction period: ““c.l. 2”. 2nd- or 3rd year 9: (1) thin, dark coloured c.l. before the reproduction period (“c.l. 2°’) (2) distinct, dark coloured c.l. at the beginning of the reproduction period (“*c.l. 2”) (3) big c.l. during the reproduction period (‘‘c.l. 3”) (4) fat, “fluffy” ovaries (the eggs and ovaria are crumbling and show a fluffy structure). Since the size of the c.l. has been estimated, it is hard to give a good classification. Especially in specimens with distinct c.l. in the middle of the reproduction period one may be uncertain of their age. In a later stage of research, however, one of us (Blokhuis) used an ocular-micrometer to measure the size of the c.l. Thus more accurate data could be obtained on the difference in size of the c.l. than by mere judgement of their dimension. The data collected of every female captured in order to judge its stage in the reproduction (Table 1) are as follows: (1) the number of 9 without c.l. or eggs (I) (2) the number of 9 without c.l. with eggs (II) (3) the number of 9 with c.l. and with eggs (III, IV, IVa) (4) the number of © with c.l. without eggs (“‘spent’’) (V) (5) size and colour of the c.l. (6) the number of eggs in relation to their position in the ovary (7) the condition of eggs and ovaries. Table 2 shows the relation between the number of eggs per female and the size of the c.l. If the size of the c.l. is a measure for the age of the female, it should be possible to find a relation between age and number of eggs. The number of ovarioles varies between 10-26. The average weekly number is about 15.4-20.0. Probably this number is too low, as counting is made difficult by the “fluffy” ovaries. Three females had few c.l. (resp.3,3 and 2) but many ovarioles (14, 18 and 20). In 5 females only one ovary had developed; in 2 females a double c.l. ring could be observed. Development of the corpora lutea. As Miss Van Haaften (in 1970) and Blokhuis (1971) obtained almost identical results, these can be summarized as follows: VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Pterostichus oblongopunctatus 5 Table 1. 1971: week 1-4: April; 5—8: May; 9—13: June; 14—17: July; 18—25: Aug.; 22—26: Sept.; 27—30: Oct.; 31—35: Nov.; 36—39: Dec. I, number of 9 without c.l., without eggs; II, number of 9 without c.l., with eggs; III, number of 9 with small c.l. (c.l. 1) with eggs; IV, number of 9 with distinct c.l. (c.l. 2) with eggs; IVa, number of 9 with large c.l. (c.l. 3) with eggs; V, number of 9 with c.l. 1, c.l. 2, c.l. 3 without eggs (“spent”). Between brackets: percentages I-II (without c.l.); III (cl. 1); IV (c.l. 2); IVa-V (c.l. 3) of the total of animals caught Week I II III IV IVa V IVa+V 1971 nr. % nr. % nr. % boe. ele % | 5 2 (47) = 2 (13) l — — 5 (40) 2 6 2 (67) — 2 (16.5) 2 — — — (16.5) 3 4 9 (48) — 7 (26) 5 = = 2 (26) 4 2 10 (33) 4 (11) ll (31) 6 — — 3 (25) 5 — 1 (12.5) 38725) 27:25) 2 = — — (25) 6 3 8 (27) 15 (37) 8 (22) 5 — | | (14) 7 — 5 (31) 6 (37) 3 (19) 2 = — = (13) 8 l 4 (30) 6 (35) 4 (24) 2 — — — (11) 9 — 9 (31) 11 (41) 4 (14) 4 | — — (14) 10 = 7 (20) 15 (46) 6 (17) 5 | = | (17) 11 l 3 (16) 12 (54) 2 (14) 4 | 1 — (16) 12 — — 14 (74) 2 (16) 2 — 1 | (10) 13 — 0 — 6 (55) 1 (18) 2 — | | (27) 14 3 — (11) 12 (43) 4 (21) 3 — 2 4 (25) 15 IS (65) 9 (53) 3 (18) = = — 4 (23) 16 1 — (50) — 1 (50) = — = — 17 — — 1 (11) 5 (56) — — — 3 (34) 18 1 — (14) — 3 (43) | — — 2 (43) 19 = = IS) = — — — 3 (75) 20 = = — — | — = = (50) 21 1 — (50) — — — — | = (50) 22 2 — (40) — — = = — 3 (60) 23 1 — (50) — = = = — | (50) 24 Be (75) — — — — | — (25) 25 5 — (100) — — — — — — 26 13 — (76) — — — 2 — 2 (24) 27 8 — (73) _ — — — | 2 (27) 28 7 — (70) — 2 — | 2 — (30) 29 1 — (33) — — — = = 2 (77) 30 1 — (50) = == — | = — (50) 31 1 — (50) = = = = — (50) 32 — — — = = = — — 33 1 — (100) = = = - — — 34 = = = = = = = — 35 e — = = = — — 36 => = = — = = = = 37 — = = — = = = 38 =O — = = — == == 39 1 — (100) — — — — — — 73 60 115 70 46 9 Il 39 yvyT= sr 6 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 Table 2. 1971. The relation between the number of eggs per 9 and the size of the c.l.; II, no c.l.; III, small c.l. (c.l. 1); IV, distinct cl. (cl. 2); IVa, large c.l. (c.l. 3) Number of eggs per 9 II III IV IVa | | 2 8 2 2 l — 17 10 3 | | 13 10 4 3 6 7 4. 5 3 2 6 3 6 6 11 5 3 7 5 10 2 | 8 8 15 | 5 9 9 16 3 2 10 3 9 2 | Il 6 5 | — 12 6 4 | | 13 6 7 | 2 14 | 9 — | 15 — 4 | | 16 — 7 | — 17 | 2 — == 18 — 2 | — 19 — — — = 20 — 2 — — 21 — = = — 22 — | — — Number of 9 60 115 70 46 Number of eggs 526 1153 326 243 Mean number of eggs 8.8 10.0 4.7 973) In the second week of April females without c.l. still constitute 44-67% of the total catch. Their numbers decrease gradually to zero — after July 14 the results are unreliable. Females with c.l. 1 are observed for the first time in the last week of April. Their share begins with 6—11% and rises to 74% in the third week of June. Later the numbers decline again, but apparently more gradually than is suggested by Table I, for it has to be corrected by those females which have grown from c.l. I to c.l. 2. The resulting decline is probably caused by a decrease of activity. The number of females with c.l. 2 attain a maximum in April (30%). After April 28 their number decreases, because either their cl. 2 may grow to c.l. 3 or by mortality of the supposed 2nd year individuals. The number of females with cl. 3: this group is at the onset of the activity represented to 40%, but as April wears on their numbers decrease rapidly to 20%, in May even to 11%. After May 26 an increase takes place again, apparently as a consequence of c.l. 2 9 developing into c.l. 3 individuals as they continue to produce eggs. The initial decrease is, obviously, caused by the dying off of the supposed 3rd year generation (PI. 2). It must be noted that the first females with eggs beneath the c.l. (April) are of the c.l. 3 group, hence, the oldest females apparently take the lead in reproduction, but within a few weeks they are already “spent” again and will probably die off soon, VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Pterostichus oblongopunctatus 7 Table 3. 1972. Size of the c.l. measured in u From 30 March to 12 April From 31 May to 14 June Size Number 9 Size Number 9 !) 19 — 7) 828u 2 23u | 85u 2 31u 3 108u 2 42u | 112p l 45u 2 123u l 46u 4 49u | From 12 April to 26 April 77u 2 Size Number 9 88u I 3lu Zur DE i From 14 June to 28 J 54u 3 rom 14 June to une 62u 2 Size Number 9 120u l — 2 29u l From 26 April to 10 May 3lu | Size Number 9 40n 69u 3 From 10 May to 31 May 72u l Size Number 9 Tu 7 a Il 1081 | 31u 2 1234 2 381 1 13lu l 42 I 58, ati From 28 June to 12 July 62u 2 Size Number 9 69u l — 0 92u 3 46u | 1084 | 69u | 77u 4 1) — = without c.l. 1234 | according to their disappearance in May (Pl. 2). A postponed reproduction of the preceding year may be possible, so that 2nd year © may die when they are two years old and only those 9 which failed to reproduce the previous year would reach their 3rd year (Vlijm & Van Dijk, 1967, have supposed this to be the case in Calathus melanocephalus). Measurements of corpora lutea. The estimate of the size of the c.l. being not quite reliable, those of freshly TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 == © 1972 third March 30 a April 12 number of individuals without c.l. second year 9 9 April 13-26 first year April 27 ot May 10 May 11-31 June 1-14 June 15-28 3 number (©) ou © June 29 July 12 N a O o size in u 100 140 Fig. 3. Size of corpora lutea in 1972 VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Pterostichus oblongopunctatus 9 caught 9 were measured by Blokhuis in 1972 by means of an ocular micrometer at a magnification of 100x (Table 3). In the first weeks of April 1972 only females with c.l. 3 (108—124u) or large c.l. 2 (77—85u) were present. One female with a c.l. of 28u probably reproduced only a short time during the preceding year. In the second half of April only one 9 with c.l. 3 was captured, but females with small c.l. 2 made their first appearance (50—60u). No females with c.1. 3 could be caught during the first part of May but the c.l. of the c.l. 2 females tend to be bigger (SS—80u) and it is in this period that females with c.l. 1 have been collected for the first time. During the second part of May, in June and the first two weeks of July the c.l. 2 have reached the size which the cl. 3 showed in April, and the c.l. 1 those of the c.l. 2. During 7 weeks (April 26-June 14) not a single female with c.l. 3 had been caught. Hence, we may conclude that females with c.l. 3 at the onset of the breeding season (first half of April) die after oviposition. The new 2nd year females take their place at the end of June. These data are a confirmation of the hypothesis that more generations are involved in reproduction (Fig. 3). A check in 1973 on individuals of the 1971 generation which were marked as young females in 1972, showed a size of the c.l. corresponding with those of second year Q (cf. Fig. 4). Consequently the age of a female can be estimated approximately by these figures though a certain overlap exists. It is evident that at a certain moment of development, Ist year c.l. 2 and 2nd year c.l. 2 do not differ in 1973 third year 99 March April second year 99 April 16-19 first year 9 9 April 26 May 4 = N a HD oo May 14 size in u number 0 50 100 140 Fig. 4. Size of the corpora lutea of individuals marked in 1972, recaptured in 1973 10 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 size. A more accurate way to determine the ages of the © taking part in reproduction is the marking-recapture method which we intend to use in the . future. A serious obstacle in the interpretation of “corpora lutea” (sizes as well as numbers) as indicators of age is the phenomenon of egg resorption (Joly, 1949, and others). The precise origin of the cl. seems to be unknown. Egg production. When dissected, only 6 out of the 291 egg-bearing females appeared to have white, glossy, elastic eggs, one female had brown, hard, malformed eggs. The eggs of all the other females proved to be “fluffy” and soft and were coloured light yellow to orange. It is possible that this coloration is due to the preservation agent, i.c. formaldehyde. i In the first week of May all females are bearing eggs. This situation is reached within one month. Decrease to zero takes more time: it takes more than 3 months before the last egg has been deposited (Table 4). Table 4. 1971. The number of egg bearing 9. Week nr. 1: March 31-April 7. Week nr. 20: August 11- August 18 Week Total 9 Q + eggs % | 31 March — — — 2 7 April 15 5 33 3 14 April 12 6 50 4 21 April 27 21 78 5 28 April 36 31 86 6 5 May 8 8 100 7 12 May 43 36 84 8 19 May 16 16 100 9 26 May 17 16 94 10 2 June 29 28 97 Il 9 June 35 33 94 12 16 June 24 21 88 13 23 June 19 18 95 14 30 June 11 9 82 15 7 July 28 19 68 16 14 July 17 12 fil 17 21 July 2 | 50 18 28 July 9 6 67 19 4 Aug. 7 4 OL 20 11 Aug. 4 l 25 In 1971, the average number of eggs was 7.7. Computing the number for each group of females shows that females without c.l. and those with small and thin c.l. have an average of 9.6 eggs; c.l. 2 females have 4.6 and c.l. 3 females 5.8 (Table 5). It can thus be concluded that the females without c.l. and with cl. I produce the majority of the eggs. The groups with c.l. 2 and c.l. 3 contribute only one third of the total. By means of Wilcoxon’s Two Sample Test the zero-hypothesis of a similar egg- VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Pterostichus oblongopunctatus Il Table 5. The average number of eggs in 1970 and 1971 1970 1971 Corrected average numbers (1971) Generation 1967 | Generation 1968 Generation 1969 4.3 4.6 (4.7 for c.l. 2-group © and c.l. 3-group 9 without in April) I) 5.8 (6.3 for c.l. 3-group © in April) Generation 1970 — 9.6 N.B. Generation 1968 corresponds with the c.l. 3 group in April 1971. Generation 1969 corresponds with the c.l. 2 group and the c.l. 3 group together, without the individuals with c.l. 3 caught in April. Between brackets the corrected averages of the number of eggs. After April the c.l. 3 group should be added to the c.l. 2 group because of the development of c.l. 2 to c.l. 3 in the course of the summer. The average of 5.5 has been computed by adding the number of eggs of the generations 1967 and 1968 which until that time could not yet be discerned. Only the c.l. 3 9 caught in April probably have an age of three years (or more). production in two groups is tested against the alternative hypothesis of a different egg-production (a= 0.05; u= 1.96). It can be concluded that group II and III do not differ significantly. This could be expected as these groups belong to the same generation: 1970-groups II and IV differ significantly (u= 6.4; p<0.0001) as do groups II and IVa (u= 4.76), but groups IV and IVa do not differ significantly (u= 1.22). Miss van Haaften, in 1970, arrived at the same results with the Median Test. CONCLUSIONS AND DISCUSSION Composition of the population. The results show that the population consists of 3 generations which are all nearly simultaneously active (Table 6, which has been derived from Tables | and 5). In June and July only 2 generations are present, as the 3rd generation dies off after oviposition in April-May and the new generation hatches only in September. Survival value. As the larval stage is supposed to be rather vulnerable, it is evident that, if one generation in its larval stage is severely damaged by adverse environmental (humidity or temperature: too high or too low) and/or biotic (disease, parasites, predators) influences, the next year still two additional generations survive to assure the persistence of the population: the last 4 years about 40% of the females taking part in reproduction consisted of 2 or 3 year old (Table 7). At the beginning of 1971 (March 31-May 5), the rate of 2nd and 3rd year egg- bearing females stood at 44%, the total number captured, i.e. 2nd and 3rd year 9 without eggs included, even at 51%. As reproduction goes on these numbers decline respectively to 25% 9 with eggs and to 24% of all females captured in the period May 6-June 2 and to 30% © with eggs and 36% © of all © captured in the period June 3-July 7. The ratio of this last period is too high owing to a failure in the correct estimation of the size of the c.l. which in the future can be eliminated by the micrometer method (cf. p. 9). 12 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 Table 6. Reproductive activity in 1971 Date Age Reproductive activity Generation April Ist year without c.l., without eggs 1970 Ist year without c.l., with eggs 1970 Ist year with c.l. 1, with eggs 1970 2nd year with c.l. 2, with eggs 1969 3rd year with c.l. 3, with eggs 1968 3rd year with c.l. 3, “spent” 1968 May as April but 3rd year 9 disappear June Ist year with c.l. 1, with eggs 1970 Ist year spent with c.l. l and c.l. 2 1970 2nd year with c.l. 3 and eggs 1969 2nd year spent with c.l. 3 1969 July Ist year with c.l. 1, c.l. 2 and eggs 1970 Ist year spent with c.l. 2 1970 2nd year spent with c.l. 3 1969 August Ist year with c.l. 2 and eggs 1970 l st year spent with c.l. 2 1970 2nd year spent with c.l. 3 1969 September Hatching of the new generation 1971 Ist year spent with c.l. land cl. 2 1970 2nd year spent with cl. 3 1969 October New generation 1971 Ist year spent with c.l. I 1970 2nd year spent with c.l. 3 1969 The first year generation (1970) produced a large number of eggs (average 9.6). The second year generation (1969) had an average production of 4.7, the third year generation (1968) one of 6.3. According to Miss Van Haaften, the generations of 1969 and 1968 produced about the same average (4.3 and 5.5, respectively). Data of 1972 have already confirmed that the population increased with about 10%: the generation of 1970 may have had a big part in this increase. Survival, however, is not dependent of meteorological factors only: predation, disease and the available amount of food will be important too, but reliable data are, thus far, lacking. Comparison with other carabid beetles. It is a special advantage that Van Dijk (l.c.) studied the age composition of Calathus melanocephalus (L.) which is, contrary to P. oblongopunctatus, a ‘‘winter Table 7. Composition of the population in 1969, 1970 and 1971 Year Locality % more than Method one year old 1969 Baarn 49% by estimating the size of the c.l. (by De Bruyn) 1970 Baarn 32% by estimating the size of the c.l. (by Van Haaften) 1971 Baarn 39% by estimating the size of the c.l..(by Blokhuis) 1969 Wijster 37% capture-recapture (Van Dijk, personal communication) VAN HEERDT, BLOKHUIS & VAN HAAFTEN: Pterostichus oblongopunctatus 13 breeder” (i.e. the larvae develop during winter time). In the former species young adults hatch in spring and only mature in the course of July, while in P. oblongopunctatus the adults hatch in September, mature in the course of the autumn and start oviposition towards the end of April. In P. oblongopunctatus, egg- production by Ist year females as well as by 2nd and 3rd year females starts earlier in the season but finishes about July 21, whereas in C. melanocephalus it continues well into September. As a matter of fact, P. oblongopunctatus has a lead over C. melanocephalus as at a certain moment (end of April, beginning of May) three generations may reproduce simultaneously, contrary to C. melanocephalus where at best two generations do so. ACKNOWLEDGEMENTS We are grateful to Dr. P. J. den Boer and Dr. T. S. van Dijk (Dr. W. Beyerinck Biologisch Station, Wijster) for their stimulating interest and useful advice in our investigations. We thank Dr. W. J. Sluiter for reading the manuscript and Miss M. A. de Bruyn for putting some preliminary results at our disposal. REFERENCES Boer, P. J. den, 1967. Zodlogisch onderzoek op het Biologisch Station te Wijster. — Med. Bot. Tuin Belmonte Arb. 11. ———, 1968. Spreading of risk and stabilization of animal numbers. — Acta Biotheoretica 18: 165—194. Dijk, T. S. van, 1973. The age-composition of populations of Calathus melanocephalus L. — Oecologia 12: 213—240. Joly, P. 1949. La fonction ovarienne et son controle humoral chez les Dytiscides. — Arch. Zoöl. exp. 84: 49— 164. Vlijm, L. et al., 1961. Ecological studies on Carabid beetles I: Calathus melanocephalus. — Arch. Neerl. Zool. 14: 410— 422. Vlijm, L., & T. S. van Dijk, 1967. Ecological studies on Carabid beetles II: General pattern of popula- tion structure in Calathus melanocephalus (L.) at Schiermonnikoog. — Z. Morph. Oekol. Tiere 58: 396— 404. TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 Pil a b Ovaria and oviducts of Pterostichus oblongopunctatus. a, corpora lutea; b, egg in pedicel PLZ TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 1, 1976 “sJONpIAIpul Z sjuasasdai 20jq si 8340 4equinu JoJo, 1? gum ‘sBBa inoygim 2 ay aacgo 5669 yim ‘2 244 aA0qp pup Japun 5662 yim ‘|? ays aapun s66a yim 12 Noy ‘5663 Yum ‘12 anoygim ‘5662 gnoyyım 2/62 IL6I Ul 2949 sAnonpoldal ay} Jo AoAINg ne À eu PDEELII9° . AFLEVERING 2 1976 TIJDSCHRIFT VOOR ENTOMOLOGIE “LIBRARY (AUG 4 7 * DE NEDERLANDSE ENTOMOLOGISCHE VER NIGINGS Y UITGEGEVEN DOOR INHOUD TELLEMAN and A. D. J. MEEUSE. — Anthecological relations between repu- __ tedly anemophilous flowers and syrphid flies. I. The possible role of | syrphid flies as pollinators of Plantago, p. 15—31, fig. 1—2, pl. 1—3. chrift voor Entomologie, deel 119, afl. 2 Gepubliceerd 7-VII-1976 ANTHECOLOGICAL RELATIONS BETWEEN REPUTEDLY ANEMOPHILOUS FLOWERS AND SYRPHID FLIES I. THE POSSIBLE ROLE OF SYRPHID FLIES AS POLLINATORS OF PLANTAGO by P. STELLEMAN and A. D. J. MEEUSE Hugo de Vries Laboratorium, University of Amsterdam With two figures and three plates ABSTRACT Regular visits by syrphid flies of the genera Melanostoma and Platycheirus to the inflorescences of the reputedly anemophilous Plantago lanceolata L. for the purpose of pollen consumption have often been reported. An effective pollen transfer could be established in the present study. Observations made in the field indicated that the flies feed on Plantago laceolata pollen in the early morning. Several aspects of their behaviour already suggest them to be effective pollen vectors. By means of Scanning Electron Microscope observations the presence of pollen grains of P. lanceolata on the body of the flies could be demonstrated. Experiments in which artificially dyed pollen of P. lanceolata was deposited on inflorescences in anthesis, while the stained pollen was subsequently recovered from the stigmas of other spikes, demonstrate that they are efficient pollinators of this plant. The amount of pollen transferred was estimated and, at least in certain habitats, is thought to be responsible for a considerable part of the geitonogamous and heterogamous pollinations. The technique employed may prove to be useful for the study of comparable cases in plants and animals. I. OBSERVATIONS IN THE FIELD AND PRELIMINARY STUDIES The taxon Syrphidae (Diptera) almost exclusively consists of species which in the adult stage are typically anthophilous and feed chiefly on pollen or nectar. Differences in body size, and in the length of the proboscis, are considerable, so that a rather wide range of flowers can be visited: Among the Diptera, the Syrphidae are in general the most important pollinators, although their signifi- cance is not as great as that of the social hymenopterous group of the Apidae (Kugler, 1970; Faegri & Van der Pijl, 1971). As may be expected, syrphid flies normally visit flowers or inflorescences of typically entomophilous plants, but observations in the field have shown that some reputedly anemophilous plants are visited regularly by various species of Syrphidae for the purpose of pollen consumption. However, the published records provide relatively few reliable data. The most important ones, to be mentioned presently, refer mostly to visits by representatives of the closely related genera Melanostoma and Platycheirus of the subfamily Syrphinae. One of the earliest reports is by Miller (1873), who noted that in the surroundings of Lippstadt (W. Germany) insects often gather or consume pollen on the spikes of Plantago lanceolata. Beside Hymenoptera, he reported several 15 16 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 Syrphidae among which Melanostoma mellinum (Linnaeus) is specially mentioned as a frequent pollen-eating visitor. Muller pointed out that this fly even shows a preference for wind-pollinated flowers: he found this insects as a regular pollen consumer on Plantago media L., Artemisia dracunculus L., Scirpus lacustris L., and several grasses such as Anthoxanthum odoratum L., Poa annua L., and Festuca pratensis Huds. It was also found on a whole range of entomophilous plants where, apart from pollen, it sometimes also ingested nectar. Müller believed that visits to anemophilous plants by this type of insects might result in cross-pollination. Representatives of the closely related Platycheirus were not reported as visitors of wind-pollinated taxa. Ludwig (1881, 1884) noticed a very large number (several thousands) of syrphids of the genera Melanostoma, Platycheirus and Melithreptus (Sphaerophoria) on the panicles of Molinia coerulea (L.) Moench in Saxony (E. Germany). An appreciable number were caught by the proboscis between the paleae or the anthers and in that case they usually appeared to be infected by a parasitic fungus (Entomophtora spec.), while living flies were often found to be present also. Ludwig suggested that the purplish-mauve colour of the anthers acts as a strong attractant for the visitors. He observed the same flies on Phleum pratense L., Helictotrichon pubescens (Huds.) Pilger, Dactylis glomerata L., and Plantago lanceolata. The extensive monograph of Knuth (1898—1905) records a number of cases of Melanostoma mellinum visiting various anemophiles. Apart from citing earlier authors he also reported personal observations made near Kiel and in the North Frisian islands (W. Germany), where he frequently noted several specimens of this fly on inflorescences of Alopecurus pratensis L., Phleum pratense, and Anthoxanthum odoratum. He also mentioned visits of this species to Plantago arenaria W. & K. (in a botanical garden) and to Sanguisorba minor Scop. Knuth did not attribute any significant role to these visits in effective pollen transfer. In addition he mentioned an appreciable number of visits of Melanostoma mellinum (and other species of the genus) and of Platycheirus species to entomophilous flowers. Drabble & Drabble (1927) reported that at one time (in Great Britain) they saw inflorescences of Phleum pratense literally covered with Melanostoma mellinum and Hylemia variata (Fallén) (Diptera, Anthomyidae), which were busy gathering pollen. They believed that these insects are undoubtedly effective pollinators. Dactylis glomerata was also visited for foraging purposes but its inflorescences do not seem to be as attractive as those of other grasses. Porsch (1956) gave an extensive survey of Coleoptera and Diptera recorded as pollen consumers on anemophilous plant species, but his lists are chiefly compiled from earlier publications and do hardly contribute any new data on syrphids. For our present study the publication by Van der Goot & Grabandt (1970) proved to be important. They studied the pollen content of the digestive tract of several anthophilous species of Syrphidae belonging to the genera Melanostoma, Platycheirus, and Pyrophaena. The senior author had repeatedly (in the Nether- lands) observed these flies visiting the inflorescences of Gramineae and of Plantago lanceolata. In order to assess their possible preference for anemophilous plants the ingested pollen was studied. It appeared that the diet of several species, viz., Platycheirus clypeatus (Meigen), P. fulviventris Macquart, P. scambus (Staeger) STELLEMAN AND MEEUSE: Anthecological relations 17 and P. angustatus Zetterstedt, consisted almost exclusively, and that of others, such as Melanostoma mellinum (L.) and M. scalare (Fabricius), partly, of pollen of the above-mentioned anemophiles. (It may be mentioned in passing that Grabandt has found pollen of Plantago and of grasses in related syrphids collected in Canada; pers. comm.). In all other species of syrphid flies examined the diet normally consists of pollen of entomophilous plants, anemophilous pollen apparently only being taken occasionally. All observations cited and pollen analyses clearly indicate a certain degree of consistency in the visits of the Melanostoma-Platycheirus group to several anemophiles, more particularly to Plantago lanceolata and a number of grasses. The question arises whether these flies can indeed act as effective pollinators for these reputedly anemophilous plants and if so, what the effectiveness of pollen transfer by insects is in comparison to the anemophilous pollination. As mentioned above, some authors suggest that pollen transfer by insects may be of some importance, but this conclusion is solely based on recorded observations of visits. In our opinion such listings of visitors, although indispensable, are only indicative of a possible pollen transfer, and only an experimental analysis can lead to more definite conclusions. Such an approach has hitherto never been attempted, as far as can be ascertained. The studies reported in the present paper were primarily intended to demonstrate the possible incidence of an effective pollen transfer by syrphid flies from one inflorescence to another, both through field studies and experimentally. Three methods were employed: (1) Field studies, mainly centred around the patterns of behaviour of the syrphids, as far as important in connection with a possible pollen transfer; (2) the use of the SEM technique to study the precise localisation of pollen grains attached to the flies after a visit; (3) experiments with stained (marked) pollen to demonstrate actual pollen transfer in the field (chapter II). For practical reasons it was deemed necessary, for the time being, to limit our investigations mainly to P/antago lanceolata. The results obtained by Van der Goot & Grabandt (1970) rendered the choice of the potential pollen vectors an easy one; the following and (locally) relatively common species seemed to be the most likely candidates: Melanostoma mellinum, Platycheirus clypeatus, P. fulviventris, and P. scambus. Since it is often impossible to identify each species with certainty in the field, especially flying specimens and those paying a fleeting visit to an inflorescence being hard to recognize, the flies were, generally speaking, treated as a group. This admittedly is a somewhat weak element in our deductions, but we had no option. Exact identification is always possible when the specimen is caught and can be examined, but we refrained from this practice as much as possible. Not only is it a time-consuming procedure, but we also did not want to extract so many individuals from the ecosystem within the scope of our investigations. Identifica- tions of killed specimens were reduced to an unavoidable minimum. Fortunately we have obtained the strong impression from our observations in the field that the behaviour of all species included in our study is very similar, and for the purpose 18 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 of the present investigation identical. Whenever the terms “syrphids” or “flies” are used in the following text, unless stated otherwise, always representatives of the Melanostoma-Platycheirus group are meant. The report covers investigations carried out in 1973 and 1974; the field work was carried out from about the beginning of May till the end of September. Low-lying sites near water offered good opportunities. The main areas for study selected were the Naardermeer and (less intensively) the Ankeveen broads. Incidental observations were also made elsewhere in the province of Noord-Holland. The SEM observations were made at the Laboratorium voor Electronenmicroscopie of the University of Amsterdam. Observations in the field Observations in the field indicate that syrphids habitually visit inflorescences of Plantago lanceolata. Spikes in the male stage of anthesis (P. lanceolata is protogynous) are predominantly visited in the early hours of the morning after sunrise. There is a marked correlation between the beginning of the (male) anthesis and the activity of the flies. Both processes seem to be temperature- dependant and commence in the temperature range between 10°—13°C, but may to some extent also be incited by other factors. As soon as in a population of P. lanceolata the male anthesis has begun, i.e., when the first stamens have emerged, a few syrphids begin to arrive and descend on the inflorescences to start feeding on pollen. Normally these events take place between 5.00 and 7.00 hrs, but on relatively cool mornings, or when bad weather conditions prevail, the period may shift to a later time in the morning. In the beginning the male phase of anthesis develops slowly and the number of visitors remains limited. When the temperature rises and the weather is not too unfavourable, the anthesis rapidly becomes more general and the frequency of the visits also increases appreciably in a short time span. Every day there is a relatively short period of an optimum of visits which seldom exceeds a couple of hours and falls between about 6.00 and 11.00 hrs, but the frequency usually diminishes already before 10.00 hrs and only rarely stays high after 11.00. After the decrease the visits diminish in number, and soon become more or less incidental to cease completely at about 11.00 hrs; the flies become less active after feeding and descend to various parts of plants to rest, but especially in the beginning they often perform cleansing and preening movements with their legs. The sequential phenomena just described are probably connected with the following circumstances: (i) More or less constant and at least fairly favourable weather conditions during the early morning induce a certain regularity in both the course of anthesis and the frequency of the visits until the optimum is attained; (ii) Dissipation of the pollen from the anthers principally by air currents (wind power!) results in the exhaustion of the pollen in the anthers which reduces their importance as a source of food; (iii) The flies become satiated after having consumed a large quantity of pollen and start their resting phase. STELLEMAN AND MEEUSE: Anthecological relations 19 Unfavourable weather in the morning, but sometimes during the previous evening or night, may bring about appreciable deviations from the regular pattern of anthesis and the visiting rhythm. Our observations concerning the male phase of anthesis agree almost completely with the results of Hyde & Williams (1946) obtained at Cardiff (Wales). Continuous observation of a group of flowering plantains during which all visits were recorded revealed that some of the spikes in anthesis are markedly more frequently visited than most other ones and that a sometimes appreciable number of spikes is not visited at all. The cause of this selective preference is obscure. Quite regularly two or three flies are present on the same spike, but larger numbers are usually not encountered: eight specimens were seen on a single spike on only one occasion. The pattern of behaviour of the flies suggests that some activities may be of importance for a possible transfer of pollen. This requires some presuppositions, VIZ., (a) that pollen grains indeed become attached to the body of the fly, and (b) that the pollen can subsequently be transferred to the stigmatic area of a different flower. Some characteristics of behaviour established on the basis of numerous observations are the following. It appears that as a rule the syrphids fly in the direction opposite to the prevailing movement of the air immediately above the stand of vegetation. This mode of flight was used to advantage in experiments to be described in the second chapter. The approach flight towards a spike in anthesis is not necessarily followed by a landing on the same spike; quite frequently the insect hovers in the air in front of it for a short while to proceed to a different inflorescence where the movement may be repeated before the fly decides to descend on a third spike, etc. The final landing is preceded by a series of brief thrusting movements aimed at the inflorescence. The fly usually alights directly upon the anther-bearing zones, but not rarely on the portion of the spike above the stamens in which only the female phase of anthesis has proceeded (i.e., on the stigma-bearing zone); in the latter case the fly walks downwards towards the stamen-bearing part of the inflo- rescence. _ When manipulating the anthers to feed on the pollen, the fly is usually attached to some place on the spike by means of its second and third pairs of legs and clutches an anther with its front legs. It subsequently places its proboscis in the dehiscence fissure of a theca and works it over along its full length. As soon as the theca seems to become empty the anther is given a half turn by the front legs so that the opposite slit comes within reach of the tongue. After having thus systematically emptied an anther the fly grabs another one but may also move owards the stigma-bearing zone of the spike or take off. The syrphids seem to nrefer freshly dehisced anthers, but somewhat older ones are also used for foraging, especially later in the morning. Pollen consumption is repeatedly interrupted by preening (preening and cleaning of the body also take place when the insects are resting after feeding as we have seen). 20 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 The duration of a single visit to a flowering spike varies considerably. From over 900 observations it could be deduced that visits lasting from less than 30 seconds to 2 minutes constitute about 50% of all vistis, and that 87% of the visits do not last longer than 15 minutes. Visits of longer duration were recorded now and then; two exceptional cases of visits, lasting 56 and 58 minutes, respectively, were noted. During the working over of the anthers the flies may take up different positions. This may be of importance in connection with the possible contamination of the insect body with pollen and in some instances also with the possible deposition of pollen grains adhering to the body on pollen-receptive stigmas. The most characteristic postures are the following: (i) the insect sits inside the stamen zone; contamination of different portions of the body with pollen is possible (PI. 1 Fig. 1); (ii) the fly takes up a practically vertical position distally of the stamen zone with the head facing the anthers downwards; pollen may adhere to the frontal part of: the body, and the second and third pairs of legs may come into contact with receptive stigmas and thus transfer attached pollen grains (PI. 1 Fig. 2); (iii) the insect is seated on the already wilted portion of the inflorescence below the stamen-bearing zone in a vertical position with the head in the upward direction; pollen may adhere to the frontal parts of the body but also to its under surface (PI. I Fig. 3); (iv) the fly clings in a nearly horizontal position to one or more stamens in the lower part of the whorl of stamens; contamination of particularly the lower surface of the body with pollen is possible (PI. 1 Fig. 4); (v) in the case of spikes in which the zone of male anthesis has almost shifted to the top: the insect sits on the tip of the inflorescence in a more or less horizontal position; pollen grains will become attached to mainly the frontal parts of the body, and deposition can take place as long as receptive stigmas are present (PI. | Fig)! More or less intermediate positions occur frequently, and during feeding the fly may shift its position. The syrphids are repeatedly found on the stigma-bearing zone of the inflo- rescence, more particularly in the following situations: (a) the fly alights on this part and after a shorter or longer sojourn proceeds to the stamen-bearing zone; sometimes it stays for a while, usually cleaning and preening its body; (b) after feeding on the pollen the insect walks from the stamen-bearing zone towards the stigma-bearing part and usually stays there for some time preening its body; it sometimes returns to the male zone to resume pollen consumption, but may also fly away (PI. 1 Fig. 6). If the female phase of anthesis has not terminated, the legs or the ventral side of the body may get in touch with receptive stigmatic surfaces, which may result in pollen transfer. In the first situation described above (a) the chance of cross- pollination is greater than in the second (because in the latter case the pollen is likely to originate from the same spike). SEM studies of the captured flies After the presence of pollen grains on the bodies of flies captured immediately STELLEMAN AND MEEUSE: Anthecological relations 21 after a visit to a Plantago inflorescence had been established by examination under a binocular dissecting microscope, a more detailed study was made by means of SEM microphotographs. For practical reasons only attention was paid to the ventral surface of the body (including the extremities). The study served a dual purpose: the pollen grains present could be identified more accurately, and the distribution of the pollen across the fly could be studied. Methods The flies were captured when they were resting on vegetative parts of plants and never when they were sitting on inflorescences so as to avoid possible contamina- tion with the large quantities of pollen shaken out of the anthers when one is handling the killing jar. For each fly a separate, clean jar was used to avoid indirect contaminations as much as possible. After a captured fly had become sufficiently desiccated it was mounted with its dorsal side on an object support by AT SX lA O7 À 4, Fig. I. “Pollen map” of a syrphid fly (see text). The black dots represent pollen grains of Plantago lan- ceolata (dots situated outside the outline of the body represent pollen grains attached to setae protru- ding from the surface of the fly body) LN 22 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 means of silver cement and coated with gold in an Edwards (model E 12 E) vacuum-coating apparatus. The coated specimens were systematically scanned for adhering pollen grains at suitable magnifications in a Cambridge Mark II Stereoscan (acceleration potential 10 KV). A schematic outline drawing of a fly was used to map the localisation and the species of the identified pollen grains; for each specimen a so-called pollen map (an example of which is shown in Fig. 1) was made in this way. Special and characteristic situations were photographically recorded. Results In all 32 syrphids were studied which belonged to all four species mentioned in the introduction. On all but one of these specimens the presence of pollen grains could be demonstrated. This pollen was almost exclusively of Plantago lanceolata; : only in two cases a few gramineous grains were noted. The number of grains per fly varied a great deal: from a single one to a recorded maximum of 179 with a mean of 44. The localisation of the pollen on the ventral body surface was as follows: on the head only a few grains were observed, if any, and even around the mouth there were very few or none. The thorax carried the greatest number as a : rule, especially in the pleural parts and on the hairs found in this part of the body. The abdomen bore fewer grains, mainly on the belly. Of the leg parts the femur carried the largest number of grains; between the three pairs of legs no important differences were found. Discussion The scanning technique demonstrated conclusively that pollen grains of Plantago lanceolata adhere to the bodies of the syrphid flies. The relatively few number of specimens studied do not permit an accurate assessment of the distribution of the pollen on the bodies of the flies, but it seems as if the head does not carry many grains. The grains are sometimes found on the glabrous cuticle, but more often adhere to the dermal setae and do not always touch the cuticle. The attachment is almost certainly attributable to a cementing substance, whose presence (in small amounts) in Plantago was shown by Pohl (1930) and Knoll (1930) almost simultaneously. On SEM microphotographs (compare Plate 2) the pollen grains, when lying close together, are frequently connected by a mass forming a conspicuous meniscus. Also between a grain sticking to a hair and the support such a substance is present. Pollen grains pretreated with ethanol before SEM examination do not show this phenomenon, so that it appears to be highly probable that the connecting matter is hardened pollen ‘Kitt’? covered by a gold film. The almost complete absence of grass pollen on the flies is of special interest. Conceivably, its relatively great rarity has something to do with the incidental nature of visits to species of Gramineae for feeding purposes. It is also known that the surface of gramineous pollen grains is only very thinly covered with Pollenkitt, although the absolute quantity may vary appreciably from species to species (Pohl, 1930); this small amount of cementing substance may account for the poor sticking capacity of the grains (or for a short duration of their attachment during the flight STELLEMAN AND MEEUSE: Anthecological relations 23 of the insect). One must also bear in mind that the handling of the flies during the capture and their subsequent preparation may rub off a number of grains from the fly body, so that the SEM images do not necessarily present the situation immediately before the insects were caught: grass pollen may be rubbed off more easily than the plantain pollen. Concluding remarks Field observations of syrphid flies regularly feeding on P/antago lanceolata indicate that these insects may easily become contaminated with pollen. More casual visits to the inflorescences of grasses with easily accessible anthers may also result in the adherence of pollen to the body. The efficiency of an insect as a pollinator is largely dependent on the constancy of its habits and its patterns of behaviour. If it frequently moves from one plant in anthesis to the inflorescence of a different individual of the same plant species the chance of transfer of specific pollen is appreciable. However, the efficacy of zoophilous pollination is also determined by the amount of pollen carried by the animal vector on its body, which amount differs from plant species to plant species owing to the varying quantities of pollen and to the presence or absence of other aids for the adherence of the grains to the insect body (such as a rough or spinulose ectexine). As far as can be deduced from the available data, grass pollen is less effectively transferred than plantain pollen, which may be interpreted as a greater reliance upon pollination by syrphid flies of Plantago lanceolata as compared to the grasses, at least in the area under investigation. This needs verification, and this relation should also be studied in other habitats. The different amounts of plantain and grass pollen recorded on captured flies support the idea of a greater dependence of Plantago lanceolata on the flies in its pollination strategy. The repeated visits of syrphids to different individuals in stands of plantain render an efficacious pollen transfer highly probable. It also enabled the demonstration of the transfer of pollen from one plant to another to be dealt with in the next chapter. II. EXPERIMENTS WITH STAINED POLLEN OF Pantago lanceolata L. After the transfer of pollen of Plantago lanceolata by certain syrphid flies had been shown to be highly probable (see the first chapter), experiments were carried out to demonstrate the actual translocation of pollen grains from one plant to another by means of pollen marked by artificial staining. It has already conclusively been shown by means of SEM microphotographs that pollen grains adhere to various parts of the fly body, but although there were cogent indications of the role of the insects as effective pollen vectors, there remained some uncertainty whether a deposition by the flies of pollen grains on the receptive stigmatic surfaces of a flower in the female stage of anthesis actually takes place. By using artificially coloured grains partly replacing the normal pollen the identification of pollen grains deposited on other inflorescences was made possible, so that the displacement of this pollen could be traced. In addition, it was 24 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 hoped that from such experiments certain ideas concerning the efficacy of a biotic pollen transfer under field conditions could be gleaned. The marking of pollen by staining Some preliminary tests had shown that pollen of P/antago lanceolata can readily be dyed by means of aqueous solutions of biological stains such as 0.5 % Methylene Blue and 0.5 % Neutral Red. Attempts to dye pollen grains still enclosed in the ripe anthers by spraying the inflorescences with such staining solutions were not very successful. Only a relatively small amount of pollen present near the slit of dehiscence appeared to have absorbed the dye, but the solution did not penetrate far enough. The anther wall may be more or less water- repellent. It was decided to use a technique in which pollen collected in the field was dyed in the laboratory as follows. Pollen from spikes bearing numerous stamens in anthesis was harvested by beating the spikes over a metal tin. To obtain a quantity sufficient for subsequent processing at least 20-30 inflorescences must be shaken out. A small quantity of the collected pollen is poured on a disc of filter paper covering the bottom of a petri dish and spread out evenly by means of a fine paint brush. The dish is sprayed with the dye solution from a flower sprayer until the filter paper has assumed the colour of the solution. The (open) dish is left to dry in a suitable place at room temperature, which takes 2-3 hrs. To assess the condition of the grains the dish is placed under a dissecting microscope and the grains are touched with a fine preparation needle: if they are loose and can be easily shifted (or even jump away) the pollen is sufficiently desiccated. It is subsequently transferred to a glass tube _ with a brush and can be kept for some time until needed. Pollen treated in this way was examined for some properties required for the purpose of the present investigation: (i) Stained and normal (= untreated) pollen grains observed under a micro- scope at magnifications of 100—200 diam. appear to be always clearly distinguish- able from one another; (ii) When marked pollen was poured on to receptive stigmatic surfaces and the stigmas were subsequently examined under a dissecting microscope, such treated pollen grains appeared to adhere firmly to the stigmatic papillae and were clearly distinguishable in situ from untreated grains; (iii) When living syrphid flies of the taxa studied (Melanostoma, etc., see the first chapter of this paper) were placed in a tube with dyed pollen, left in the tube for several minutes, and narcotised for examination under the microscope at a magnification of about 40 diam., they proved to carry a substantial amount of pollen on their bodies; the localisation of the pollen grains was very much the same as in the case of contamination with untreated pollen. The treated grains apparently do not differ appreciably from undyed ones in some properties essential for their transfer by flies. To remove possible remaining doubts, the power of adherence of untreated and of dyed pollen was tested by locking up syrphids in a petri dish containing a small amount of a mixture of untreated and dyed grains in a known quantitative relation. After a brief sojourn STELLEMAN AND MEEUSE: Anthecological relations 25 the flies were narcotised and their bodies were examined for adhering pollen. From counts of the two kinds of grains for each fly (42 specimens) the relation between untreated/dyed pollen was estimated. The result was as follows: mean ratio untreated/dyed pollen in the mixture: 4.41 + 0.12 [mean ratio untreated/dyed pollen on the fly bodies: 5.18 + 0.54]. Experiments in the field: methods For the experimental set-up the following reasoning was followed (see also the situation sketch, Fig. 2): When of two groups of flowering plants in a sufficiently direction Saw BE Da = a wind a e A tential testslid. d P “receptor spikes er spikes min. 40 cm Fig. 2. Situation sketch of a field experiment (explanation in text). For the sake of clarity the rosette of leaves is drawn as if they are adpressed to the ground (as in Plantago major and P. media) dense stand of Plantago lanceolata one (A) is used as the donor group by applying dyed pollen to the spikes in the manner to be described below, and the other group (B) is left undisturbed, a syrphid fly visiting one or several donor spikes is likely to become covered with pollen including a number of treated grains. If it subsequently alights on untreated spikes there is a distinct possibility of dyed pollen becoming deposited on the stigmas of these spikes, which can be established by examining such spikes under a dissecting microscope (PI. 3 Fig. B). During the practical execution of the experiments the possible transfer of wind- borne pollen had to be taken into account. The relative position of the group of donor spikes was chosen in such a way that it always remained below the wind in respect of the untreated spikes present in the neighbourhood. A very favourable circumstance is that the syrphids tend to fly against the direction of the air current over the site, in the figure from group A towards group B. It goes without saying that this increases the efficacy and the reliability of the experiment. In order to establish if nevertheless an undesired pollen transport of stained 26 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 pollen took place, test slides were placed between the two groups which were coated with egg-white glycerol on the one side to catch air-borne pollen grains; these slides were placed at a height a little below the average height of the donor spikes, with the smeared side facing the latter. The distance between the groups of donor plants and receptor plants must not be too short (at least 40 cm) so as to avoid the risk of an undesired pollen transfer by direct contact between plants or by air turbulence. The area chosen for such experiments was always limited enough to be easily surveyable from one observation post and thus to permit the reliable recording of visits to receptor spikes by flies coming from donor spikes by following the flight of the latter. After some trials a method was worked out to bring about a contamination of flies with dyed pollen which agrees to all intents and purposes with the mode of contamination with untreated pollen during a normal visit to an inflorescence. Dyed pollen is placed on a flowering spike as follows: A sufficient quantity of dyed : pollen is put in a glass tube of about 10 cm long and 1.5—2 cm in diam. By shaking and tapping the tube whilst turning it about its longitudinal axis in a nearly horizontal position the pollen is evenly distributed over the inner wall. Such a tube is placed over a spike in male anthesis with a well-developed zone of stamens, the inflorescence during this and the subsequent operations being held in a horizontal : position by bending its stalk carefully. By pushing the spike to and fro and, if necessary, by turning the tube a bit, some of the dyed pollen is rubbed off and sticks mainly to the anthers (but also to other parts of the inflorescence such as the stigmatic zone). The spikes must be dry; damp ones moisten the inside of the tube too much so that the pollen sticks too firmly to the tube and becomes useless. Such a treatment of the inflorescences does not deter flies from visiting them normally, even if the whorl of stamens has assumed a somewhat unnatural colour. The dyed pollen grains are as readily consumed as untreated ones; in a number of cases a marked bluish-green discolouration of the abdomen of a fly was observed after the animal had been feeding copiously on spikes treated with stained pollen. It may, therefore be taken for granted that the deposition of marked pollen grains on a spike does not affect the normal behavioural pattern of the flies in the least. The added pollen is also dispersed in a normal way by air currents as could be deduced from the gradual disappearance of the greyish discolouration, after some time, of a spike treated with pollen dyed blue, even if no insect visits took place. Examination of flies captured after they had visited a donor spike revealed the presence of coloured pollen grains on the fly body (PI. 3 Fig. A). Experiments in the field: results Altogether 22 experiments were carried out, the duration of each single experiment being 60 to 150 minutes. The number of (treated) donor spikes was 3-6 and the receptor spikes usually numbered 3—4. Of about 90 spikes which were gathered in the field and might carry stained pollen grains ultimately only 67 could be used for the records, because some mishaps occurred which rendered them unusable (some became soaked with water, etc.). The following qualifications need some explanation. STELLEMAN AND MEEUSE: Anthecological relations 27 Fresh stigmas: stigmas with a shiny, somewhat vitreously white appearance; the distal parts may be suffused with brown. Old stigmas: stigmas which have mostly or completely turned brown and have often become desiccated. The duration of the receptive phase of the stigmas is not exactly known, but nevertheless the distinction made above, based on a different appearance, corresponds most probably with the receptive phase and the phase of post- anthesis of the stigmas. After the experiments were concluded, a publication by Zeisler (1938) came to our notice in which a simple method is described to assess the degree of ‘maturation’ (receptiveness) of the stigmas by a chemical reaction (with H,O,). If this test is applicable to Plantago stigmas, some relevant additional information may thus be acquired. Whole spike: includes bracts, perianth members, stigmas, and a zone of stamens. Stigma ratio (s.r.): number of stigmas with stained pollen grains total number of stigmas Control with test slides: The pollen counts were made of a rectangular area of 10 mm by 20 mm in the middle of the slide at a magnification of 100—200 diam. Not in a Single instance stained pollen was recorded, but a few undyed grains were usually present; the latter may have been deposited by air turbulence or Table 1. Presence or absence of dyed pollen on spikes of Plantago lanceolata L. Marked pollen present in Marked pollen absent in On fresh stigmas 43 spikes 24 spikes On fresh and/or old stigmas 48 spikes 19 spikes On whole spikes 50 spikes 17 spikes convection currents, or as fall out of the local pollen rain from somewhat higher levels. In a number of cases test slides were also placed in a downwind position and these almost invariably had caught large quantities of both stained and unstained pollen. The (67) receptor spikes were examined for the presence of marked pollen grains and the results are as follows (Table 1). Table 2. Distribution of dyed grains over spikes of Plantago lanceolata L. Number of dyed grains Number present On fresh stigmas On old stigmas On whole spike (0) (24) (19) (17) 1-10 25 15 11 11-20 8 14 11 21-50 7 14 18 51-100 2 2 5 more than 100 1 3 5 43 48 50 28 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 The following survey shows the number of individual grains in the cases in which dyed pollen had been transferred (Table 2). In the accompanying tabel (Tabel 3) the stigma ratio of 41 spikes with transferred dyed pollen grains, and the Table 3. Stigma ratio (s.r.) and number of grains on 41 spikes with transferred dyed pollen grains ST. number of SIT. number of Sir number of grains grains grains 3/9 - 8 10/10 - 136 Vile = 3 1/10 - 4 2/2 - 19 6/10 - 36 6/1 A= 63 7/9 - 47 ar "=" 20 MOSE 8 3/6 - 19 18/8 682 Al anr 4/8. - 26 2/6 - 2 8/10 - 47 SIATE 12 6/9 - 26 1/29 - 2 VA - 2 4/11 - 4 2/4 - 9 W237 4 2/26 - 9 4/8 - 31 1/21 - 12 1/8 - 1 2/11 - 7 2/4 - 3 W/22 RE 1 5/17 - 10 4/14 - 14 3/3) 2219 2/3 - i vl - 2 3/40 6 Diss 5 2/14 - 5 VO er IO Aje RO 3/21 - 13 number of grains on the stigmas is shown. Of two of the 43 spikes with stained pollen grains the number of fresh stigmas was inadvertently not counted. Discussion The examination of the test slides indicates that an abiotic transportation of pollen against the direction of the wind is negligible. The conclusion that can be drawn from the results of the first survey (Table 1) is that syrphid flies are instrumental in the transfer of pollen from one spike to another and can do this repeatedly. The absence of marked pollen grains on potential receptors can have several reasons, the most important being: (a) that the spikes were not visited by flies, or (b) that the spikes were visited, but by flies that did not carry stained pollen grains. One can only speak of an effective pollination if pollen is deposited on fresh (and presumably receptive) stigmas. This was the case in 20 experiments and in 64% of the total number of spikes examined, so that the general conclusion remains valid. One of the surveys (Table 2) gives a good idea of the quantities of marked pollen grains transferred by the animal pollen vector, but the results do not require a special discussion. From the viewpoint of efficacy of pollen transfer the number of stigmas on which dyed pollen had been deposited is important. Table 3 shows that there are appreciable differences in this respect, both in stigma ratio and in the number of stained grains. This is to be expected if one considers that a number of mutually independent factors is involved in the pollen transfer to the stigmas, viz., (a) the total number of visits to a spike, STELLEMAN AND MEEUSE: Anthecological relations 29 (b) the duration of these visits, (c) the mode of landing, behaviour and/or displacement of the pollen vector, (d) the amount and the location of the pollen on the body of the visitor, and (e) the number and the length of the stigmas on the particular spike. The first two factors could be quantified by the registration of the visits, but there was no correlation with the quantity of deposited marked pollen. Matters were complicated by the uncertainty in many cases whether a visitor was indeed a carrier of dyed pollen. The dyed grains were especially deposited on the tips of the stigmas as may be expected because these protruding parts are the most likely to come in touch with the insect body. It is also noteworthy that stained grains were localised on the stigmas in groups. Another survey (Table 4) demonstrated that of all pollen-bearing stigmas only Table 4. Distribution of stained pollen grains over the stigmas 36 stigmas (24.3%) with 1 grain 24 stigmas (16.2%) with 2 grains 24 stigmas (16.2%) with 3 grains 64 stigmas (43.3%) with 4 or more grains 24% had caught a single dyed grain, whereas in 76% of the cases two or more grains were present. Since the gynoecium contains 2(-3) ovules the theoretical chance of efficacious fertilization is rather great. The experiments described in the previous chapter render it highly probable that pollen treated with stains has practically the same properties as untreated pollen at least as far as its power of adhesion to the insect body and of its successful translocation is concerned. It is, in our opinion, not necessary to put its equivalence with regard to adhesion to the stigmatic surface to the test: the sticky stigmas readily retain all sorts of particles (some of which are larger than the pollen grains) and their capacity to capture pollen grains by their adhesive properties certainly exceeds the affinity of the pollen grains to the surface of the insect body. Various considerations render the same chances of being transferred by biotic agencies. Considering that untreated pollen seems to adhere slightly better to the insect body, the transfer of untreated pollen by flies can be accepted beyond reasonable doubt, even if the transfer of untreated grains from a donor spike to a receptor spike could not be directly demonstrated: untreated pollen found on receptor spikes may have been deposited by a fly coming from a donor spike but just as well from various other sources outside the experimental area. One must bear in mind that the duration of the experiments was usually shorter than the time-span of the daily visiting period (although it always included the phase of optimum activity of the flies), so that an — albeit relatively small—part of the potential pollen transfer was not recorded. This only means that the rate of biotic pollen transfer is undoubtedly slightly higher than it appeared to be under our experimental conditions. The fact that the flies often move about on the same spike (and preen themselves) favours the incidence of geitonogamy, and so does their habit of frequently moving towards inflorescences in the immediate neighbourhood (which may well belong to the same individual of Plantago 30 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 lanceolata), but our experiments have shown that effective cross-pollinations are by no means rare. As far as the relative importance of efficiency of the biotic and the anemophilous form of pollen transfer is concerned, the present experiments do not permit a clear verdict. A specially adapted method (by means of fertilisation experiments?) would have to be worked out for the purpose. An ecologically interesting and relevant circumstance is that the syrphid flies of the Platycheirus-Melanostoma group as a rule visit the Plantago inflorescences during the first hours of the morning (see the first chapter). At that time of the day the wind is often hardly noticeable or at least very weak, so that presumably the relative share of the biotic pollen transfer in respect of that by air currents will be greater than later during the day. Local differences in the relative rates of the two types of pollination associated with different environmental conditions and habitat variation (such as landscape and vegetation type and microclimatological differences) cannot be precluded. The question in how far these variations may be correlated with, or reflected in, modifications of the inflorescence is of interest. Delpino (1870, cited and discussed in Muller, 1873) thought he could recognise forms with differences in structural adaptation, which was accepted with some diffidence by the latter author. One should never loose sight of the fact that a biotic transfer of pollen of Plantago lanceolata as described in the present paper can only take place where this plantain and the potential pollinators occur sympatrically, which need not always be the case. Field observations and experiments in other areas are indicated, and it will be necessary to extend the investigation by comparing the anthecological syndromes of other species of Plantago with the situation in P. lanceolata. The technique for the dying of pollen for purposes of specific identification after a possible transfer can probably be improved and extended to be used effectively for the study of other cases in which different taxa and plants and other (potential) pollen vectors are involved. Coldwell (1951) has developed a useful method to study the dispersal pattern of air-borne pollen (of different species of conifers) by using pollen labelled with radioactive substances (isotopes). This technique is rather laborious, however, and other workers who have attempted to apply this method have not been successful (B. J. D. Meeuse, priv. comm.). The use of artificially stained pollen permits a more or less small-scale approach to various anthecological problems, more particularly in border-line cases between en- tomophily and anemophily. ACKNOWLEDGEMENTS The authors are highly appreciative of the active assistance of Mr. V. S. van der Goot (Amsterdam), especially with the identification of the syrphid flies. The technical assistance of Mr. C. Bakker (Werkgroep Scanning-Electronenmicro- scopie, Amsterdam) and of Messrs. F. D. Boesewinkel, L, Dijkhuizen, C. L. Hansson, H. J. Koerts Meyer, A. Smit and J. Vuijk (Hugo de Vries Laboratorium) is thankfully acknowledged. STELLEMAN AND MEEUSE: Anthecological relations 31 The permission granted by the „Vereniging tot Behoud van Natuurmonumenten in Nederland” to do field work in the Naardermeer nature sanctuary deserves grateful memorisation. REFERENCES Coldwell, R. N., 1951. The use of radioactive isotopes in determining spore distribution patterns. — Amer. J. Bot. 38: 511—523. Delpino, F., 1870. Applicazione della teoria Darwiniana ai fiori ed angli insetti visitatori dei fiori. Dis- corso pronunziato del Dr. Erm. Mueller di Lippstadt. Versione dal tedesco, e annotazioni. — Bullet. della Soc. Entom. Ital. 2: 140—159. Drabble, E., & H. Drabble, 1927. Some flowers and their Dipteran visitors. — New Phyt. 26: 115—123. Faegri, K., & L. van der Pijl, 1971. The principles of pollination ecology (2nd. ed.). — Pergamon Press, Oxford. Goot, V.S. van der, & R. A. J. Grabandt, 1970. Some species of the genera Melanostoma, Platycheirus and Pyrophaena (Diptera, Syrphidae) and their relation to flowers. — Ent. Ber. 30: 135—143. Hyde, H. A., & D. A. Williams, 1946. Studies in atmospheric pollen III. Pollen production and pollen incidence in ribwort plantain (Plantago lanceolata L.). — New Phyt. 45: 271—277. Knoll, F., 1930. Über Pollenkitt und Bestäubungsart. — Z. Bot. 23: 609—675. Knuth, P., 1898—1905. Handbuch der Blitenbiologie I—III. — Engelmann, Leipzig. Kugler, H., 1970. Blütenökologie (2nd. ed.). — Fischer, Stuttgart. Ludwig, F., 1881. Weitere biologische Mitteilungen. I. Molinia coerulea als Fliegenfangerin. — Bot. Centralbl. 8: 87. ———, 1884. Uber den Fliegenbesuch von Molinia coerulea. — Bot. Centralbl. 18: 123. Müller, H., 1873. Die Befrüchtung der Blumen durch Insekten. — Engelmann, Leipzig. Pohl, F., 1930. Kittstoffreste auf der Pollenoberflache windblitiger Pflanzen. — Beih. bot. Centrbl. 46, - (I): 286—305. Porsch, O., 1956. Windpollen und Blumeninsekt. — Osterr. Bot. Z. 103: 1—18. Zeisler, M., 1938. Uber die Abgrenzung der eigentlichen Narbenflache mit Hilfe von Reaktionen. — Beih. bot. Centrbl., sect. A, 58: 308—318. u Tui ' ‘ ‘ t dur. bal j TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 Pl. 1 Different positions of syrphid flies on inflorescences of Plantago lanceolata (explanation in text) TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 PINS 3 le SEM microphotographs of pollen grains of Plantago lanceolata on the bodies of syrphid flies. 1, X 600; 2, X 720; 3, X 900; 4, X 900 Pa ù un) tar TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 2, 1976 Pl. 3 A. Part of the head of a syrphid fly to which normal (light coloured) and stained (dark coloured) pollen grains are attached, X 50. B. Stigma of an inflorescence of Plantago lanceolata collected after a visit by a fly, showing the presence of untreated (light coloured) and stained (dark coloured) pollen grains, X 50 dt i N N ‘ . Na Seiad a > È n è x £ stia DEEL 119 AFLEVERING 3 1976 TIJDSCHRIFT VOOR ENTOMOLOGIE MUS. COMP, Zels: a LIBRANY 4 UITGEGEVEN DOOR 7 (AUG 1 7 15,5 i HARVARD 3 DE NEDERLANDSE ENTOMOLOGISCHE VERENIGINERSITY. È 4 a . bee INHOUD c. VAN ACHTERBERG. — A preliminary key to the subfamilies of the Braconidae _ (Hymenoptera), p. 33—78, fig. 1—123. Gepubliceerd 7-VII-1976 ec A PRELIMINARY KEY TO THE SUBFAMILIES OF THE BRACONIDAE (HYMENOPTERA) by C. VAN ACHTERBERG Rijksmuseum van Natuurlijke Historie, Leiden With 123 figures Abstract A preliminary key to the subfamilies of the Braconidae is given. The subfamilies are redefined in the key and in a short description. Especially the apomorphous characters, which may be used to define the subfamilies, are discussed. INTRODUCTION The family Braconidae can be separated from its sister-group, the Ichneu- monidae, by at least two synapomorphous characters. The most important is the rigidly joined 2nd and 3rd tergites of the metasoma and, secondly, the absence of the second recurrent vein. The rigid plate formed by the 2nd and 3rd tergites may be weakly sclerotized (as in the Aphidiinae) or may be completely fused, e.g., in the Hybrizontinae and Acaeliinae. The first worker, who tried to give a classification of higher groups was Wesmael (1835), whose classification concerned only the Palaearctic region. The Braconidae were divided into two divisions by Wesmael (p. 11): the “braconides endodontes” and the “braconides exodontes’’. The latter group is what is now called the Alysiinae. The “‘endodont’’-Braconidae were divided into four sub- divisions (p. 14): (1) the “polymorphes”, (2) the “‘cryptogastres’’, (3) the “areo- laires”, and (4) the “cyclostomes”. The “polymorphes” contain the Aphidiinae, Euphorinae, Helconinae, Ichneutinae and Opiinae. The “cryptogastres” are an aggregate of Cheloninae and convergent groups, belonging to the Helconinae and Meteorideinae. The ‘‘areolaires”’ embrace the convergent groups Microgaster- inae, Acaeliinae, Orgilinae and Agathidinae. Finally the ‘“‘cyclostomes”’ contains the Doryctinae, Rogadinae and Braconinae, as treated in the present paper. The first reviser of this system was Foerster (1862), who gave a very elaborate sub-division and defined 26 ‘‘subfamilies’’, to which he added the suffix ‘‘-oidae’’. The following systems were based more or less on the Foerster-system, but several “subfamilies” were given tribal rank in one large subfamily. E.g., the Bra- conoidae, Euspathioidae, Hecaboloidae, Doryctoidae, Hormioidae, Rogadoidae and Rhyssaloidae became in Fahringer (1925) the tribes Braconini, Spathiini, Hecabolini, Doryctini, Hormiini, Rhogadini, Exothecini and Pambolini in one subfamily Braconinae, in effect the old “cyclostomes” of Wesmael. The “‘Sigal- phoidae” became a part of the Helconinae, while the “Chelonoidae” are the 33 34 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 present Cheloninae and Meteorideinae. The ““Microgasteroidae” contain the Microgasterinae and Acaeliinae, and the Agathidinae were divided by Foerster in the “Agathidoidae” and the “Eumicrodoidae”, while the last mentioned also contains the Orgilinae p.p. The “Pachylommatoidae” and “Aphidioidae” are called Hybrizontinae and Aphidiinae, respectively, in the present paper. The “Euphoroidae” and “Perilitoidae” form the Euphorinae, from which quite recently Zemiotes Foerster was excluded and added to the Zelinae (Mason, 1973). The “Brachistoidae” of Foerster are of composite nature, they belong partly to the Helconinae and partly to the Zelinae (as defined in the present paper). The “Livphronoidae”, together with Pygostolus Haliday from the “Bla- coidae”, form the tribe Centistini in the Euphorinae. The remainder of the “Bla- coidae” belongs to the Helconinae. The “Ichneutoidae” and “Helconoidae” became the Ichneutinae and Helconinae p.p., respectively. Foerster’s “Macro- centroidae” contain the Macrocentrinae and Zele Curtis of the Zelinae; this: grouping is still widely used, e.g., by Eady & Clark (1964) and Capek (1970). The ‘“Diospiloidae” belong to the Helconinae and Orgilinae. Except for Gnaptodon Haliday (which is included in the Rogadinae), his “Opioidae’” are fully com- parable with the Opiinae. Ultimately his ““Alysioidae” and “Dacnusoidae” are united by Griffiths (1964: 831) in the subfamily Alysiinae; this was confirmed. through the study of the larvae by Capek (1970: 861). The system used by Marshall (1885: 9) is essentially that of Wesmael, but he added as a sixth division the “Flexiliventres” for the Aphidiinae, because of their flexible and weakly sclerotized metasoma. In 1891 he added as a seventh division the “Pachylommatidae”, now called Hybrizontinae. These mainly large groups were subdivided according to the system of Foerster with some modifications, but also resulting in 26 “subfamilies” (for which Marshall used the suffix ‘‘-ides’’) for the Palaearctic region. The key by Ashmead (1900: 111) to the subfamilies of the Braconidae is the first general key, intended to be used in more than one region. Ashmead separated the Alysiinae as a family Alysiidae, while the remaining genera were divided among 17 subfamilies. His division and interpretation of the genera was often incorrect, because he used a modification of the Foerster system without examining the types of many genera. The second and latest general key to the subfamilies, which was used more extensively, is the key published by Szépligeti in 1904. It divides the Braconidae into 31 subfamilies, one of which (the Lysiognathinae) belongs in the Ichneumo- nidae. From the remaining 30 subfamilies the following are removed and reduced to lower rank in the subfamilies mentioned in parentheses: Cenocoelininae (Helconinae), Gnathobraconinae (Rogadinae), Aphrastobraconinae (Braconinae); Exothecinae (Rogadinae), Spathiinae (Doryctinae), Hecabolinae (Doryctinae), Pambolinae (Rogadinae), Hormiinae (Rogadinae), Sigalphinae (Helconinae), Calyptinae (Helconinae), Liophroninae (Euphorinae), Blacinae (Helconinae), Cardiochilinae (Microgasterinae), Diospilinae (Helconinae), Meteorinae (Eupho- rinae) and Dacnusinae (Alysiinae). The remainder, 14 subfamilies, are augmented to 22 in the present paper, because of three newly described subfamilies (Yp- sistocerinae Cushman, 1923; Telengainae Tobias, 1962; Mesostoinae Van Achter- C. VAN ACHTERBERG: The subfamilies of the Braconidae 35 berg, 1975) and a rearrangement of five groups according to Capek (1970 and 1973) and Tobias (1967). These five subfamilies are the Acaeliinae and Zelinae (according to Tobias), the Orgilinae, Meteorideinae and Neoneurinae (modified after Capek). The two subfamilies erected by Fahringer (1936: 586) in his fairly general (but rather confusing) key are not accepted in the present paper. His Aneurobraco- ninae (containing the genus Aneurobracon Brues) is provisionally treated in this paper as a tribe of the Agathidinae, while the Pseudodicrogeniinae (containing only the genus Pseudodicrogenium Fahringer) is included as a tribe in the Braconi- nae. The morphology of both types examined clearly justify this transfer (Fig. 120—122). After Szépligeti, several others have given subfamily-keys for the Palaearctic region (e.g., Fahringer, 1925; Tobias, 1971), while Marsh (1963: 522) gave a key for the Nearctic region. This key was disregarded in his later key to the Nearctic genera (1971: 841), because “there is some disagreement as to the limits of the various subfamilies in the Braconidae,....’’. For the terminology used, see Van Achterberg, 1976. DISCUSSION The main reason for the confusion in the systematics of the Braconidae is formed by the many convergent evolutionary trends, occurring in this group. Some of the most important trends are the following: |. The reduction of the veins of the wing, e.g., the Blacini in the Helconinae versus the Centistini in the Euphorinae. 2. The forming of rows of setae, usually one row per tergite. In this respect I disagree with Griffiths (1964: 842), who stated that “the arrangement of hairs in single rows on the gastral segments, being found in almost all other Alysiinae and many other Braconidae is beyond doubt plesiomorph’’. I am convinced that the opposite is true: many species with many plesiomorphous character-states have the tergites evenly setose. Still more convincing is the fact that most Ichneu- monidae (the sister-group, in most aspects clearly less evolved) have the meta- somal tergites evenly setose. 3. The formation of a carapace, formed by the three basal tergites of the metasoma, e.g., in the Helconinae (Brachistini), Cheloninae, Microgasterinae (Fornicia Brullé), and Rogadinae (Tobias & Dudarenko, 1974). 4. The ovipositor becomes shorter, sometimes more or less curved, and its sheaths become often wider. 5. The reduction of the segments of the palpi; the plesiomorphous condition of the maxillary and labial palp is 6 and 4 segments, respectively. Lower numbers occur in e.g., the Braconinae, Alysiinae, Helconinae, Neoneurinae and Hybri- zontinae. 6. The selection of hosts other than Coleoptera larvae, especially Lepidoptera and Diptera. 7. The development of endoparasitism. 8. The parasitism of adult insects by the Neoneurinae, Aphidiinae and Eupho- rinae. 36 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 9. The reduction of the number of antennal segments occurs in almost all groups. 10. The development of a petiolus at the basis of the first metasomal tergite: e.g., Doryctinae (Spathiini), Euphorinae (Meteorini) and Zelinae (Zemiotini). 11. The forming of a hypoclypeal depression and a more or less concave clypeus: e.g., Braconinae, Opiinae, and Euphorinae (Cosmophorini). 12. The twisting of the mandible: e.g., Macrocentrinae, Braconinae and, very extremely, in the Alysiinae. 13. The formation of a dorsope and/or a laterope, while often also the TA carinae become more developed, until beyond the middle of the first tergite: e.g., Helconinae, Rogadinae, Alysiinae and Euphorinae. 14. The spiracle of the first metasomal tergite becomes situated near the middle of the tergite: e.g., Euphorinae and Zelinae (Zemiotini). 15. The radial cell becomes short and/or narrow, a tendency occurring in almost all subfamilies. 16. The first metasomal tergite becomes depressed, especially laterally: e.g., Mesostoinae, Braconinae, Microgasterinae, Acaeliinae and the tribe Proteropini of the Ichneutinae. 17. The convergent development of an “ophionoid facies” (Gauld & Huddles-: ton, 1976) an adaption for nocturnal activity, occurring in the Zelinae, Euphori- nae, Macrocentrinae and Rogadinae. 18. The mandibles of the larvae become slender and toothless: e.g., Euphorinae, Neoneurinae, Aphidiinae, Alysiinae, Opiinae, Microgasterinae p.p., and Rogadi- nae p.p. 19. The reduction of the cephalic structures of the larvae, especially of the hypostoma: e.g., Euphorinae, Orgilinae, Helconinae (Blacini & Brachistini), Neoneurinae, Ichneutinae (Muesebeckiini), Opiinae, and Doryctinae. 20. The labial and stipital sclerites of the larvae become long and slender: eg, Euphorinae, Helconinae, Zelinae, and Opiinae. Mainly owing to the research by Capek upon the cephalic structures of the final instar larvae, I have tried to construct a key to the subfamilies proposed by Capek and Tobias. Some changes seem inevitable to me, e.g., the fusion of the subfamilies Rogadinae and Exothecinae. There are too many genera in these groups, that connect the groups in more than one character. The mummification of caterpillars by Rogas and closely allied genera is not sufficient (however peculiar it may be) to separate them as a subfamily. The Centistinae of Capek (1970) are included in the Euphorinae, according to Capek (1973). The Adeliini s.l., a tribe of the Microgasterinae in Capek (1970), are redefined and partly maintained as the tribus Miracini in the Microgasterinae; the other part is treated as a separate subfamily, the Acaeliinae, according to Tobias (1967 and 1971, who used the name Adeliinae). The Orgilinae (the tribe Orgilini of the Agathidinae in Capek, 1970) and the Braconinae (the tribe Braconini of the Braconinae in Capek, l.c.) are treated as separate subfamilies. The three other tribes (Exothecini, Hormiini and Pambolini) of the Braconinae in Capek are included in the Rogadinae. The Spathiinae of Capek (1970) is at most a tribe of the Doryctinae, as pointed out by Capek (1973: 267). C. VAN ACHTERBERG: The subfamilies of the Braconidae 37 More problematic is the position of the genera Zele (and Zemiotes), Charmon (= Eubadizon of Capek), Acampsis and Sigalphus. Capek (1970) included Zele and Macrocentrus in his Macrocentrinae following, for example, Nixon (1938) and Eady & Clark (1964). But Watanabe (1969: 319) considered the most adequate location to be in a tribe Zelini of the subfamily Helconinae; also Capek (1973) gave them a tribal rank. The larvae are distinct from the larvae of the Helconi- nae: the labial sclerite is pentagonal, while it is transverse (Cenocoeliini) or longitudinal (height larger than width in other tribes) in the Helconinae; they are endoparasites of Lepidoptera, while Helconinae are almost exclusively endoparasites of Coleoptera; and the imagines of Zele and its relatives have several synapomorphous characters in respect to the Helconinae as treated in the present paper. It is therefore that I follow Tobias (1967, 1971), who gave this group subfamily rank, and I also include Charmon and Zemiotes. The adult morphology of Charmon (e.g. the genitalia of the male, cf. Fig. 36, 40 in Tobias, 1967), its biology (also endoparasites of Lepidoptera) and the regular shape of the emergence opening from the cocoon (irregular in the Orgili- nae, its near relatives) indicate the relationship with Zele. The cephalic structures of the final instar larvae of Charmon (Fig. 15 in Capek, 1970) are different from Zele, because of the absence of the hypostomal spur, but the slender shape of the hypostomal parts and of the stipital sclerite make a relationship with the Zelinae more likely than with the Orgilinae (cf. Fig. 16, 17 in Capek, 1970) as treated in this paper. Still more complicated is the placement of Acampsis, Sigalphus and Meteoridea. The larvae of Acampsis and Sigalphus have slender mandibles with a very wide base (Fig. 34, 35 in Capek, 1970), unlike the basally slender mandibles of the Cheloninae (Fig. 36, l.c.). Capek overlooked this difference, because he included them in the Cheloninae (1970: 871), even in his key to the larvae (1973: 261) where he mentioned as the key factor for the Cheloninae “. . . (mandibles) without or only with a small base”. Thus owing to the shape of some cephalic structures of the larvae and because of many differences in the adult morphology (e.g., nervellus broken, postpectal carina absent), I agree with Tobias (1967: 659), who gave this group subfamily rank. Unfortunately Tobias (l.c.) used the name “Sigalphinae”, a name for a long time (incorrectly) used for a group now included in the tribus Brachistini of the Helconinae. Capek (1970: 871) erected the sub- family Meteorideinae for Meteoridea, because the biology of this genus is rather peculiar. It is (at least partly) comprised of gregarious endoparasites of Lepi- doptera-larvae, which let pupate the host larva and thereafter the parasites spin their cocoons inside the host cocoon. The stout stipital sclerite, the very wide base of the mandible of the larvae with its apical half slender and toothed (cf. Fig. 12, 34, 35 in Capek, l.c.), combined with similarities in the morphology of the adults (e.g., nervellus broken, rather short radial cell, first discoidal cell petiolate and Ovipositor sheath wide) give some indications about its relationship with the “Sigalphinae” of Tobias. Of the characters mentioned at least the slender apical half of the mandibles of the larvae, the stout ovipositor sheath of the adults and the rather short radial cell are synapomorphous. Therefore I propose to unite the Meteorideinae of Capek and the Sigalphinae of Tobias into the Meteorideinae s.l. TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 38 (pepn1oxe yooqeseny snanauob1]0404) 2EBUT193SP3010 1 2BUTI[ 220 2euTISÂTY (pepntoxe 4ept1eH uopo4doup) aeuttdg BeUTOISOSEN aeutesuatzay, (pepnrour ]aeusaM SNAQUOTIILSTH) seutuodeig (pepnrour AepiteH uopogdpug) seurpedoy (pepnro -X2 [eeusam SNOW -0434824) 2eut19410q 2eut12201s1sdx 1aded juasead aEBUITIUDOIPIE) pue “(pepnrout yoeqeseny snanauob1]0404) “d:d seurzrogse3ororn SeUITIYooIpieg pue ‘-d‘d seurydre3rs ‘-d-d oeurygesy ‘-d-d oeut1e1se30191Y SeUTTTYDOTPIeI pue ‘d:d oseutydte3ts ‘*d*d seuryze3y ‘d:d seurzegse3ororn "d:d seprozegse3oIorn ‘d:d ,,S9IIBTO21V,, ‘d:d aeutzaqse3010 IW ‘d:d seutiraqsez0191y ‘d:d »eurı1a3sedoısıy ‘d:d seprozagse3ororn eepTIsÁIv 344 JO SePUISNUDET eeursnu2e( eeursnuoed pue SeUIIsSÁIV eeprosnuoed pue SeUIIsSÁTy pue aBUTISATY SAITIUEJQNS eu] pue ‘9epIOISATY .S893UOPOXH,, (pepnrour (pepnrout (pepnrout (pepnrout KeprreH uopogdoun) £eptieH uopoadpun) £eptieH uopozdnuy) kepıTey uopozdoun) aeuttdg aeuttdg aeuttdg aeptotdg ‘d:d | saydiomkqog,, eeuTuodeigojseaydy aeutTuooRIg pue s9eutuoveag 2BUTUOIPI1Y ‘d:d seprouose1g aeuttyyeds ay JO TUITUIOH pue 2euTuo2e1qou}eus Tutfoqueg seq113 eu] pue ‘aeutoay OXY pue seutpesouy aeprofessAuy ‘aeutloqueg ay} UT TUIPESOYN pue ‘oeprorwioH ‘eeutIuIOH pue ‘turjessAyy ‘*d*d eeprouoseig eeurpedoy ‘aeutpesouy “IUTI9YJ0OXH SAGIII ‘eeprope3oy 8201507249, aeurrygeds ou] jo turryjeds oqrıy ay} pue seutpesoyy aeprol 2U3 JO TuIToqeoeH pue “(pepnjour jeew -S2M SNAQUOLIISIH) Tut3041oq S9QUII -0q299H pue ‘aepto -tyjzedsng ‘(pepnro -UT ]9aewusaM SHIQUOT -2481H) eeptojoA10q aeuttyjeds pue ‘(pap -NTOUT ]9aEwsaM SNIAU —-042951H) 2eut119410 oeuttuyjeds pue ‘(pep -NTOUT Teewusam Sno -049981H) veur394ıoq IS6I SASTAIEM 2 Yoqosonw 7061 ‘133311dazs 0061 ‘peeouysy ‘19351904 gegı ‘Toeusem T981 Jaded juasoid ou ur pesodoid uolsiAip ay} YNM paredwos sioyine wos Aq UdAIZ Jepruoseig ay) JO SUOISIAIG ‘| 9JQRL 39 C. VAN ACHTERBERG: The subfamilies of the Braconidae aeutquoztiqhy eeurrprudy SBUTINIUOSN aeurıoydny aeUuTT T3109 seuTT eZ 2PUTUOOTSH 2BUTIJU929019EN eeutpru3e8y SeuTopr1osIeW (pepnrour yoeqeseny snansuobi] 040g) aBUTINIUYII 3euruojaug aeutjemuoy Axeg eeurtprudy 2PUTINSUOSN eeurzoydousog pue ‘*d°d seurioudng *d*d oeutoe]g ‘d:d seutioydng pue <“*d*d seuroerd ‘*d*d seuruooren *d*d seuroerg pue ‘*d*d seuruooren OBUTIFZUIIOIIEN Seurpryresy *d*d »eutuojayy pue ‘-d’d seuruooren SBUTINIUMIT ‘d:d »euruofew aeuUIJewWWOT ÁXed eeurrprudy ‘d:d seuryze3y eeuruoiydorj pue ‘-ded oeutroetg ‘d:d, seurzoezen ‘eeutrioudng *d*d seurftdsorgq pue ‘ded aeurygedy *d*d seurzoejen pue ‘dd eeur3d47e9 ‘*d°d oeut1]u29019EY ‘ded eeutuoreu) pue ‘seurydre8rs ‘-d’d seuroerg “*d*d eeut1d47e)9 ‘dd eeurrrdsotq ‘oeururjeooousg ‘aeuruooTeH ‘ded seurigusooroen ‘ded eeutry1e3y ‘d:d »euruojayn pue ‘d:d seuryze3y OBUTINIUYIT *d*d seuruojfauy aeutumoTAxeg eeuttprudy *d*d oeut191se30191Y ‘d:d seuroerg ‘-d’d seurioegen ‘aeurıoydng *d*d seuroerg pue ‘*d'd eeut11u29019E ‘d:d oeutroe]g pue ‘-d-d seursnuoed ‘d:d seut10212y ‘-d’d seutr11u99019E ‘d:d »euruofayn pue “*d*d seurydre3rs “*d*d seuroerg ‘aeutuo9T>H ‘ded seutr11u29019EN ‘d:d seurprygesy -d’d aeutuo]ayg pue ‘dd eeutptuy1e3y OBUTINIUYIT *d*d seuruofayy eeprogewwor Ayoed aeptotprydy ‘d:d »eproı1a3se3o1aTy eeprouoiuydotT pue ‘-d’d seprooerd ‘aeproıoydny ‘ded sepro3TIraeg *d*d sepropoaorung pue ‘-d*d seprortdsorq *d*d aeptojstyoeig pue ‘*d*d aepro1111124 ‘-d-d septo11u29019EY aepIoyd]e3Is pue ‘*d*d seprorrdsorg ‘*d°d seprooerg ‘-ded aeptoqstyorig “aeprouooTeH ‘d:d seproigueooioen ‘d:d eepropoiorung pue ‘aeproptuze3y ‘*d*d seprouoreug eeprogneuyor ‘d:d seprouoreug ‘d:d „seydıow4Tog,, ‘d:d „saydzouktod, ‘d:d „seireroeiv, ‘d:d ,saydioWA]od,, n8913S8303d429,, pue ‘*d*d ,saydiowA]og,, *d*d „seydıow4Tod, ‘ded „seıtefogay, ‘ded ,,s211se30d419, ‘ded „seydzoukTog, ‘d:d ,,s213se303d419,, 40 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 The division of the Braconidae given by Wesmael (1835), Foerster (1862), Ashmead (1900), Szépligeti (1904), and Muesebeck & Walkley (1951) are compared with the division given in the present paper in Table 1. KEY TO THE SUBFAMILIES OF THE BRACONIDAE S.L. 1. Mandible unidentate (Fig. 10); antenna inserted on the top of the head (Fig. 8, 9); ovipositor issuing near middle of metasoma (Fig. 8); whole body densely setose (Fig. 8); associated with termites ...... Ypsistocerinae (p. 44) — Mandible with 2—7 teeth; antenna inserted below top of head, at height of the eyes (Fig. 5, 11, 25); ovipositor issuing near apex of metasoma (Fig. 11, 68, 72, 89, 95, 108, 109); body less setose; associated with other groups ........... 2 2. Mesoscutum protruding above pronotum (Fig. 11); epistomal (medially) and scutellar sutures absent (Fig. 12, 15); ovipositor bent upwards (Fig. 11); clypeus. straight medially, above an indistinct hypoclypeal depression (Fig. 12) ...... sy MAA I SRO ee a I RR aN PAR: Mesostoinae (p. 44) — Mesoscutum not strongly protruding above pronotum, if exceptionally slightly protruding, then ovipositor straight or bent downwards; epistomal and clypeal sutures almost always complete; hypoclypeal depression absent, if present, then clypeus more or less concave and differently shaped (Fig. 27, IMI AEREE La NR O Ce OES Go 06 3 3. Hypoclypeal depression comparatively wide, deep and more or less round dorsally (Fig. 27, 28, 35, 37, 41), exceptionally the face takes part in the hypoclypeal depression (tribe Pseudodicrogeniini), resulting in a concave face and a flat clypeus (Fig. 121, 122); metasoma rather often sculptured, some- times its second tergite with a semi-circular suture (Fig. 38) .............. 4 — Hypoclypeal depression absent, if present (Fig. 5, 52), then more or less shallow, narrower and straight dorsally or nearly so; face and clypeus more or less convex, not concave (Fig. 85, 86); metasoma (if intermediate) often smooth behind the first tergite (Fig. 63), its second tergite without semi- circular suture, at most with a more or less chevron-shaped impression (Fig. 63) PORRO jones OI NI IO VIII no oc 8 4. Antenna situated at a protuberance, in front of the eyes (Fig. 97—99); venation of wings reduced (Fig. 105); maxillary palp with 4 segments; para- sites of adult Coleoptera ............ tribe Cosmophorini of the Euphorinae — Antenna more or less situated between the eyes (Fig. 25, 27, 34, 42), without distinctly developed protuberance; venation usually more complete (Fig. 19, 23, 31, 36, 43, 44); maxillary palp with 5 or 6 segments; parasites of larvae of various groups Of INSECKS” Aln LE 5 5. First metasomal tergite strongly flattened basally and laterally (Fig. 22, 26); dorsope absent or nearly so, if present, then dorsal carinae are situated later- ally, above the spiracles; prepectal carina absent; occipital carina absent dorsally; maxillary palpus with 5 segments, if with 6 segments (tribe Histero- merini, Fig. 25) then hind femur strongly compressed, disk-shaped (cf. Fig. 20) and fore tibia with a cluster of spines on two-third of its outer surface (Fig. 21); 10. C. VAN ACHTERBERG: The subfamilies of the Braconidae 41 nervulus interstitial or nearly so (Fig. 23) or antefurcal (Fig. 19), very excep- tienallypostfurcal(Fig--120) … Lu wrs corset 2m Lia Braconinae (p. 45) First tergite not or slightly flattened, usually with dorsope and/or dorsal carinae (Fig. 33, 38, 40, 46); dorsal carinae are removed from the spiracles; occipital carina usually (partly) present dorsally; prepectal carina variable; maxillary palp with 6 segments; hind femur not disk-shaped, at most weakly compressed; fore tibia with a row of spines (Fig. 32) or without spines (Fig. 39); nervulus usually postfurcal (Fig. 31, 36, 43, 44) or absent, exceptionally ater AS BEE EE 6 . First metasomal tergite immovably joined to the second tergite, without dorsal carinae (Fig. 29, 30); three basal tergites about as long as half of metasoma .. RT RS ER NE bla à Telengainae (p. 45) First tergite flexibly joined to the second tergite, almost always with dorsal carinae, at least basally (Fig. 40); if exceptionally the first tergite is immovably joined to the second tergite, then the formed carapace is about as long as the FIESCHI SIRO MORE E ROLO OO eased la RC spule eye ms OT fl . Fore tibia with short, often rather thick spines (Fig. 32), if intermediate, then occipital and prepectal carinae complete; hind coxa usually with an anterio- ventral tubercle; if nervulus is present, then sides of first brachial cell parallel (IE LG, a N RP EP er Doryctinae (p. 45) Fore tibia without spines (Fig. 39); occipital carina often partly, or com- pletely absent; prepectal carina variable; hind coxa without tubercle; if intermediate then first brachial cell widened apicad (Fig. 43) and nervulus E SER e ata man. Lera con ori à kin Rogadinae (p. 45) . Mandibles with inner side out, their tips not touching when closed (Fig. 48, 50); mandibles usually with 3—7 teeth or lobes (Fig. 49, 51), seldom with medial tooth large and both lateral teeth small ........... Alysiinae (p. 46) Mandibles normal, their tips touching when closed (Fig. 5, 52, 85, 86, 90); mandibles with two teeth, exceptionally with three teeth ................. 9 . Spiracles of first metasomal segment on its weakly sclerotized pleuron (Fig. 56, 59); prepectal and occipital carinae completely absent ............ EREN O N E Rl Dae Microgasterinae (p. 46) Spiracle of the first segment on its strongly sclerotized tergite (Fig. 61, 63, 77, 80, 82, 83); prepectal and occipital carinae variable ................. 10 Subdiscoidella present, often consisting of a more or less disconnected and yellowish or brownish stripe; nervellus often weakly pigmented posteriorly in respect to its anterior half (Fig. 65, 67, 73); if intermediate, then nervellus Biken EME MOA) ip Aden. ts Adha ET libere hei lana BS 11 Subdiscoidella absent (Fig. 74, 75, 79, 81, 84, 92); nervellus not broken, seldom its posterior half weakly pigmented in respect to the anterior half IPAM ODOM) oer TE ete PT are. er 12 . First discoidal cell sessile or subsessile (Fig. 65); cu 1 often absent; occipital carina completely absent; second tergite often with a chevron-shaped impres- sion (Fig. 63); fore side of radial cel shorter than the pterostigma (Fig. 65) ... 42 We TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fe ait ale hI NEN SCORES OEY LR SO eee et E Agathidinae (p. 47) First discoidal cell distinctly petiolate and cu 1 always present (Fig. 64, 67, 73); second tergite without chevron-shaped impression; fore side of radial cell equal in length to the length of pterostigma or somewhat longer (Fig. 64, 67, 73); occipital carina present, at least laterally ..... Meteorideinae (p. 47) . Postpectal and prepectal carinae completely present (Fig. 71); first metasomal tergite immovably joined to the second tergite, forming with the third tergite arigid'carapace (Fig. 72)) RR RR EIRE Cheloninae (p. 47) Postpectal carina absent (Fig. 87, 91), exceptionally a ventral remnant present; prepectal carina variable; first tergite usually movably joined to the second tergite (Fig Wo) oO AR OS PE Re De, PO 13 . First metasomal tergite flattened laterally and concave medially, almost invisible and immovably joined to the second tergite, forming with the third. tergite a shield that covers slightly less than the basal half of the metasoma (Fig. 61); cuqu | from the pterostigma, usually far from r 1 (Fig. 62); transverse carina of propodeum strongly developed; occipital carina completely present N eae wearer pny es a ARE PRE SEON Acaeliinae (p. 47) First tergite distinctly convex (Fig. 53, 80, 83), if seldom somewhat flattened, : then occipital carina absent; first tergite usually flexibly joined to the second tergite; if the first tergite is immovably connected, then the three basal tergites cover almost the complete length of the metasoma; cuqu | from the radius (Fig. 75, 79, 118); transverse carina of propodeum less developed ........ 14 . Metasoma inserted above the hind coxae (Fig. 77, 78) .................. 15 Metasoma inserted between the hind coxae, at least partly (Fig. 87, 91, 103,106, 109) U ER ees EN VERE RA Fee Ea IAA 18 . Metasoma inserted far above hind coxae (Fig. 78); occipital carina present dorsally: 42%: #3 PI ae eee tribe Cenocoeliini of the Helconinae Metasoma inserted near hind coxae (Fig. 77); occipital carina absent dorsally E ERI We Pr tah AIA CR EREN 16 . Trochantellus with apical teeth externally (Fig. 77); submediellan cell large (Fig. 75, 81); middle lobe of mesoscutum more or less protruding (Fig. 76) ha PO ia OI Ra I LIE Macrocentrinae (p. 48) Trochantellus without apical teeth; submediellan cell small (cf. Fig. 23); middle lobe of mesoscutum not protruding "MERE 17 Metapleural flange absent but with a rather wide and thin carina (cf. Fig. 91); radial cell comparatively wide (cf. Fig. 101); occipital carina present laterally’. „2405 AEEA et Rs UN ET ARRE Orgilinae (p. 50) Metapleural flange present (cf. Fig. 78, 87); radial cell very narrow (cf. Fig. 65); occipitalicarimajabsent) at Re RR re Agathidinae (p. 47) . Maxillary palp short, with 2 or 3 segments (Fig. 3, 5); first brachial cell almost square (Fig. 4) or rqu present (Fig. 7); parasites of ants ................. 19 Maxillary palp usually longer, with 4—6 segments; first brachial cell elongate and rqu absent;’parasites of other groups ee 20 . First brachial cell almost square (Fig. 4); fore basitarsus very slender (Fig. 2); 20. 21. DO. 23, 24. 25: C. VAN ACHTERBERG: The subfamilies of the Braconidae 43 clypeus very narrow, longer than wide (Fig. 3); ovipositor straight; malar Spacsiconeave posterioriy til Ai. aren MG oe Hybrizontinae (p. 48) First brachial cell elongate (Fig. 7); fore basitarsus stout (Fig. 6); clypeus wide, wider than long (Fig. 5); ovipositor strongly curved ventrad; malar space without concavity posteriorly ................. Neoneurinae (p. 49) Nervellus absent and radial cell (so far present) moderately long (Fig. 93, 96); first brachial cell comparatively narrow (Fig. 93, 96); metasomal tergites weakly sclerotized (Fig. 95); first metasomal tergite usually slightly or not widened apicad; clypeus comparatively small (Fig. 94); pleural suture weakly developed; occipital carina present, at least laterally; parasites of aphids en TUTI ORA, STO AS Aphidiinae (p. 49) Nervellus present, if indistinctly developed or absent, then radial cell very short (Fig. 118), first brachial cell and clypeus wider; tergites distinctly sclerotized; first tergite and occipital carina variable; pleural suture usually widgland'crenulate; parasites of other groups !.....2..... en. 21 Basal vein strongly curved at its anterior end (Fig. 88); occipital carina com- LEE AREA oi che ii N ee à Ichneutinae (p. 49) Basal vein scarcely or not curved anteriorly (Fig. 92); occipital carina usually PES EI SRI ee TRIES PR NEE IE 22 First tergite of metasoma flattened basally and laterally, without dorsal carinae (Fig. 82); occipital carina absent; anterior tentorial pits very large (Fig. 90); radial cell short and comparatively high (Fig. 92); basal vein more HS Een bent: parasifes'ofsaw-flles n. she dateren aten mon an sn en ern hr ne EN tribe Proteropini of the Ichneutinae (p. 49) First tergite scarcely or not flattened (Fig. 53, 83); occipital carina and/or dorsal carinae present; anterior tentorial pits much smaller (Fig. 52, 115); radial cell longer, if short, then narrow (Fig. 118); parasites of other groups nt ARE RI RA SIR ARTE, E I RN RESINE E ERA ue is 23 Prepectal carina absent and metasoma short, oval (Fig. 53), its first tergite flexibly joined to the second tergite; hypoclypeal depression present or absent; notauli and precoxal suture more or less reduced; parasites of Diptera ..... rade leo venin Siani ze Le Opiinae (p. 46) Prepectal carina almost always present; metasoma usually more slender (Fig. 111—113), sometimes tergite immovably joined to the second tergite; notauli and precoxal suture usually less reduced; almost always parasites of BEBEREEBUNS(O. citi hanstes beneeglial id adik ask) aioe seo; First metasomal tergite petiolate (Fig. 111, 113), sometimes petiolus rather short (Fig. 112) or length more than 3.2 times its apical width (Fig. 114, 116); spiracle of first tergite usually situated medially or behind the middle of the E a la dii ici ae 25 First tergite sessile (Fig. 80, 83), shorter than 3.2 times its apical width; spiracle usually situated in front of the middle of first tergite (Fig. 80, 83) io net ie outa ror ammel Lis ten til LIL) ell 26 Radiellan cell widened distad (Fig. 107), sometimes with an interradiella; whole surface of 4th and Sth tergites of metasoma densely setose (Fig. 111); 44 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 GUqui2ipresenti HN Ble NCIS ENNE tribe Zemiotini of the Zelinae (p. 50) — Radiellan cell narrowed or scarcely widened distad, without interradiella (Fig. 117—119); if cuqu 2 present, then 4th and Sth tergites with one row of setae pertergite (Fig. 112,113). rn en peewee tld: "Euphorinae (p. 50) 26. Laterope deep (Fig. 109) and cuqu 2 absent, if intermediate, then ovipositor sheats wide, shorter than three times its maximum width (Fig. 109) ....... EN EE A Er tribe Centistini of the Euphorinae (p. 50) — Laterope shallow (Fig. 87) or absent, if present then cuqu 2 present and/or ovipositor sheaths very slender, much longer than three times its maximal Widths siii ee en RENEE en CRC 27 27. Dorsal carinae of first tergite present, at least basally (Fig. 80, 83); if cuqu 2 absent and dorsal carinae of first tergite weakly developed, then dorsal surface of propodeum distinctly shorter than its posterior surface (Fig. 87) and radial cell comparatively wide (Fig. 84); mainly parasites of larvae of Coleoptera’: cna ae ae ee een Helconinae (p. 48) — Dorsal carinae absent (Fig. 91, 103, 106), if present then cuqu 2 absent, dorsal surface of propodeum scarcely separated from its posterior surface and of more or less equal length (Fig. 91, 103, 106) and radial cell narrow (Fig. [01):;parasites of Lepidoptera-larvaen RO eae 28: 28. Anal lobe of hind wing distinctly developed and usually with a more or less developed aqu’(Fig. 100, 104), if aqu’ absent, then metapleuron with a thin, more or less protruding carina (Fig. 103); distal abscissa of radius more or lessicurved (Fig: 1004104) ER Re ren Zelinae (p. 50) — Anal lobe of hind wing comparatively narrow, without aqu’ (Fig. 101, 102); if anal lobe is intermediate, then metapleuron without a thin carina, at most with a small flange (Fig. 106); distal abscissa of radius straight (Fig. 101, ID e ee Orgilinae (p. 50) SHORT DESCRIPTIONS OF THE SUBFAMILIES Ypsistocerinae (Fig. 8—10) Small subfamily, which contains the two genera Ypsistocerus Cushman and Termitobracon Brues. They live probably as parasites in the nests of termites in the Neotropical region. This group possesses many apomorphous characters, e.g., the unidentate mandibles (Fig. 10); the far retracted hypopygium (Fig. 8); the densely setose body (Fig. 8); the highly inserted antenna (Fig. 9) and the strongly reduced palpi and eyes (Fig. 9, 10). Mesostoinae (Fig. 11—18) Small subfamily, containing only the genus Mesostoa Van Achterberg from the Australian region, of which the biology is unknown. As pointed out by Van Achterberg (1975: 158) almost all characters are apomorphous, e.g., the absence of the occipital and prepectal carinae (Fig. 11); the absence of the precoxal and scutellar sutures (Fig. 11, 15); the flattened first metasomal tergite without carinae C. VAN ACHTERBERG: The subfamilies of the Braconidae 45 (Fig. 14); the smooth propodeum; the concave frons (Fig. 18); the upcurved ovipositor and the compressed legs (Fig. 16). Braconinae (Fig. 19—27; 120—122) Large subfamily, consisting of solitary or gregarious ectoparasites of larvae of holometabolous insects. The less evolved species often parasitize Coleoptera, the more evolved species also Lepidoptera, Hymenoptera-Symphyta and Diptera. According to Capek (1970: 862) the host larva is paralysed at egg-deposition, the parasite-larva feeds on the paralysed host and forms its delicate cocoon at a sheltered place. The cephalic structures of the larvae are remarkably homo- geneous. For the most parts none is reduced except for the mandibles, as in the main part of the Rogadinae as defined in the present paper. Therefore Capek enlarged the concept of the Braconinae, in my opinion incorrectly because it is based on an aggregate of plesiomorphous characters. It only indicates that this group is less evolved than could be expected from the morphology of the adults. The adults have several apomorphous characters, e.g., the flattened first tergite (Fig. 22, 26); the reduction of the occipital and prepectal carinae and the maxillary palp consisting of 5 segments (except for the Histeromerini). Telengainae (Fig. 28— 30) Small subfamily containing only the Palaearctic genus Telengaia Tobias. The biology is unknown. The shape of the metasoma is peculiar (Fig. 29, 30), but I am not sure about its position; it may be only a member of the Rogadinae. Doryctinae (Fig. 31—36) Rather large subfamily, consisting of ectoparasites of larvae of (wood-boring and bark-mining) Coleoptera. Less common also other hosts in plant tissues are attacked. The host-larva is paralysed before egg-deposition. The cephalic structures of the larvae are similar to those of the Braconinae. The morphology of the adults and their biology indicate their comparatively close relationship to the ancestral stem of the Braconidae. Some of the few apomorphous characters are the spines of the fore tibia (Fig. 32); the often large hypoclypeal depression (Fig. 35) and the usually distinctly developed dorsope (Fig. 33). Rogadinae (Fig. 37—44, 46) As treated in the present paper a large group and rather heterogeneous because in this group the transition to endoparasitism has taken place. This is accompanied with some pecularities as the tooth-less mandibles of the larvae (but already in the Braconinae occur larvae with only a few teeth on their mandibles) and the mummification of the host-caterpillar. In at least one genus closely related to Rogas Nees (viz., Bucculatriplex Viereck) the pupation takes place in the host- pupa in stead of in the mummified host-larva. Some species are solitary or 46 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 gregarious ectoparasites of larvae of Lepidoptera (Hormius Nees and its relatives; also Oncophanes Foerster, but O. campsolechiae Watanabe (probably belonging to Clinocentrus Haliday) is reported to be a gregarious endoparasite of larvae of Lepi- dopterous leaf-rollers; their woolly cocoon is formed outside the host). Others are solitary ectoparasites of Hymenoptera-Symphyta (Phanomeris Foerster s.l.) or solitary endoparasites of mining larvae of holometabolous insects (Colastes Haliday), while Rogas Nees and its allied genera are mainly solitary or gregarious endoparasites of larvae of Lepidoptera, which pupate within the mummified larva or in the pupa. Some apomorphous characters are the often deep hypoclypeal depression; the reduction of the occipital carina in several genera; the endopara- sitism of the main part of the subfamily and the often distinctly developed dorsope of the first metasomal tergite. Alysiinae (Fig. 45, 47—51) One of the largest subfamilies of the Braconidae of which the larval cephalic structures, e.g., the simple and smooth mandibles, resemble those of the larvae of the Opiinae. Almost all species are solitary endoparasites of larvae of Diptera, but some are gregarious (Aphaereta Foerster). The egg is usually deposited in the larva, seldom in the egg (Polemochartus Schulz) or probably sometimes in the (pre-)pupa (Aphaereta Foerster). The delicate cocoon is made in the puparium of the host. The most striking apomorphous character of this group is the rotation of the mandibles of the adults in a way that the outer side becomes the inner side. Other apomorphous characters are the absence of the prepectal and the occipital carinae, together with the development of additional teeth on the mandible and of dorsope. Opiinae (Fig. 52, 53) Rather large subfamily, consisting of endoparasites of larvae of Diptera; pupation in the puparium of the host. Ciosely related to the Alysiinae but the mandibles are normally attached (but more or less twisted) and the occipital carina is almost always present laterally. Some of its apomorphous characters are the smooth mandibles of the final instar larvae, the absence of the prepectal carina, the comparatively short and stout metasoma, the more or less reduced notauli, precoxal and pleural suturae and the tendency to develop a hypoclypeal depression. Microgasterinae (Fig. 54—60) Large subfamily, consisting of endoparasites of larvae of Lepidoptera. Often gregarious; the eggs are deposited in the egg of the host or in the early instar larvae. In the Microgasterini and Cardiochilini the larvae pupate outside the host; the gregarious species often spin together in a common web. In the Miracini (which are endoparasites of leaf-mining Lepidoptera) the larvae pupate in the host-cocoon. The wing venation is often strongly reduced (Microgasterini, Mira- C. VAN ACHTERBERG: The subfamilies of the Braconidae 47 cini), but the most important apomorphous characters are the position of the spiracle of the first metasomal segment and the absence of the prepectal and of the occipital carinae. Agathidinae (Fig. 64—66) Rather large subfamily containing solitary or gregarious endoparasites of larvae of Lepidoptera. The egg is deposited in the early instar larva, the parasites pupate outside the host larva. The presence of a more or less distinct subdiscoidella (Fig. 65) and the sometimes comparatively high insertion of the metasoma are remarkable, but are probably old plesiomorphous characters; also the larvae are in several aspects (e.g., the shape of the mandibles) rather slightly derived. Some apomorphous characters are the small second cubital cell (or cuqu 2 absent), posterior half of nervellus more or less weakly pigmented; the absence of the occipital carina; the short radial cell (Fig. 65); and the second metasomal tergite often has a more or less chevron-shaped impression (Fig. 63). Meteorideinae (Fig. 63, 67, 68, 70, 73) Small subfamily, solitary or gregarious endoparasites of larvae of Lepidoptera. According to Capek (1970: 859) Acampsis Wesmael and Sigalphus Latreille may deposite their eggs in the egg of the host and pupate outside the host larva, while Meteoridea Ashmead pupates inside their host pupa. This group shows many plesiomorphous characters in both the larval and the adult morphology. Some of the few apomorphous characters are its endoparasitism, the comparatively wide ovipositor sheaths, the more or less developed dorsal carinae and deep laterope of the first metasomal tergite. Cheloninae (Fig. 69, 71, 72, 74) Rather large subfamily, which consists of solitary endoparasites of larvae of Lepidoptera. The egg is deposited in the egg of the host. The delicate cocoon is attached to the remains of the host larva. This group possesses many apomorphous characters, e.g., the metasomal carapace, the short and high radial cell, the presence of the postpectal carina, the often densely setose eyes, the basally slender mandibles of the final instar larvae and the often indistinct hypostomal suture. The deposition of the egg in the host egg is clearly an adaptation to the habits of its hosts, which live at more or less hidden places. Acaeliinae (Fig. 61, 62) Small subfamily, its members closely resemble the Miracini of the Microgasteri- nae (cf. Fig. 62 with Fig. 55). They share also their hosts, viz., both are endo- parasites of larvae of leaf-mining Lepidoptera, also the pupation is in the cocoon of the host; this may account for their resemblance. The larvae are distinguishable from the larvae of the Microgasterinae by the presence of a mandibular base (cf. 48 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 27, 28 in Capek, 1970). The flattened shield formed by the three basal tergites of the metasoma is peculiar, they are fused almost invisibly (Fig. 61). Some other apomorphous characters are the strongly developed transverse carina of the propodeum, the position of the cuqu 1 and r 1 and the reduction of several other veins, e.g. the nervellus (Fig. 62). Macrocentrinae (Fig. 75—77, 81, 86) Subfamily of moderate size, containing solitary or gregarious endoparasites of larvae of Lepidoptera. The parasite pupates outside the host. Polyembryony is known of the gregarious species, of which the specimens spin together in a common web. The most strikingly apomorphous character of this group is the toothed trochantellus (Fig. 77), others are the specialized depressed area at the second tergite laterally; the high attachment of the metasoma, the reduction of the’ occipital carina, the more or less protruding middle lobe of mesoscutum (Fig. 76) and the usually absent dorsal carinae of the first metasomal tergite (Fig. 77). Hybrizontinae (Fig. 1—4) Small subfamily with one genus, Hybrizon Fallen, which contains endoparasites of larvae of several ant-genera. The naked pupae remain in the host nest. Pos- sesses many apomorphous characters, e.g., the short palpi (Fig. 3), the concavity behind the malar space, the slender legs, the peculiar venation of the wings (Fig. 4), the slender clypeus and probably completely grown together 2nd and 3rd tergites (Fig. 1). Helconinae (Fig. 78—80, 83—85, 87) Large subfamily of which almost all species are endoparasites of larvae of Coleoptera. The few exceptions may be Blacus Nees (some derived species have . been bred from larvae of Diptera) and Dyscoletes Haliday (bred from larvae of Mecoptera). This subfamily is difficult to characterize by apomorphous charac- ters, except for the endoparasitism. Formerly this name often served for a hetero- geneous group, clearly used as a rest group of remaining genera. Actually the. group of genera included in this paper can be characterized by several more or less “reticulate”’ occurring apomorphous characters. The most important apomor- phous characters in the Cenocoeliini are the highly inserted metasoma (Fig. 78), the concave frons, as occurs also in the Helconini and less distinctly in the Brachistini. The Helconini (Fig. 83) have usually the dorsal carinae distinctly developed, often reaching behind the middle of the first tergite as in the Blacini (Fig. 80) and Brachistini. A dorsope is present in the Blacini (Fig. 80) and Ceno- coeliini (Fig. 78), and less commonly in the Helconini (Fig. 83). The first discoidal cell is (sub-)sessile in the Helconini (p.p.), Brachistini (Fig. 84) and Blacini (but seldom shortly petiolate). The reduction of the veins aqu 1 + 2 and aqu’ occurs in the Helconini, Brachistini and Blacini, as also the reduction of the sculpture on the metasoma and of the precoxal suture. The reduction of veins leads to the loss C. van ACHTERBERG: The subfamilies of the Braconidae 49 of the cuqu 2 in the Brachistini (Fig. 84) and Blacini. In this subfamily the tribe Helconini is clearly less derived than other tribes; it possesses many plesio- morphous character-states. The final instar larvae have also many plesiomorphous characters (Fig. 9 in Capek, 1970), which supports the hypothesis about the evolution of the Braconidae from exoparasites of sheltered living larvae of Coleoptera. Ichneutinae (Fig. 82, 85, 88—90, 92) Rather small subfamily, consisting of endoparasites of larvae of sawflies (Ichneutini, Proteropini) and of lepidopterous leafminers (Muesebeckiini). The egg is deposited in the egg of the host but the larva developes after pupation of the host. The cocoon is formed inside the cocoon of the host. Rather hetero- geneous group, e.g., the final instar larvae of the Muesebeckiini have (at least in one genus) slender and smooth mandibles, stipital sclerite robust and hypostoma absent, while the Ichneutini have robust and toothed mandibles (as the Pro- teropini) and the cephalic sclerites present and robust (Fig. 33 in Capek, 1970). The cephalic sclerites are also present in the Proteropini but are very slender (Fig. 32, l.c.). Also the imagines differ considerably, but with the information available it seems better to unite the three tribes in one subfamily. The most important apomorphous characters are the absence of the occipital carina, and also more or less of the prepectal carina in the Muesebeckiini and Proteropini; the short and comparatively high radial cell in the Ichneutini and Proteropini (Fig. 88, 92) (which resembles the Cheloninae), the short ovipositor with its more or less widened sheaths (Fig. 89), the more or less strongly curved basal vein and the absence of aqu’ (Fig. 88, 92). Neoneurinae (Fig. 5— 7) Small subfamily, consisting of endoparasites of adults of worker ants; the development is in the gaster of the host. The cocoon is attached to the remains of the host. The most important apomorphous characters are the short palpi (Fig. 5), the peculiar venation of the wings (Fig. 7), the concave coxae dorso- apically and the strongly bent ovipositor. Aphidiinae (Fig. 93—96) Rather large subfamily, containing solitary endoparasites of aphid nymphs and adults. Praon Haliday and Dyscritulus Hincks emerge from their host and spin their cocoon directly beneath the aphid, whereas all other genera pupate within the mummified host skin, mounted on the surface of the plant. The cephalic structures of the larvae are like those of other Braconidae, according to Capek (1970, p. 848). The apomorphous characters of this subfamily are especially distinct in the larvae, because of the long and funnel-shaped spiracles without a division into atrium and closing apparatus, of the simple and smooth mandibles 50 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 and of the disk-shaped antennae. The apomorphous characters of the adults are the reduction of the wing venation (Fig. 93, 96), the specialized female genitalia and the weakly sclerotized metasoma (Fig. 95). Zelinae (Fig. 91, 100, 103, 104, 107, 111) Small subfamily, which contains solitary endoparasites of larvae of Lepidoptera. The pupation is outside the host larva. The apomorphous character of the larvae in respect to the Orgilinae is the slender stipital sclerite; Zemiotes Foerster has also smooth mandibles and Zele Curtis has the hypostomal spur absent. The adults have few apomorphous characters, e.g., the first metasomal tergite is slender and without dorsal carinae (Fig. 103), even petiolate in Zemiotes Foerster (Fig. 111). Orgilinae (Fig. 101, 102, 106) Small subfamily containing species with often many apomorphous characters. They are endoparasites of larvae of Lepidoptera, which pupate outside the host larva; the emergence opening of the cocoon is irregular. The larvae are character- ized by the absence of the hypostoma (Fig. 16, 17 in Capek, 1970). Apomorphous characters of the adult parasites are the absence of the aqu’ (Fig. 10}, 102); the more or less reduced metapleural flange (Fig. 106) and occipital carina dorsally; the more or less small radial cell of fore wing and anal lobe of the hind wing (Fig. 101, 102); cuqu 2 absent (Fig. 101) or if present, forming a small second cubital cell (Fig. 102). Euphorinae (Fig. 97, 99, 105, 108—110, 112—119) Large and diverse subfamily, consisting of solitary or gregarious endoparasites of larvae of Lepidoptera and Coleoptera (Meteorus Haliday), of solitary or gregarious endoparasites of adult Coleoptera (and more seldom of their larvae), (e.g., Perilitus Nees s.l., Ropalophorus Haliday in Curtis, Cryptoxilos Viereck, Stre- blocera Westwood), of adult parastic and aculeate Hymenoptera (Syntretus Foers- ter, the only exception to the rule that Braconidae are primary parasites), of adult Neuroptera (Chrysopophthorus Goidanich), of nymphal and adult Heteroptera (Wesmaelia Foerster, Aridelus Marshall, Leiophron Nees, Holdawayella Loan) and of nymphal and adult Psocoptera (Leiophron Nees). In Meteorus Haliday the cocoon of some species hangs from a long thread. The most important apomorphous character of the larvae are the smooth and short mandibles. The apomorphous characters of the adults are rather “‘reticulate”’ as in the Helconinae, e.g., the wing venation is reduced in the Cosmophorini (Fig. 105), the Euphorini (Fig. 118, 117, 119) and to a lesser degree in the Centistini, the parasitism of adult insects also occurs in these tribes; a dorsope occurs in the Centistini (p.p.), Euphorini (p.p.) (Fig. 110) and in the Meteorini (p.p.) (Fig. 112, 113); the notauli and the precoxal suture are reduced in the Centistini and in the Euphorini; the comparatively wide ovipositor sheaths, together with a more or less robust and bent ovipositor in the Centistini (Fig. 109) and Euphorini (p.p.; Fig. 108), and the C. VAN ACHTERBERG: The subfamilies of the Braconidae SI spiracles of the first metasomal tergite are situated in the middle of the tergite or behind the middle in the Euphorini (Fig. 110, 114, 116) and in the Meteorini (p.p.; Fig. 112, 113). The possible relations between the subfamilies are depicted in Fig. 123. The following groups may be recognized: Group A: The old “‘cyclostomes” of Wesmael, together with the later formed subfamilies Telengainae and Mesostoinae. They share the hypoclypeal depression, the apically more or less concave clypeus (Fig. 27, 28, 35, 41, 12, 37, 121) and the more or less flattened first metasomal tergite (Fig. 14, 22, 26, 30, 33, 38, 40). Group B: Specialized endoparasites of larvae of Diptera with the pupation in the host-puparium. The larvae have smooth, sickle-shaped mandibles and the labial sclerite absent or at least broadly interrupted ventrally. The adults have the prepectal carina absent and a more or less oval metasoma (Fig. 53). Group C: Specialized endoparasites of larvae of Lepidoptera; the egg is deposit- ed in the eggs of the host or in the early instar larvae. The larvae of the Acaeliinae and of the Microgasterinae-Cardiochilini have the tips of the mandibles bifid; the larvae of the Microgasterinae and of the Cheloninae have mandibles without or only with a small, scarcely differentiated base. The adults often have the eyes densely setose, the radial vein is often shortened (Fig. 54, 55, 57, 58, 62, 74) and the first metasomal tergite is more or less flattened in the Microgasterinae and Acaeliinae (Fig. 56—61). Group D: Specialized endoparasites of larvae of Lepidoptera; the larvae share the robust mandibles with a long, toothed blade (but the blade is intermediate in Meteoridea Ashmead). The imagines have the radial cell rather small (Fig. 65, 67, 73) and the laterope deep and usually large (Fig. 68, 70). Group E: Specialized endoparasites of larvae of Lepidoptera; the larvae have the labial sclerite transverse and the epistomal arch and hypostoma are absent. The adults share the tendency to loose the dorsal carinae of the first metasomal tergite. Group F: Generally very specialized endoparsites, especially of adult insects. The larvae have toothless, more or less wedge-shaped mandibles. The adults share the tendency to have the spiracle of the first metasomal tergite situated near the middle (Fig. 95, 109, 110, 112—114, 116); the wing venation is often very specialized (Fig. 7, 93, 96, 105, 117—119). The position of the Hybrizontinae within the Braconidae is rather uncertain, the situation of the spiracle near the middle of the tergite (Fig. 1) it shares with group F, but this may be a convergential development. The Ichneutinae, Macrocentrinae, and Helconinae form separate groups on their own. The larvae of the Ichneutinae show some similarity with the larvae in Group D, but the mandibles have a more or less developed triangular base. The adults also show some similarity, e.g., the reduction of the wing venation and the flattened first metasomal tergite in several groups. The larvae of the Macrocentrinae have a transverse labial sclerite with two processes ventrally, and polyembryony occurs in the genus Macrocentrus Curtis. The adults differ from the Helconinae by the shape of the trochantellus (Fig. 77), 52 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 of the first and second metasomal tergites of the mesosoma (Fig. 76), and of the head (Fig. 86). Finally the Helconinae; both in biology and morphology a slightly derived group. The larvae have wedge-shaped and toothed mandibles with a more or less distinctly differentiated base and the adults show the tendency to develop long dorsal carinae and distinct dorsope (Fig. 78, 80, 83). ACKNOWLEDGEMENTS I wish to express my sincere thanks to Dr. J. Decelle (Tervuren) and Dr. J. Papp (Budapest) for the loan of types and to Mr. T. Huddleston (London) for the correction of the English text. LITERATURE Achterberg, C. van, 1975. A new genus, Mesostoa gen. nov., from W. Australia, belonging to a new sub- family (Hym., Braconidae). — Ent. Ber., Amst. 35: 158-160, Fig. 1-8. ———, 1976. A revision of the tribus Blacini (Hym., Braconidae, Helconinae). — Tijdschr. Ent. 118 (7): 159-322, Fig. 1-476. Ashmead, W. H., 1900. Classification of the Ichneumon flies of the superfamily Ichneumonoidea. — Proc. U.S. nat. Mus. 23: 1-220. Capek, M., 1970. A new classification of the Braconidae (Hym.) based on the cephalic structures of the final instar larvae and biological evidence. — Can. Ent. 102 (7): 846-875, Fig. 1-58. ———, 1973. Key to the final instar larvae of the Braconidae (Hym.). — Acta Inst. forest. zvol.: 259- 268, | Fig. Cushman, R. A., 1923. A new subfamily of Braconidae (Hym.) from termite nests. — Proc. ent. Soc. Wash. 25 (2): 54-56, Fig. 1-3. Eady, R. D., & J. A. J. Clark, 1964. A revision of the genus Macrocentrus Curtis (Hym., Braconidae) in Europa, with description of four new species. — Entomologist’s Gaz. 15: 97-127, Fig. 1-92. Fahringer, J., 1925-28. Opuscula braconologica, Palaearktischen Region, 1: 1-606, Fig. 1-12, | table. ———, 1936. Uber einige merkwirdige und seltene Hymenopteren-Gattungen aus Afrika. — Fest- schrift 60. Geburtstage Prof. Dr. Embrik Strand 1: 568-590 + Fig. 1-27. Foerster, A., 1862. Synopsis der Familien und Gattungen der Braconen. — Verh. naturh. Ver. preuss. Rheinl. 19: 224-288, 1 Fig. Gauld, I. D., & T. Huddleston, 1976. The nocturnal Ichneumonoidea of the British Isles, including a key to genera. — Entomologist’s Gaz. 27: 35-49, Fig. 1-20. Griffiths, G. C. D., 1964. The Alysiinae (Hym., Braconidae) parasites of the Agromyzidae (Dipt.). |. General questions of taxonomy, biology, and evolution. — Beitr. Ent. 14: 823-914, Fig. 1-38, 2 graphs, 2 tables. Hedgvist, K.-J., 1955. Studien über Braconiden. I. — Ent. Tidskr. 76 (2-4): 92-98, Fig. 1-2. Marsh, P. M., 1963. A key to the Nearctic subfamilies of the family Braconidae (Hym.). — Ann. ent. Soc. Am. 56: 522-527, Fig. 1-29. ———, 1971. Keys to the Nearctic genera of the families Braconidae, Aphidiidae and Hybrizontidae (Hym.). — Ann. ent. Soc. Am. 64: 841-850, Fig. 1-11. Marshall, T. A., 1885. Monograph of British Braconidae, 1. — Trans. ent. Soc. London (1): 1-280 + 6 plates. ———, 1891. Les Braconides. In: André, E. (ed.), 1891-96. Species des Hyménoptères d'Europe et d’Algérie, 5: 1-628 + 19 plates. Mason, W. R. M., 1973. Recognition of Zemiotes (Hym., Braconidae). — Proc. ent. Soc. Wash. 75 (2): 213-215. Muesebeck, C. F. W., & L. M. Walkley, 1951. Family Braconidae. In: C. F. W. Muesebeck et al. (ed.), Hymenoptera of America north of Mexico, Synoptic catalogue. — Agriculture Monogr. 2: 90- 184. C. VAN ACHTERBERG: The subfamilies of the Braconidae 53 Nixon, G. E. J., 1938. Notes on the taxonomy and synonymy of Zele Curtis and Macrocentrus Curtis (Hymenoptera, Braconidae). — Bull. ent. Res. 29: 415-424, Fig. 1-2. Shenefelt, R. D., & C. F. W. Muesebeck, 1957. Ashmead’s Meteoridea (Hym., Braconidae). — Proc. ent. Soc. Wash. 59 (3): 129-134, Fig. 1-4. Szépligeti, G. V., 1904. Hymenoptera, Fam. Braconidae. In: Wytsman, P. (ed.), 1902-32. Genera Insec- torum, 22: 1-253 + 32 Fig. Tobias, V. I., 1962. A new subfamily of braconids (Hym., Braconidae) from Middle Asia. — Trudy zool. Inst. Leningr. 30: 268-270, Fig. 1-4. (Russian) ———, 1967. A review of the classification, phylogeny and evolution of the family Braconidae (Hym.). — Ent. Obozr. 46 (3): 645-669, Fig. 1-43. (Russian) ———, 1971. Review of the Braconidae (Hym.) of the USSR. — Trudy zool. Inst. Leningr. 54: 156- 268, Fig. 1-112. (Russian) Tobias, V. I., & G. P. Dudarenko, 1974. General trends in evolution of Braconidae (Hym.) abdomen. — Vest. Zool. 1974 (3): 65-72, Fig. 1-2. (Russian) Watanabe, C., 1937. A contribution to the knowledge of the braconid fauna of the Empire of Japan. — J. Fac. Agric. Hokkaido Univ. 42 (1): 1-188 + i-iv + Fig. 1-15. ———, 1968. Notes on the genus Cosmophorus and Orgilus in Japan with description of a new species (Hym., Braconidae). — Insecta matsum. 31: 1-6 + Fig. 1-11. ———, 1969. Notes on the genera Zele Curtis and Xiphozele Cameron with special reference to the species in Japan (Hym., Braconidae). — Proc. ent. Soc. Wash. 71 (3): 318-328, Fig. 1-9. Wesmael, C., 1835. Monographie des Braconides de Belgique. — Nouv. Mém. Acad. sci. R. Bruxelles 9: 1-252 + 1 plate + 2 tables. 54 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 1—4, Hybrizon buccata (de Brébisson), 9, Netherlands, Meijendel. 1, basal half of metasoma, lateral aspect; 2, fore leg, lateral aspect; 3, head, frontal aspect; 4, wings. Fig. 1, 3: 2.1 times scaleline; Fig. 2, 4: scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 55 1.0 mm Fig. 5—7, Neoneurus auctus (Thomson), 9, Lappland, Enontekiò. 5, head, frontal aspect; 6, fore leg, lateral aspect; 7, wings. Fig. 8, Ypsistocerus manni Cushman, 9; habitus, lateral aspect. Fig. 9, 10, Ypsistocerus vestigialis Cushman. 9, head, frontal aspect; 10, mouth parts, ventral aspect. Fig. 5: scale- line; Fig. 6, 7: 1.2 times scale-line; Fig. 8: 0,5 times scale-line. Fig. 8-10 after Cushman, 1923 56 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 FRONTAL ASPECT / / # LATERAL ASPECT Fig. 11—18, Mesostoa compressa Van Achterberg, 9, holotype. 11, habitus, lateral aspect; 12, head, frontal aspect; 13, wings; 14, abdomen, dorsal aspect; 15, mesonotum, dorsal aspect; 16, hind leg, lateral aspect; 17, antenna, frontal aspect; 18, head, dorsal aspect. Fig. 11, 13, 16, 17: scale-line; Fig. 12, 14, 15, 18: 1.5 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae SY Fig. 19-21, 24, Histeromerus mystacinus Wesmael, 9, Netherlands, Asperen. 19, wings; 20, fore leg, anterio-lateral aspect; 21, detail of fore tibia, posterio-lateral aspect; 24, hind tarsus, lateral aspect. Fig. 22, 23, Bracon urinator Fabricius, 9, Switzerland, Saas-Fee. 22, first metasomal tergite, dorsal aspect; 23, wings. Fig. 19: scale-line; Fig. 20, 24: 1.2 times scale-line; Fig. 21, 22: 2.5 times scale-line; Fig. 23: 0.6 times scale-line 58 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 1.0 mm Fig. 25, 26, Histeromerus mystacinus Wesmael, 9, Netherlands, Asperen. 25, head, lateral aspect; 26, propodeum, first and second metasomal tergites, dorsal aspect. Fig. 27, Bracon urinator Fabricius, 9, Switzerland, Saas-Fee; head, frontal aspect. Fig. 28-30, Telengaia ventralis Tobias. 28, head, frontal aspect; 29, metasoma, lateral aspect; 30, metasoma, dorsal aspect. Fig. 25-27: scale-line; Fig. 28, 30: after Tobias, 1962; Fig. 29: after Tobias & Dudarenko, 1974 C. VAN ACHTERBERG: The subfamilies of the Braconidae 59 31 Fig. 31-33, Doryctodes imperator (Haliday), © , Netherlands, Cadier. 31, wings; 32, fore tibia, anterio- lateral aspect; 33, first and second metasomal tergites, dorsal aspect. Fig. 34-36, Dendrosoter protuberans (Nees), 9, Netherlands, Best. 34, head, dorsal aspect; 35, head, frontal aspect; 36, wings. Fig. 31, 33: 0.6 times scale-line; Fig. 32: 2.5 times scale-line; Fig. 34-36: scale-line 60 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 37, 38, Gnaptodon pumilio (Nees), 9, Netherlands, Waarder. 37, head, frontal aspect; 38, meta- soma, dorsal aspect. Fig. 39, 40, 43, Rhyssalus clavator Haliday, 9, Netherlands, Wijster. 39, fore tibia, posterio-lateral aspect; 40, basal half of abdomen, dorsal aspect; 43, wings. Fig. 41, 42, Rogas excavatus (Telenga), 4, Italy, Riva s. Garda. 41, head, frontal aspect; 42, head, dorsal aspect. Fig. 37-39: 2.5 times scale-line; Fig. 40-42: 1.2 times scale-line; Fig. 43: scale-line C. vAN ACHTERBERG: The subfamilies of the Braconidae 61 Fig. 44, Rogas unipunctator (Thunberg), 9, Netherlands, Oostkapelle; wings. Fig. 45, 51, Polemochartus liparae (Giraud). 45, 3, Netherlands, Haaren, three basal segments of metasoma, dorsal aspect; 51, ©, Netherlands, Eindhoven, mandible, ventro-lateral aspect. Fig. 46, Rogas drymoniae Watanabe, 1937; habitus, dorsal aspect. Fig. 47-49, Tanycarpa punctata Van Achterberg, 9, holotype. 47, propodeum, first and second metasomal tergites, dorsal aspect; 48, head, frontal aspect; 49, detail of mandible, lateral aspect. Fig. 50, Chaenusa bergi (Riegel), 9, paratype; head, frontal aspect. Fig. 44, 45: 0.5 times scale-line; Fig. 46: 0.1 times scale-line; Fig. 47, 51: 2.5 times scale-line; Fig. 48: 1.2 times scale-line; Fig. 50: 1.8 times scale-line; Fig. 49: 3.2 times scale-line 62 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 52, 53, Opius (Apodesmia) polyzonius Wesmael, 9, Netherlands, Asperen. 53, head, frontal aspect; 54, metasoma, dorsal aspect. Fig. 54, Apanteles falcator (Ratzeburg), 9, Netherlands, Waarder; wings. Fig. 55, Mirax cremastobombyciae (Fullaway), 9, holotype; wings. Fig. 52: 2.5 times scale-line; Fig. 53, 54: scale-line; Fig. 55: 1.7 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 63 Fig. 56, Mirax cremastobombyciae (Fullaway), 9, holotype; basal half of abdomen, dorsal aspect. Fig. 57, Microplitis theretrae Watanabe, after Watanabe, 1937; habitus, dorsal aspect. Fig. 58, Micro- gaster takeuchii Watanabe, after Watanabe, 1937; habitus, dorsal aspect. Fig. 59, Apanteles falcator (Ratzeburg), 9, Netherlands, Waarder; metasoma, dorsal aspect. Fig. 60, Cardiochiles japonicus Wata- nabe, after Watanabe, 1937; habitus, dorsal aspect. Fig. 61, Acaelius spec., 9, Netherlands, Oost- voorne; metasoma, dorsal aspect. Fig. 56, 61: 2.5 times scale-line; Fig. 59: scale-line; Fig. 57, 58: 0.2 times scale-line; Fig. 60: 0.1 times scale-line 64 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 62, Acaelius spec., 9, Netherlands, Oostvoorne; wings. Fig. 63, 65, Earinus nitidulus (Nees), 9, Netherlands, Schaarsbergen. 63, first and second metasomal tergites, dorsal aspect; 65, wings. Fig. 64, Meteoridea japonensis Shenefelt & Muesebeck, after Shenefelt & Muesebeck, 1957; wings. Fig. 62: | 4.0 times scale-line; Fig. 63, 65: scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 65 Fig. 66, Braunsia matsumurai Watanabe, 9, after Watanabe, 1937; habitus, dorsal aspect. Fig. 67, 68, Acampsis alternipes (Nees), 9, Netherlands, Den Haag. 67, wings; 68, habitus, lateral aspect. Fig. 69, Chelonus (Microchelonus) tosensis Watanabe, after Watanabe, 1937; 9, habitus dorsal aspect; à, apex of metasoma, apical aspect. Fig. 70, Siga/phus irrorator (Fabricius), 9, Netherlands, Naaldwijk; meta- soma, lateral aspect. Fig. 66, 0.1 times scale-line. Fig 67, 68, scale-line. Fig. 69, 0.4 times scale-line. Fig. 70, 0.6 times scale-line; t = tooth 66 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 71, 72, 74, Ascogaster instabilis Wesmael, 9, Netherlands, Asperen. 71, mesosoma, ventro-lateral aspect; 72, metasoma, lateral aspect; 74, wings. Fig. 73, Sigalphus irrorator (Fabricius), 9, Netherlands, Naaldwijk; wings. Fig. 71, 72, 74: scale-line; Fig. 73: 0.5 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 67 Fig. 75—77, Macrocentrus thoracicus (Nees), 9, Netherlands, Meijendel. 75, wings; 76, mesoscutum, lateral aspect; 77, propodeum and first metasomal segment, lateral aspect. Fig. 78, 79, Cenocoelius analis (Nees), 9, Netherlands, Wijster. 78, propodeum and first metasomal segment, lateral aspect; 79, wings. Fig. 80, Blacus (Ganychorus) striatus Van Achterberg, 9, holotype; propodeum and first metasomal tergite, dorsal aspect. Fig. 75: 0.5 times scale-line; Fig. 76-79: scale-line; Fig. 80: 2.5 times scale-line 68 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 81, 86, Austrozele brevicaudis (Szépligeti), 2, lectotype. 81, wings; 86, head, frontal aspect. Fig. 82, Proterops nigripennis Wesmael, 9, Netherlands, Kralo; basal half of metasoma, dorsal aspect. Fig. 83, Elachistocentrum similis (Szépligeti), 9, lectotype; propodeum and first metasomal tergite, dorsal aspect. Fig. 84, Eubazus (Brachistes) lapponicus (Thomson), 9, lectotype: wings. Fig. 85, Ichneutes spec., 9, Netherlands, Waarder; head, frontal aspect. Fig. 81: 0.4 times scale-line; Fig. 82: scale-line; Fig. 83: 1.8 times scale-line; Fig. 84: 0.7 times scale-line; Fig. 85: 1.2 times scale-line; Fig. 86: 0.8 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 69 5 D TT 2a, È S Fig. 87, Eubazus (Foersteria) tibialis (Haliday), 3, Netherlands, Wijster; propodeum and first meta- somal segment, lateral aspect. Fig. 88, 89, /chneutes spec., 9, Netherlands, Waarder. 88, wings; 89, habitus, lateral aspect. Fig. 87-89: scale-line 70 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 90, 92, Proterops nigripennis Wesmael, 3, Netherlands, Kralo. 90, head, frontal aspect; 92, wings. Fig. 91, Charmon extensor (Linnaeus), 9, Netherlands, Naardermeer; propodeum and first metasomal segment, lateral aspect. Fig. 90: 2.5 times scale-line; Fig. 91: scale-line; Fig. 92: 0.6 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae ZU 99 Fig. 93—95, Ephedrus plagiator (Nees), 9, Netherlands, Maastricht. 93, wings; 94, head, frontal aspect; 95, metasoma, lateral aspect. Fig. 96, Aphidius ervi Haliday, 9, Netherlands, Asperen; wings. Fig. 97, Cosmophorus cembrae Ruschka, after Hedqvist, 1955; anterior part of head, lateral aspect. Fig. 98, Cosmophorus klugi Ratzeburg, after Watanabe, 1968; head, dorsal aspect. Fig. 99, Cosmophorus regius Niezabitowski, id. Fig. 93, 95, 96: scale-line; Fig. 94: 2.5 times scale-line 72 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 100, Charmon extensor (Linnaeus), 9, Netherlands, Naardermeer; wings. Fig. 101, Orgilus laevigator (Nees), 9, Netherlands, Waarder; wings. Fig. 102, Microtypus wesmaeli Ratzeburg, 9, Netherlands, Crailo: wings. Fig. 100, 102: 0.6 times scale-line; Fig. 101: scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 73 10 mm T © Fig. 103, 104, Zele infumator Lyle, 9, Netherlands, Crailo. 103, Propodeum, lateral aspect; 104, wings. Fig. 105, Cosmophorus regius Niezabitowski, after Watanabe, 1968; wings. Fig. 106, Microtypus wesmaeli Ratzeburg, 9, Netherlands, Crailo; propodeum, lateral aspect. Fig. 103, 104: 0.6 times scale-line; Fig. 106: scale-line 74 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig. 107, Zemiotes deceptor (Wesmael), 9, Netherlands, Wageningen; wings. Fig. 108, Leiophron (Leio- phron) apicalis Haliday, 9, Netherlands, Wijster; apex of metasoma, lateral aspect. Fig. 109, Ancylo- centrus ater (Nees), 9, Netherlands, Waarder; metasoma, lateral aspect. Fig. 110, Streblocera macro- scapa (Ruthe), 9, Netherlands, Waarder; first metasomal tergite, dorsal aspect. Fig. 107: 0.6 times scale-line; Fig. 108-110: 2.5 times scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae 75 ee seal oo tne 7 1.0mm = au 111 Fig. 111, Zemiotes deceptor (Wesmael), ©, Netherlands, Wageningen; metasoma, dorsal aspect. Fig. 112, Meteorus cf. sulcatus Szépligeti, 9, Netherlands, Wijster; metasoma, dorsal aspect. Fig. 113, Meteorus ictericus (Nees), 9, Netherlands, Waarder; metasoma, dorsal aspect. Fig. 114, Syntretus cf. conterminus (Nees), 9, Netherlands, Waarder; first metasomal tergite, dorsal aspect. Fig. 115, Perilitus (Microctonus) cf. deceptor Wesmael, 9, Netherlands, Putten (G.); head, frontal aspect. Fig. 116 Chrysopophthorus spec., 9, Costa Rica, Turrialba; first metasomal tergite, dorsal aspect. Fig. 111: 0.6 times scale-line; Fig. 112, 113, 116: scale-line; Fig. 114, 115: 2.5 times scale-line 76 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 Fig.117, Perilitus (Microctonus) cf. deceptor Wesmael, 9, Netherlands, Putten (G.); wings. Fig. 118, Leiophron (Leiophron) apicalis Haliday, 9, Netherlands, Oostvoorne; wings. Fig. 119, Chrysopophthorus spec., 9, Costa Rica, Turrialba; wings. Fig. 117, 118: 1.2 times scale-line; Fig. 119: scale-line C. VAN ACHTERBERG: The subfamilies of the Braconidae i Fig. 120—122, Pseudodicrogenium monstrosum Fahringer. 120, fore wing, paralectotype; 121, head, frontal aspect, lectotype; 122, head, lateral aspect, lectotype. Fig. 120, 122: scale-line; Fig. 121: 2.0 times scale-line 78 TIJDSCHRIFT VOOR ENTOMOLOGIE, DEEL 119, AFL. 3, 1976 \ [ea] = \ ta 78 Hi Fe \ => m = = Er = <5 x 5 U m E = oe A 5 È = fa RG : = (cal Ta n = = > i H N Csa E E ' A N ~ h ner BL x a m = a x N \ m N A (©) ‘ ©) x 4 a = Pi \ es = S LUI = EN NI) 2 zZ \ 5 N Sne a 3 A k = hi = Q fea] \ pe ; = = \ & = H à E = = È I = O 9 a ; H © "== . A & 2 A [ne] 4 | ic] \ oO = ss los) = H = eet So ea = \ a = fal = ES A à \ La < 0 \ =x E 5 o EN D Sd \ bet 5 DEL \ i SR a A , © m \ ar SIMS a a Ù f x 5 ato \ / Mo EF m TO È [ea] xt DRE = n \ ve / a = MU Ws \ 7 Le H x nn DI H = \ e SAS = i = > \ bl x ! H DENE (dp) in N ' (©) x x > I E x 4 N J ‘a > << Se rare) AN DS ON eee ees = = SII i È D p H \ E \ Hi w N N HI A o ~ aD & TE iI \ «x E ‘a = See \ 5 = . \ \ H oO \ o î