BULL ETIN DE L'INSTITUT ROYA L DES SCIE 1CES NATURELLES DE BE LGIQUE ENTOMOLOG IE. 70: 13-31 . 2000 BULL ETI N VA I-l ET KON INKLIJ K BELGISCH I STITUUT VOOR NATUU RWETENSCHAPPEN E TOMOLOGIE. 70:, 13-31,2000

Flight muscle developtnent and dispersal in the life cycle of carabid beetles: patterns and processes

by Konjev DESENDER

Abstract Les deviations significatives observees pour une proportion plus im­ portante de coleopteres avec des muscles de vols fonctionnels chez des femelles non matures en comparaison avec celles matures indiquent un Seasonal patterns of fli ght muscle development are documented fo r 27 tel syndrome chez Ia plupart des especes de Carabidae etudiees. Ceci carabid species, collected from di fferent habitats. The ph enology or s'applique surtout aux especes qui emergent aI a fin du printemps, juste timing of their life cycle is compared to t he p resence o f a functional avant leur periode de reproduction en ete-automne. Le pourcentage de fli ght apparatus during complete year cycles. The majo ri ty of the femelles matures avec des muscles alaires fonctionnels differe entre investigated ground beetles have never b een stud ied b efore in this espi:ces et peut etre utilise comme une mesure d' un "'oogenesis-fli ght respect. In most species there is a di stinct seasonal pattern of fli ght syndrome" ma ins deterministe. L'analyse globale de tous les resultats muscle functionality. Some species, with similar life cycle timing a nd/ suggere que les caracteristiques suivantes interviennent comme fac­ or similar habitat requirements, show the same pattern of fli ght muscle teurs expliquants les profil s o bserves au ni veau des muscles de vol: development. Inter- and intraspecific v ariability in maximal propor­ choix de !'habitat, evolution du cycle de vie, ainsi que d'autres eie" tions o f beetl es with functional flight muscles indicates fli ght muscle ments de Ia vie de l' insecte. Le developpement des muscles devol des dimorphism. The reproducti ve state of fe males (obtained through dis­ Carabidae semble di:s lors faire partie d' un ensemble de caracteres section of the ovaries) is used to test the generality of an "oogenesis­ coadaptes. fli ght syndrome" (trade-off between dispersal and reproduction), the null hypothesis being that functional fl ight m uscul ature and ripe ova­ Mots-clefs: Carabidae, pouvoir de dispersion, cycle de vie, developpe­ ri es occur independently. Observed signi ficant d eviations towards a ment des muscles alaires, oogenesis-flight syndrome, reproduction, hi gher proportion of beetles with fi.mctional fl ight muscles in unripe as dimorphisme alaire, adaptations, caracti:res coadaptes compared to ripe females indicate such a syndrome in most of the species studied. It is especially true for species emerging during late spring before their summer-autumn reproduction. The proportion of ripe females with functional fli ght muscles differs between species and can be used as a measure for a less deterministic version of an Introduction "oogenesis-flight" syndrome. A complex of factors must play a role in the expression of fli ght muscle development. Integrati on of all results s uggests habitat choice, evolu tion of life cycle timing and other [n many insect groups, species are known to show poly­ life hi story traits as ultimate factors responsible for the observed morphisms affecting their fl ight ability. The most obvious patterns of flight muscle fun ctionali ty. Flight muscle development in examples are variations in wing and fli ght muscle devel­ ground beetl es therefore seems to be part of a suite of coadapted traits. opment (HARR ISON, 1980). Wing polymorphisms are un­ Key wonls: Carabidae, dispersal p ower, life cycle, fli ght muscle der genetic and environmental control (HARRISON, 1980; development, oogenesis-flight syndrome, reproduction, wing dimor­ ROFF, 1986). phism, adaptations, sui te of coadapted t ra its Numerous observations indicate that in many insect species migration is limited t o t he post-teneral, pre-re­ productive period. These observations have led to define Resume the " oogenesis-flight syndrome" as a characteristic of insect migration (JOH NSON, 1969; for Coleoptera, see f. e. Les pro til s saisonniers du developpement des muscles alai res sont Ll NDERS et a\., 1995, MUDA et a\. , 198 1, RANKIN et a\. , reportes pour 27 especes de Carabidae, recoltees dans des habitats differents. Durant des cycles annuels complets, Ia phenologie ou Ia 1994, TADA et al. , 1991) . The inherent assumption is that periode du cycle v ital est comparee pour chaque espece i1 Ia presence mi gration (through fli ght d ispersal by means of fu nc­ d ' un appa reil devol fonctionnel. La majorite des especes de Carabidae tional fl ight muscles and developed hind wi ngs) and concernees n 'a jamais ete etudiee SOLI S eel angle avant ce travail. Chez Ia plupart des especes, il y a un cycle saisonnier de Ia foncti onnali te des reproduction are alternating physiological states. This muscles alaires. Certaines es peces qui ont un cycle de vie e t/ou des view was strengthened by the fact that migration in many exigences environnementales similaires demontrent le meme profil cases is associated with or induced by conditions promo­ saisonnier au ni veau des muscles devol. Si l'on considi:re Ia periode du cycle de vie otl le developpement des muscles de vol est le plus ting a dult diapause (i.o. w. delaying reproduction), for important, on constate une vari abilite inter- et in traspecifique, ce qui example a s horter photoperiod, lower t emperatures or suggere un dimorphisme de muscles alai res. L 'eta! reproducti f d es poor f ood conditions (RANKIN et a\. , 1986). Migrating fe melles (obtenu apres dissection des ovaires) est utilise pour t ester insects moreover often are physiologically comparabl e Ia presence d' un " oogenesis-fl ight syndrom.e" (balance entre d isper­ sion et reproduction), !' hypothese nulle etant qu· une muscul ature alaire to cli apausing insects in their possession of hypertrophic fonctionnell c et Ia maturi te des ovaires apparaissent independamment. fat bodi es and immature ovaries. Studies on the physio- 14 Konjev DESENDER '' logy and endocrinology of flight muscle degeneration and production) and is related to habitat change (TIETZE, regeneration, in a few insect species, have suggested a 1963 ; VAN Hui ZEN, 1977). Flying females were always control by juvenile hormone (FAIRBAIRN & Y ADLOWSKI, unripe. According to DEN BOER eta!. ( 1980) some species I 997; ZERA et al., 1997). A strong negative correlation (Amara.familiaris, Anisodactylus binotatus, Calathus ro­ has commonly been observed between flight muscle mass tundicollis and Nebria brevicollis) deviate more or less and ovarian mass of insects (ROFF, I 986), suggesting that from an "oogenesis-flight" syndrome. In Nebria brevi­ the construction and maintenance of the flight apparatus col/is developing flight muscles were found in immature competes with egg production for a limited internal nu­ beetles (i.e. in spring) more than in reproducing beetles trient pool (ZERA & DENNO, 1997). The metabolic rates of (in autumn), while at least some flight activity was ob­ flying insects can be 20-100 times that of resting animals served during autumn too (N ELEM ANS, 1987). MATALIN and are among the highest known (RANKIN & BURCHSTED, ( 1994, 1997) and ZHANG et a!. ( 1997) studied Harpalus 1991 ). Nevertheless, several authors ( cf. GATEHOUS E & rujipes and came to the conclusion that flight was more ZHANG, 1995) have chall enged the assumption that the prevalent in (but not limited to) young beetles before oogenesis-flight syndrome is a general phenomenon. reproduction. MEIJER ( 1974) suggested variation in flight Ground beetles are well-known for their varying de­ behaviour of the small Bembidion varium (always posses­ gree of hind wing development. Some species are con­ sing functional flight muscles): in spring relatively more stantly winged, others show a wing polymorphism or unripe females were recorded flying as compared to wing dimorphism and some species always possess re­ reproducing ones. duced wings. Carabid beetles consequently exhibit a large Clearly, there is a considerable lack of empirical data amount of variation in their respective dispersal power. on most carabid beetles in this respect. To this end, we Wing development is largely under direct genetic control have dissected large numbers of ground beetles collected in the few ground beetle species studied so far (cf. Au­ during complete year cycles in a multitude of habitats and KEMA, 1986; DESENDER, 1989a; LINDROTH, 1946). About belonging to 27 different species (20 constantly macro­ three-quarters or as much as 280 carabid species from pterous, seven wing dimorphic or polymorphic). These Belgium are known to be constantly macropterous or full­ species can be roughly divided into spring breeders (adult winged. Nevertheless wing development and, as a con­ hibernators) on the one hand and summer-autumn bree­ sequence, fli ght capacity are extremely variable between ders (larval or larval/adult hibernators) on the other hand. winged species too, as has been shown on the basis of a In this paper, we will first look at seasonal patterns of biometric approach (DEN BOER et al., 1980; DESENDE R et fli ght muscle development by comparing the phenology a!., 1986; DESENDER, 1989b ). Many winged ground beet­ or timing of the life cycle in each species to the relative les moreover not necessarily possess functional flight occurrence of a functional flight apparatus. Secondly, the muscles. Below a certain value of relative wing size, reproductive state of females (obtained through dissec­ functional flight muscles apparently are only rarely ob­ tion of the ovaries) will be used to test the generality of served (D EN BOER et a!., 1980; DESENDER, 1989b ). an "oogenesis-fli ght syndrome", the null hypothesis Genetic studies on fli ght muscle development have not being that functional fli ght musculature and ripe ovaries yet been performed in ground beetles, with one excep­ occur independently. A significant deviation towards a tion: NELEMANS ( 1987) showed that there is no simple higher proportion of beetles with functional flight mus­ genetic basis for flight muscle development in Nebria cles in unripe fe males as compared to ripe females would brevicollis. However, genetic studies on other beetles be an indication for such a syndrome. The percentage of (two species of Scarabaeidae; T ADA et a!., 1993, I 994, ripe females with functional fli ght muscles can be used 1995) have demonstrated genetic fli ght muscle dimor­ as a measure fo r a less deterministic version of an " oo­ phism, besides environmental control of fli ght muscle genesis-fli ght" syndrome. Results for species with dif­ development. Seasonal changes in the proportions of ferent life cycle timing will finally be compared in an carabid beetles with functiona l flight musculature for a attempt to give ultimate explanations or underlying evo­ given species could suggest phenotypic plasticity and lutionary processes responsible for the observed variabi­ point to environmental conditions possibly influencing lity and timing of the presence of functional fli ght mus­ flight muscle development. Some studi es have shown that culature. proportions of beetles with flight capability differ be­ tween species. Seasonal aspects, however, usually were not included. Such data could nevertheless suggest Material and Methods whether there is some kind of dimorphism or polymorph­ ism in the determination of fli ght muscle fu nctionali ty. Two groups of indirect flight muscles, situated in the The current knowledge on seasonal variation in fli ght metathorax of beetles, are essential for fli ght activity muscle development as compared to adult life cycle of (Fig. I). The medio-dorsal longitudinal muscles ground beetles is limited to some five species only. (Fig. I B) bulge the metatergum outwards when contrac­ Amara plebeja, reproduci ng during spring and hiberna­ ting and provoke the downward stroke of the wings. The ting as adult, showed a strict "oogenesis-flight" syn­ lateral dorso-ventral muscles (Fi g. I A) have an antago­ drome: mi gration occurs during autumn (post-teneral), ni stic effect and fl atten the metatergum, in this way but also during spring (after hibernation and before Te- provoking the upward stroke of the wings. Flight muscle development in carabid beetles '' 15

muscle reduction, i.o.w. individuals with a non-functional fli ght apparatus. When available, replicate year samples from different sites were studied. The ovaries of about 3000 females belonging to the same 27 species (except Amara tibialis and Clivina.fossor) were dissected in order to define their reproductive state as compared to flight muscle functionality. The null hypothesis (functional flight muscles and ripe ovaries occur independently) was tested by means of a G-test of independence (SOKA L & ROHLF, 1981 ).

Results

I. Seasonal variation offlight muscle jimctionality in spring-reproducing ground beetles Figs. 1-2 - Functional (Fig. I) and degenerated indirect flight (adult hibernators) muscles ( Fig. 2; muscle fibres replaced b y adi­ pose tissue) in the metathorax of Pogonus chal­ Fig. 3 (A-Q) summarises data obtained for I 0 carabid ceus; I A: lateral dorso-ventral muscles, I 8: me­ species, reproducing mainly duri ng spring. Amara aenea dio-dorsal longitudinal muscles was studied in 6 populations (Fig. 3, G-L), Agonum muelleri (Fig. 3, M-N) and Pterostichus versicolor (Fig. 3, P-Q) in 2 populations. The remaining species were studied in a single population year cycle (Fig. 3, Diss.ection of carabid beetles reveals that these muscles A-F, 0). These l 0 species show their reproductive activ­ can be reduced to different degrees or even totally be ity (expressed as numbers caught p er month in pitfall replaced by ad ipose ti ssue (Fig. 2). Intra-specific differ­ traps) mainly during spring, resulting in larvae develop­ ences in the state of fli ght musculature are not necessarily ing during summer and the emergence of the new beetle restricted to different individuals, but may, in certain generation during autumn (teneral beetles, indicated by species, occur during the lifetime of even a single beetle. black columns). Table I summarises the results for fe­ In such a case flight muscul ature can differentiate, be males from the same spring-reproducing species, along resorbed or autolysed and then regenerate. with G-test results and sample sizes. We di ssected the meta thorax from about I 0.000 field­ The percentage of beetles with functional flight mus­ collected beetles belonging to 27 different carabid species cles shows no clear pattern of seasonal variation in Amara (nomenclature according to DESENDER et al. , 1995), oc­ .familiaris (Fig. 3 A) and Asaphidion curtum (Fig. 3 B). cutTing in different habitats in Belgium, in order to define Both are characteri sed by well-developed hind wi ngs the state of their flight muscles. From 13 species, a high (DES ENDER, 1989b). At any time during the adult life number of indi viduals could be checked: in these cases cycle of Amarajamiliaris about 25 to 50% of the indivi­ monthly fractions of beetles with functional flight mus­ duals show functionality of the fl ight apparatus (cf. DEN cles were plotted with their respective 95% confidence BoER et al., 1980). This carabid is known to occur in limits (WONNACOTT & WONNACOTT, 1977). Jn the re­ rather ephemeral habitats, for exampl e on open, recently maining species only the obtained fractions were plotted. created, sandy sites (our sampling site was a recent The timing of the life cycle was also illustrated for each motorway verge on sandy soil). Asaphidion curtum is species. Based on data from the same sampling s ites particularly common in light forest and forest clearings. (nearl y always from pitfall year cycle series, interpolated The continuous possession (at least in a number of in­ per month), frequency distributions were given for adult dividuals) of a functional flight apparatus in these species and teneral (newly emerged) beetles (in some cases also thus can be interpreted as an adaptation to unpredictable for larvae; in one case based on seasonal occurrence of changes in their habitat. There is no statistical deviation mean number of ripe eggs in females). Data from seven from independence between reproductive state and fl ight wing dimorphic o r wing polymorphic species are in­ muscle functionali ty (Table I), i.o. w. we do not observe a cluded: here, proportions of beetles with functional fl ight significant oogenesis-flight syndrome. muscles were calculated based on macropterous indivi­ Most other spring breeders, on the other hand, show a duals only. For two high-density populations of Bembi­ significant oogenesis-flight syndrome (Table I), although dion properans (a wing dimorphic species), the monthly somewhat less deterministic. Tn each species we observe proportion of observed macropterous beetles was also at least some reproducing fe males with functional fl ight calculated and plotted in order to look for possible sea­ muscles, but the reproductive period overlaps only partly sona l changes in these fractions too. with the period of fu nctional flight muscles. Anisodacty­ For the current paper we distinguished beetles with lus binotatus and Acupalpus jlavicollis individuals all well-developed flight muscles from all states of fli ght possess flight muscles during early spring (at the onset 16 Konj ev DESENDER ''

H S8 Amara familiaris Asaphidion curium B A ~ 38 u road verge on sand woodland B B "E 29 " R ~ 28 c c ~ 18 ~ 18 6 6 H H T T 100 188 90 se se p 72 p 78 E ee E 118 ~ 50 ~ 50 E <10 E <10 H S8 ! 29 A~ : s 18 ? ? 6 18 ? ? E ? ? ? E ? ? ? ? ?

2 s 4 5 8 9 18 II 12 2 3 5 6 8 9 18 II 12

18 14 H 16 H c u 12 D ~ 14 H B 12 Acupalpus ffavicollis B 18 E ~ Ill humid grassland R 8 c 8 c 6 A 6 A u u 4 4 6 6 2 H 2 H T T 189 91! 88 se p 78 p 78 E 69 E se ~ 58 ~ 50 E 49 E <10 H 38 H S8 ! 28 r 211 6 18 6 18 E ? ? ? ? ? E ?? ? ? ? ?

2 3 5 6 8 19 II 12 2 3 4 5 7 8 9 19 II 12

72 14 Amara tibialis Agonum dorsale N H 118 u 12 dune grassland u H E fields + dry pasture F 8 18 : 50 E R 8 i '"' c 6 c S8 A u 4 ~ 29 s H 2 ~ 18 T T 100 1ee 90 90 88 se p 78 p 79 E 88 E se ~ 50 ~ 50 E 49 E <10 N 38 T 28 ~ : A A 6 18 ? 6 18 E ? ? ? ?? E

2 3 4 5 8 9 10 II 12 2 3 4 5 6 7 8 9 19 II 12

Fig. 3 (A -Q) - Phenology of th e li fe cycle in spring breedi ng gro und beetles (upper fi gure; black colu mns: teneral beetles), compared to th e seasonal occ urrence of monthl y proportions of bee tl es wit h functiona l fl ight muscles (lower fi gure; with 95% c.i . for large sa mp les,?: months withou t data, arrows indica te zero-values based on large sa mp le sizes). A: Amara .fCnnifiaris (recent motorway ve rge on sandy soil ), 8: Asaphidion curium (woodland site), C: Acupafpu sjlavicoffis (hum id grass land), D: Anisodaclylus binotalus (hum id poor gras land), E: Amara tibialis (dun e grass land) , F: Agonum dorsafe (cu lti va ted fi elds and dry pasture). Flight muscle development·in carabid beetles I I 17

79 2SB H Amara aenea H u G u H ee pasture 299 A. aenea B 58 B poor grassland " "E E 158 R 4e R c 9e c 198 A A u 28 u G 6 58 H 19 H T T 189 189 89 89 88 88 p 79 p 78 E ee E 88 R R c 58 c 58 E 4e E ole H 30 H 38 T T A 28 A 28 6 10 6 18 E ? E ? ? ? ?

2 3 5 6 7 8 9 18 II 12 2 3 4 5 6 8 18 II 12

79 aenea 4e N A. N u ee relic dune grassland u A. aenea J 8 58 8 30 road verge on sand "E "E R 4e R 28 c 9e c A A u 28 u 19 6 G H 19 H T T 189 189 118 118 88 88 p 79 p 79 E 88 E ee R R c 58 c 58 E 4e E 4e N H 9e 30 T T 28 A 28 A G 18 ? ? 6 18 ? ? E ? ? ? E ?

2 3 4 5 9 19 II 12 2 3 4 6 7 9 19 II 12

48 A. aenea H H u city lawn K u 39 L 39 B B A. aenea "E "E R R 28 city lawn 29 c c A A 19 u 19 u 6 6 H H T T 189 188 118 99 88 81! p 79 p 79 E 88 E 88 R R c 58 c 58 E ole E 4e N 39 N T 39 28 T A A 29 6 19 ? ? ? ? 6 19 ? ? E ? ? E ? ? ? ? ?

2 3 5 6 8 19 I I 12 2 3 4 5 6 8 9 19 11 12

Fig. 3 (continued) - G-L: Amara aenea (G: pasture, 1-1 : dry poor grassland , l: relic dune grassland, J: recent motorway verge on sandy so il , K-L: lawns in city of Ghent). 18 Konj ev DESENDER ''

118 289 Agonum muelleri A. muelleri N M H N u grassland road verge u 58 pasture B 158 B 48 "E "E R R tee 311 c c A A 28 u 58 u G G Ill H H T tee T tee Sllil Sllil se 118 p 7l! p 71l E ee E 611 R R c 58 c 58 E 48 E 48 H 311 N 311 T T A 28 A 28 G 19 ? ? 6 Ill E ? E ?

2 3 5 e 7 8 9 18 II 12 2 3 5 6 7 8 9 19 II 12

Ill 188 g H H u 98 p u 8 0 88 Pt. versicolor 7 B 71l dry poor grassland "B "E E 6 R ee R 5 58 c 48 c A A 3 u 311 u 6 211 6 2 H 18 H T T tee tee Sllil 98 88 se p 71l p 78 E E Be ea R R 58 c 58 c 48 48 E E N H 311 311 T T 211 A 211 A 6 Ill 6 18 ? ? ? ? ? E ? ? ? E ? ? ?

II 12 2 3 5 e 7 8 9 Ill II 12 2 3 4 5 6 8 9 Ill

4811 N u 358 Q Pt. versicolor 389 8" dry poor grassland E 258 R 289 c A !58 u IIIII G H 58 T IIIII Sllil se p 79 E 60 R c 58 E 48 H 311 T A 211 G Ill ? ? E ?

3 4 5 8 7 8 9 18 II 12

Fig. 3 (conti nued) - M-N: Agonum muelleri (M: grassland on m otorway verge, N : pasture), 0: Harpalus a/]inis (dry pasture); P-Q: Pterostichus versicolor (dry poor grasslands). I I Flight muscle development in carabid beetles 19

Table I - The oogenesis-fl ight syndrome in ground beetles: fli ght muscle development and reproductive state of ovaries in females of the investigated ground beetl es; n = number of dissected fe males; R = ovaries with ri pe eggs, 0 = ovaries immature o r only w ith corpora lutea (spent, after egg-laying), + = functional fl ight muscles, - = degenerated flight muscles; G-test-stati sti c: test of independence after comparing [number with fli ght muscles and unripe or spent ovari es to the total with unripe or spent ovaries] with [number with flight muscles and ripe ovaries to the total with ripe ovaries]; number of specimens with or without fl ight muscles within some species not comparable (indicated with an asterisk); data only used to test independence between the occurrence of reproduction and functional flight muscul ature; species ordered in three groups as in results (spring breeders, autumn b reeders and dimorphic/polymorphic species) and in alphabetic order within each group.

Species n 0+ 0- R+ R- G-test significance

Acupalpus f lavicollis 33 19 6 5 3 0.53 n.s. Agonum dOJ··sale * 11 8 6 16 II 85 3.1 7 n.s. Agonum. muelleri * 78 19 12 9 38 14.55 p<0.001 Amara aenea 324 24 59 27 214 13.2 1 p<0.001 Amara familiaris 48 7 7 18 16 0.03 n. s. Amara tibialis - Anisodactylus binotatus 14 9 3 2 4.20 p<0.05 Asaphidion curtum 11 7 8 18 30 61 0.04 n.s. Harpalus ajji.nis 33 9 5 8 11 1.60 n.s. Pterostichus versicolor 92 26 28 9 29 5.84 p<0.05

Amara biji-ons 26 3 1 22 17.23 p<0.001 Harpalus attenuatus 96 19 I I 66 51. 13 p

Bembidion properans * 219 28 II 44 136 30.77 p

of their reproductive peri od) (Fig. 3, C-D). Later on, thi s tibialis), poor grasslands (Pterostichus versicolor) to pas­ percentage decreases. These carabids also possess well­ tures and cultivated fields (Agonum dm·sale, Agonum developed hind wings (D ESENDER, 1989b). Both species muelleri, Amara aenea, Harpalus ajjinis). Some of these can b e fo und in moderately to very humid sites, wh ich species are known to hi bernate in edges of fields, hedges regularly inundate duri ng winter. The remaining species or woodland edges. The obtained patterns therefore can to some degree show two periods a year with a more be interpreted in terms of habitat change performed by elevated fraction of beetles with fu nctional fli ght muscles beetles, though not necessarily in many individuals by (Amara tibialis and Harpalus ~[finis show one such per­ means of fl ight dispersal. iod, but the number of individuals studied was low in Species studied from multiple popul ations show com­ these species). In nearl y a ll species the periods with a parable patterns. The different populations of Amara higher i ncidence of functional fl ight muscul ature are aenea, for example, show a clear alternation of the period clearl y complementary to the reproductive period of reproduction and p eri ods with a hi gher proportion of (Fig. 3, E-Q). Without excepti on, these carabids li ve in individuals with functional fl ight muscles. The maximal open habi tat types, ranging from dune grasslands (Amara proportion of beetles with functional flight muscles how- 20 Konjev DESENDER '' ever shows a large interpopulation variability. These there is a significant deviation from independence differences can nevertheless at least pa1tly be interpreted between r eproduction and the occurrence of functional when comparing the sampling s ite characteristics. High fli ght muscles, but the number of dissected b eetles is percentages of beetles wi th functional flight muscles are low. A s imilar result is obtained for Hmpalus rujipes observed on a recently created motorway verge (Fig. 3, J), (Fig. 4 E), but here the percentages of beetles with whereas very low values a re found in a population occur­ flight muscles are hi gher, while the G-test is not signifi­ ring in a relic site from an old, highly decalcified, dune cant (but low sample s ize). Hcupalus rujipes, known as a grassland area (Fig. 3, I). These results indicate that a weed seed predator, prefers fi elds w ith a more or less complex of factors plays a role in the expression offlight developed h erb layer (its larvae mainly feeding on small muscle development (not only habitat characteristics, but pl ant seeds), thus possibl y requiring regular (re)coloni sa­ also age of the population si nce initial colonisation, mean tion. individual body size influenced by environmental condi­ The remaining Hmpalus species (Fig. 4, 8-1) show a tions during ontogeny, etc.; cf. D ESEN DER , 1989a). hi gher proportion of beetl es with functional flight mus­ In Agonum dorsale and Agonum muelleri (two species cles at the beginning of their phenology curve (coinciding available in sufficiently large numbers), the number of w ith the appearance of the new generation), followed by a individuals with fli ght muscles is reduced to zero during sharp decline. Dissection of females reveals that those winter. This s uggests that a number of specimens must w ith ripe eggs in the ovaries never possess functional develop and subsequently autolyse or resorb their fli ght fli ght muscles, demonstrating a complete (highly signifi­ muscles at least t wo or three times during their lifetime. cant) oogenesis-flight syndrome ( Table I). All these They first develop their flight muscles after emergence Harpalus species a re more or less bound to sandy soil , during a utumn in order to fly in search of a convenient usually w ith poor grassy vegetation. Di fferences between hibernation quarter. Then, they autolyse their muscles, maximal proportions of beetles with flight muscles are followed by regeneration during early spring in order to relati vely large between species, but somewhat less reli­ search for a suitable reproductive site and, once again, able due to relatively low sample s izes. H01palus tare/us resorption at the onset of reproduction. (Fig. 4, G-l), with low fractions of beetles w ith functi onal Compari son of the observed maximal proportion of fli ght muscles, shows some variability between popula­ beetl es with flight muscles and the species-specific hind tions. The differences between the two populations of wing development (DESENDER, 1989b), shows that those Harpalus attenuatus can be in part the consequence of species with small est relative wing size (Agonum d01 ·sale, a possible detection problem: proportions with functional Amara tibialis) show very low percentages of beetles fli ght muscles in th e dune grassland (Fig. 4, 8-C) could with fli ght muscles. have been bi ased to lower valu es because of the very hi gh proportion of newly emerged tenerals in this sample. In such beetles flight muscle development is sometimes 2. Seasonal variation ojjlight muscle jill1ctionality difficult to describe. If such indi viduals develop post­ in summer-autumn-reproducing ground beetles teneral functional flight muscles, flight muscle develop­ (larval or larval/adult hibernators) ment would need at least some days to some weeks after th e beetl es have emerged and already could have been Data on I 0 species are s hown in Fig. 4 (A-P). For Nebria active on th e so il surface (cf. SMITH , 1964). brevicollis, Hwpalus tare/us and Harpalus attenuatus, Leistusfulvibarbis, Leistus rufomarginatus and Nebria data are given on respectively 4, 3 and 2 year cycle series. brevicollis (Fig. 4, J-0), three species with a high number These species can be grouped into different categories: of dissected specimens, show s trong simil arities in their (I) Amara bifi-·ons, a s pecies with summer r eproduction pattern s of seasonal flight muscle development. Again , (short adult period), (2) species belonging to the genus beetles with functional fli ght muscles are largely re­ Harpalus, probably with summer reproduction, some stri cted to the non-reproductive peri od, i.e. th e emergence species in an annual (larval hibernation) to bi enni al cycle period during late spring, before the summer aesti vation. (larval and adult hibernati on), (3) Leistus jidvibarbis, These species suggest the occurrence of a distin ct (s ig­ Leistu .s· rufomarginatus and Nebria brevicollis, all known nificant) oogenesis-flight syndrome (Table I). Differ­ to emerge in spring, followed by an adult aestivation ences in th e fractions of beetl es with fli ght muscles can dormancy and r eproduction during a utumn, ( 4) Trechus be interpreted in terms of the habitats preferred by th e quadrislriatus, with its new generation appearing in sum­ different species. Both Leistus species occur in wood­ mer, reproducing in autumn till early next spring. The lands, especiall y in moderately humid to wet sites. Nebria number of beetles with fu nctional flight muscles in non­ brevicollis, on the other hand, is a very e Luytopic species reproductive as compared to reproductive fe males is from forests, parkland and grasslands. g iven in Tabl e I, along with the results of G-tests of Trechus quadristriatus (Fig. 4 P), a common species on independence, as we ll as sample s izes. certain types of cultivated fields, is a n example of an Amara biji-ons prefers dry sandy habitats with poor autumn breeder emerging during summer (Jul y-August) vegetati on. Our resul ts (Fig. 4 A) show no c lear differ­ and reproducing from autumn till earl y next spring. Once ence between the seasonal activity peak and the monthl y again , immature beetl es from the new generation show a proporti ons of beetl es with flight muscles. Nevertheless, much m ore e levated proportion (nearly I 00%) of indiv i- I I Flight muscle development in carabid beetles 21

811 Amara bifrons .---- 38 Harpalus attenuatus 68 poor grassland A 8 "u "u poor grassland B" 48 B E "E 28 R 38 R c 28 c uA A 18 ; u 18 6 H H T T 188 188 Q8 88 88 118 p 18 p 711 E 1111 E 1111 R sa R c c 6fil E 48 E 48 311 T 311 " 28 T A "A 28 ; 18 ? ?? ? ? ? ? ...,.,..,.? ?? ; Ill ? E E ? ? ? ? ? 2 4 7 8 II 18 II s s e 12 2 s 4 s II 7 II g 111 11 12

S8 188 ,-- H. aftenuatus Q8 H. rubripes H H t-- u dune grassland c u 88 poor grassland D B 28 B 78 E" "E 118 R R Sfil c c ... A 18 A u u 38 G 6 211 H H 111 T T _r= b 188 1filfil Q8 811 89 88 p 78 p 711 E 89 E 611 R R c S8 c 6fil E 48 E ... 38 S8 T T 211 "A 28 "A G 18 ? ? 6 18 E ?? ? ? ? E ? ?? ? ? ?

3 4 s 9 18 11 12 2 s s II 8 II 18 11 12

12 H. rufipes E u 28 u" 18 poor grassland " B 8 "E 8 E" R R 6 c c 18 A A u 4 u G 2 8 H H T T 188 188 Q8 811 88 118 p 78 p 711 E 88 E 611 R R c sa c 6fil E 48 E 48 38 311 T T "A 28 "A 28 G 18 ? ? 6 18 ? ? ? ? ? E ? ? ? ? ? ? E ? ? ?

2 3 4 s e 7 8 9 18 11 12 II 2 s 4 5 II 8 g 18 II 12

Fig. 4 (A- P) - Phenology of the li fe cycle in summer-autumn breeding ground beetles (upper fi gure; bl ack colu mns: t eneral beetles), compared to the seasonal occurrence of monthly proportions of beetles with functio nal flight muscles (lower figure; with 95% c.i. fo r large amples,?: months without data, arrow · indicate zero-values ba eel on large samp le sizes). A: Amara biji·ons (poor grassland on sanel y soil), 8-C: Hmpalus a/tenua/1/s (8: poor grassland on motorway verge on sandy so il ; C: dune .grassland), D: !-lcnpalus rubripes (poor grassland), E: Harpalus rt(/ ipes (poor grassland), F: Hmpalus ru{tpalpis (poor grassland on motorway verge on sanely soil ). I' 22 Konjev DESENDER

88 r-- N 78 Leistus fulvibarbis N G u 58 u 89 humid valley forest r-- Harpalus tardus B 48 - "B 58 "E dry grassland E R R 49 311 ..__ c .--- c !!9 A 28 A u u 2lil 6 18 19 H ~ T T r- n 1118 1118 • •B8 89 p 78 p 79 E B8 E 89 R c 58 ~ 59 E 48 E 49 N 58 N !!9 T A 28 ! 29 6 18 6 18 E ? ?? ? ? ? E

2 ' 4 s 8 8 g 19 II 12 3 4 S 8 7 8 9 18 II 12 12 19 L. rufomarginatus N H. tardus 18 u H woodland K 8 dune grassland 8 8 N 8 E" u R 6 4 instar I "B 2 c 4 E A R 8 u ; 2 0 s H F 4 T 1118 L 3 II • A 2 89 ~ I p 79 : 14 E 89 R 12 c 59 C II E 49 N 39 ~ : Ill T 8 A 29 H 4 6 19 T 2 E ?? ? ? ? ?

2 3 4 5 7 e 18 II 12 18 • N 16 H. tardus N 88 u 14 poor grassland ~ 78 B 12 B B8 "E E 59 R 19 R 48 c 8 c A 6 A 38 u 4 u 28 & 6 H 2 H 18 T T 189 188• •89 89 p 78 p 78 E 89 E 88 R c 59 ~ 58 E -48 E 48 N 58 N 39 T 28 A 29 ! 6 Ill 6 18 E ? ? ? ? ? ? ? E

2 3 4 s 8 8 9 Ill II 12 2 3 4 S 8 7 8 g Ill II 12

Fig. 4 (continued) - 0-K: 0 -I: f-!cupalus tare/us (0: dry g rassland , H: dune grassland , 1: poor grassland), J: Leistusjitlvibarbis (humid vall ey fo rest), K: Leis/us rufomcug inatus (woodland site; phenology of the three larval instars added). Flight muscle development in carabid beetles 23 ' '

ee Nebria brevicollis 89 L II 68 u 78 N. brevicollis N 48 instar I II 31 8 • grassland + scrub u E" 519 28 R B 48 "E 18 c 38 R A • u 28 0 S8 ; f' 48 H II 38 T L II 188 A 28 R 118 v II 88 A p 78 E 218 E R c 519• A 1lil c u E 48 111 Ill II 38 6 T H 58 A 28 T 6 18 E ? ? ? ? ? ?

It 2 3 4 5 e 7 8 II 18 11 12 u 168 11188 148 II II 138 N. brevicollis "E u 128 R 1111 pasture 0 B" ... c 6811 E .. A R 89 u 71 & c ee H A 58 T u 411 ... 8 38 118 H 28 T 18 81 1. p 71 118 E R • • c 68 p 78 E 411 E II 31 R • T c 58 A 28 E 411 6 18 II 38 ? T E ? ~ ~ ~ 28 A ; 11 ? 2 3 4 5 e 8 II 18 11 12 E ? ? ?

2 3 4 6 e 7 8 II 11 11 12

4.11 It N. brevicol/is Trechus quadristriatus p u • II 3.5 li8 deciduous forest u cultivated fields 8 3.8 " B E 411 " 2.5 R E 38 R 2.8 c A 28 c 1.5 u A 6 u t.ll 18 & H 8.5 T H 188 T 118 ••118 88 p 78 • E p 78 R E 88 •519 R c c 68 E 411 E II 38 ... T It 31 28 T A A 28 6 Ill ; E ? ? ? ? ? ? ? t ~ ~ .. E

2 3 4 5 II 7 8 9 18 It 12 2 3 4 5 II 7 II 9 .. II 12

Fig. 4 (continued) - L-P: L-0: Nebria brevicollis (L: beech forest; phenology of the three larval instars added; M: deciduous forest; N: grassland with scrub; 0 : pasture), P: Trechus quadristriatus (cultivated fi elds). I I 24 Konjev DESENDER duals with functional fli ght mu sc les as compared to beet­ of hi ghest number of ripe eggs carried by the females in les at a later stage of their life cycle. The flight period of their ovaries (Fig. 5 G). this species is indeed restricted to a short period between Ca lathus rotundicol/is, a species from li ght forest and July and September. Dispersal by flight is probably an woodland edges, again seems to show an oogenesis-flight adaptation to the temporary characteristics of the repro­ sy ndrome. Only at the onset of the seasonal activity cycle, duction habitat (e.g. due to crop rotation, ploughing a nd at most 20% of the beetles possessed functional fli ght other management practices). Although several fem ales mu sc les. have ripe ovaries and at the same time functional fli ght Finally, Bradycel/us hcnpalinus shows a pronounced muscles, a significant oogenesis-flight syndrome is ob­ oogenesis-flight syndrome: functional flight muscles and served (Table I). possible di spersa l by fli ght are limited to the post-teneral and pre-reproductive period. Dry grassland, heathland as well as wetland are the preferred habitats of thi s rather 3. Seasonal variation of wing morph .fi'equencies and eurytopic species. flight muscle jimctionality in wing dimorphic and polymorphic ground beetles Discussion Data on seven wing dimorphic or polymorphic carabid species a re summarised in Fig. 5 (A-H) and in Table I Because we did not study actual flight behaviour in (dissected females). Three populations were studied of ground beetles, our results g ive indirect ev idence only the wing dimorphic Bembidion properans, two popula­ fo r seasonal variation in the occurrence of dispersal by tions of the wing dimorphic Bradycellus hwpalinus, one flight. Studies based on flight observations only, on the population of the wing dimorphic Clivina fossor and other hand , suffer from other shortcomings, because such Calathus rotundicollis, and one population of the wing observations are highly influenced by meteorological polymorphic Pogonus chalceus, Pterostichus minor and va ri ation. In addition, fli ght observations, especially if Pterostichus vernalis. Most of these species are reprodu­ gathered by light trapping, are much more difficult to cing during sp ring (adult hibernators), with th e exception quantify in terms of local population sizes. Data on fli ght of Calathus rotundicollis (summer reproduction; larva l musc le development (as in our study) give an idea of the hibernator) and Bradycel/us hcupa/inus (reproduction maximal proportion of individuals in a population that during autumn till next spring). might be able to fl y. Bembidion properans (Fi g. 5 A-C) does not show Proximate factors influencing the onset and duration of much diffe rences between observed wing morph frequen­ flight in carabid beetl es predominantly act as strong in­ cies per month (two populations) as opposed to the clear hibitors for flight. VAN HuiZEN (1979) showed that most seasonal changes in functional fli ght muscle frequencies ground beetles show no flight acti vity below l7°C (see in all populations studied. The results suggest the occur­ also 1-I ONEK & PULPAN, 1983), during precipitation and at rence of an oogenesis-flight syndrome during spring (see wind speed exceeding about 5 m/s. Day-acti ve spec ies also Table I), while only very few beetles with functional appea r to require strong insolation prior to flight beha­ fli ght musc les are observed during autumn. viour. For many carabids from our regions, these imposed A hi gher incidence of beetl es with fli ght muscles in physiological constraints reduce the number of days per spring and in autumn is observed in Pterostichus vernalis year, suitabl e for flight, to a few days to some weeks, with and Pterostichus minor, again suggesting the occurrence extreme va ri ation between years and sites. This has been of a seasonal shift in habitat (mi gration towards or away confirmed by long term series of fli ght observations in from an overwintering habitat, cf. DESEN DER et al. , 198 1). ground beetl es (e.g.: li ght traps: 1-IONEK & PuLPAN, 1983 ; In the case of Pterostichus vernalis, however, maximal window traps: VAN 1-I UIZEN, 1979; DEN BoER et al. , proportions of beetles with functional fli ght muscles are 1980) .d Published ata on the fli ght periods in the speci es quite low. Obviously these numbers a re f urther reduced from our study (compiled in Table 2) largely correspond not only during reproduction but also during hibernation to our observed seasonal pattern of flight muscle func­ diapause. tionality. Species which are known occasionall y to per­ Clivinajossor (only limited data and no mac ropterous form mass fli ghts a lso show nearly complete fli ght mus­ females ava il abl e) shows a low maximal proportion of cle functionality during s uch periods (e.g. Bradycellus beetles with fli ght mu sc les, which seems to coincide with hwpalinus, cf. KERSTENS, 196 1; Trechus quadristriatus, the reproducti ve peri od. Interestingly, thi s spec ies has a cf. LACMAN , in press). largely s ubterranea n way of life and d oes not seem to If the species that we have investigated are grouped perform habitat change in the pasture studied (cf. DESEN ­ into spring breeders (short larval cycle in summer, adult DER, 1983 ; DESEN DER & POLLET, 1985). hibern ation) and summer-autumn breeders (long larva l The sa ltmarsh inhabiting Pogonus chalceus does not cycle till nex t spring or summer) , the latter contain more show an oogenesis-fli ght syndrome, alth ough high num­ species with fun cti onal flight musc les in the post-teneral bers of beetles were investigated (Tabl e I). Despite being and pre-reproductive period than the fo rmer. Such species low, th e yea rl y peak of beetles with functional flight fl y during a short time span of the year (sometimes a few mu scles (DESEN DER, 1985), co incides with the moment days only w ith mass flight, cf. 1-IONEK & PUL PAN, 1983). II Flight muscle development in carabid beetles 25

188 Bembidion properans B. properans A Hsae c 88 city lawn u pasture p 78 E ea "­B ~ sa E 48 ~- c2llll A l : u 6 18 s 188 E ? ? ? H T 188 811 N u 21111 811 p 7B B E BB "E 168 R ~ sa E 48 c 188 A ~ : A ~ sa s 18 ? H E ?? T 188 811 2 3 4 S 8 7 8 II II II 12 88 minor p 78 N 38 Pt. E ee u marshland forest ~ sa B E 48 E" 2ll ~ R A : c s 18 ?? A IB E ? ? u 6 IB II 12 H 2 s 4 5 8 T 188 liB BB 188 p 78 B E ee 88 p 78 ~ sa E 40 E ee ~ sa ~ : E 48 A 6 Ill ? ? N 38 E . . ! 211 6 18 2 s 4 5 e 1 8 9 18 11 12 E ? ?

68 Pt. vernalis H 3611 u - "u 40 pasture E " 3118 8 ~2511 E" 38 R2llll R c rsa A u 188 6 H sa T 188 811 88 811 p 78 p 7B E 88 E BB ~ 58 ~ 58 E 48 E 411 " 38 A~ : ! 2ll 6 18 6 IB E ? ? ? E

2 s 4 s e 8 II 18 II 12 2 S 4 5 8 7 8 II II II 12

Fig. 5 (A-H) - Wing dim orphi c and polymorphic carabid spec ies: ca lculated percentages with functiona l fli ght mu sc les based on the number of mac ropterous indi viduals onl y, except for D, E and G; see lege nd Fi g. 3 fo r further ex pl anation. A-C: Bembidion properans (A-8: two recent law ns in the city of Ghent; uppermost fi gure represents the seasonal va ri ation in the percentage (with 95% c. i.) of winged bee tl es for thi s wing dim orphi c spec ies, middle figure shows phenology of th e li fe cycle (bl ack co lumns: te neral beetles), compared to the seasonal occurrence of monthl y propo rti ons of beetles with functional fli ght mu sc les (lower fi gure; with 95% c. i. fo r large sa mpl es,?: months without data, arrows indi cate ze ro-va lu ~s based on large sample sizes).; C: pasture), D: Pterostichus minor (marshl and forest) , E: Pterostichus vernalis (pasture). 26 Konjev DESENDER

:2lil Clivina fossor 14 Bradycel/us harpalinus H F H u grassland u 12 dry hayfield 8 8 HI E" "E R R II 18 c c II A A u .. u ; .. 2 H H T T IBB IBB liB liB 8B 8B p 7B p 7B E 8B E 8B R R c SB c SB E 48 E 49 H sa N sa T T A :2lil A :2lil ; 18 ? ? ? 6 Ill E ? ? ? E ? ?

2 3 .. s e 7 18 II 12 2 3 .. s II 7 8 9 18 II IZ e ee N Pogonus cha/ceus N u 5 u sattmarsh G se 8 4 B "E "E 48 R 3 R 3B 0 c F 2 A :2lil u E 8 Ill 6 H 6 T s 1118 1119 ge ge ae se p 7B p 78 E 68 E 118 R R c se c se E ..., E 48 H 3B N sa T T A :2lil A :2lil 6 Ill 6 18 E t E

2 3 5 8 7 g 18 II 12 z 3 s II 7 8 g 18 II IZ 88 N u 78 Calathus rotundicol/is 88 woodland H 8 E" 58 R 48 c sa A u :2lil 6 H 18 T 118 118 se p 78 E liB R c se E <48 N 3B T A :2lil 6 18 E ? ?? t ?

2 3 s II 8 g 18 It 12

Fig. 5 (continued) - F: Clivinafossor (grass land), G: Pogonus chalceus (sa ltm arsh; num ber of ripe eggs per fe male pl otted as ind ica tion of reproductive peri od), 1-1 : Ca lathus rolundico!lis (woodland site) , 1-J : Bradyce!lus harpa!in us (I: dry hayfield , J: dry poor grassland). Flight muscle development in carabid beetles ' ' 27

Table 2. - Compilation of literature observations of flight and main annual flight period in the ground beetle species, studied in thi s paper; species ordered in three groups as in re sults (spring breeders, autumn breeders and dimorphic/polymorphic species) and in alphabetic order within each group: number of records: * = low numbers, X= hi gh numbers; number of symbols refers to the number of references with the species; data from: B ASEDOW & DICKLE R, 198 1; BRIEL, 1962; BRIGGS, 1965; D EN BOER, 1971 ; D ES ENDER, l986b and unpublished data; GREENSLADE & SOUT HWOOD , 1962; H AECK, 1971 ; H ONEK & P UL PAN , 1983 ; KADAR & S ZENTKI RALY I, 1983; K ERSTENS, 1961 ; L ACMAN, in press; Li 1DROTH , 1945; M ATALI N, 1994, 1997; M EIJ ER, 1974; Y ANHERCKE et al., 1980; V AN H UIZEN , 1979, 1980; ZHANG et al., 1997; Z ULKA , 1994.

Species number of records flight period(s)

A cupalpus jlavicollis X* ? Agonum dOl-sale *** Apri l-May, August Agonum muelleri *** April, September Amara aenea XXXX** April-May, (September) Amara familiar is XXX*** April-July Amara tibialis * April-May Anisodactylus binotatus XXX* April-May, (September) Asaphidion curtum **** April-June Harpa!us afjinis X* May-June Pterostichus versicolor ** May, end July

Amara bifi·ons XX***** July-August Harpalus attenuatus · ? ? Harpalus rubripes * ? Harpalus rufipes XXXXXX*** July Harpalus rufipalpis X ? Hmpalus tardus ? ? Leistus jitlvibarbis ? ? Leistus nifomarginatus ? ? Nebria brevicollis * May Trechus quadristriatus XXXXXX** * July-September

Bembidion properctns ** April-May Bradycellus harpalinus XXXXXX* * * * ':' July-August Calathus rotundico/lis * early August Clivina jossor *** May-June Pogonus chalceus ? ? Pterostichus minor X* ? Pterostichus vernalis **** May, September

Many spring breeders, on the other hand, with a winter­ THI ELE, 1977), as opposed to a lot of day-active spring diapause between emergence and reproduction, show a species. Summer-autumn breeders like Bradycellus har­ higher incidence of functional flight muscles before as palinus, Harpalus rl!fipes and Trechus quadristriatus are well as after reproduction (in that case mainiy post-ten­ caught especially flying at night (cf. HoNEK & PULPAN, era!). Seasonal variability in fli ght muscle development 1983 ; MATA LI N, 1997; ZH ANG et a!. , 1997). The high in these species co incides with two known fli ght periods temperatures needed for fli ght further reduce the number per year (references, cf. Table 2) related to habitat change of days with suitable conditions for fli ght in night-active to and from overwintering sites and sites for reproduction. species. Spring carabid beetles, on th e contrary, are prob­ After emergence and fli ght activity, adults of many sum­ ably less constrained for flight because of their daytime mer-autumn breeders directl y start reproduction. From activity and indeed regul arl y show two flight periods a th en onwards the reproducing generation graduall y dies year, related to habitat change. In some cases (e.g. Bem­ off. The very short flight periods in th ese spec ies can also bidion properans) the spring pea k appears to be more be a partl y consequ ence of other life hi story as pects. important than th e autumn peak (following from data on Many of these ground beetles indeed are known to be fli ght muscle functionali ty as well as from literature data exc lu sivel y active during the ni ght (DES ENDER et al. 1984; on flight periods). This cou ld refl ect a synchronous de- 28 Konjev DESENDER

velopment of flight muscles after winter diapause in order therefore concluded that the "oogenesis-flight syn­ to be able to recolonise the reproduction h abitat as drome" in many cases is labile and controlled mainly quickly as possible. During a utumn the appearance of by the environment. the new generation i s much m ore spread, while short­ If di spersal (migration) contributes to individual fitness ening of day length inhibits vitellogenesis and thus repro­ (survival and direct or indirect reproductive output), we duction (TH IELE, 1977). It is therefore possible that sui­ expect that species, emerging during summer, will show table hibernation quarters are then more easily reached by flight activity immediately before their reproduction. At less energy-demanding activities such as walking instead that moment, climatological conditions are most suitable of flying (more " time" available?). and females are not yet carrying the extra weight of eggs A majori ty of the carabid beetles investigated seems to (males not yet the extra weight of very much enlarged exhibit a significant "oogenesis-flight" syndrome: repro­ gen ital accessory glands). In spring-breeding species this duction and reduced fli ght musculature are thus not in­ is not the case. Climatological conditions during early dependent events in their life cycle, i.o.w. there is a trade­ spring and autumn are much more regularly below thres­ off between reproduction and functional fli ght muscula­ hold values fo r fl ight. This might give an explanation fo r ture. Reproducti on as we ll as flight being both very the higher incidence in such species of females with ripe energy-consuming activities, it seems plausible that re­ eggs simultaneously with functional flight muscles (i.e. a sorption of flight muscles liberates energy (cf. ZERA & less deterministic oogenesis-flight syndrome). DEN o, 1997) and/or creates space to accommodate en­ In some of the species investigated, flight muscles are larging reproductive organs (cf. SOLBRECK, 1986; T ADA et resorbed after suitable hibernation quarters have been al., 1991 ). Flight activity during reproduction probably is reached, and are regenerated next earl y s pring. This in­ hampered or even made impossible in some species dicates that the mere maintenance of a functional flight because of an increased individual body we ight (cf. MA­ apparatus (even without flight activity) involves high TA LI N, 1997) and consequently higher wing loading. energy investments. Consequently, migration and dia­ These hypotheses should be tested experimentally. Spe­ pause do not run strictly in parallel in such cases. cies where some individuals will and others will never be Functional fl ight muscles are expected to develop only able to develop fli ght muscles (genetic flight muscle when there is some selective advantage involved. This dimorphism) present interesting cases for such experi­ enables for example a rapid exchange between habitats ments. for spring species while yielding much higher chances of In many ground beetle species, especiall y spring bree­ survival in protected winter quarters. Extreme examples ders, we observe a proportion of mature females with are species from habitats inundated during winter. An functional flight muscles, indicating a less deterministic even stronger selection for the continuous presence of version of the oogenesis-flight syndrome. This could functional flight muscles is expected to occur in species have important side effects for the colonisation chances from very unpredictable habitats (i. e. a habitat that cannot of such species. When migration i s re lated mainly to continuously be inhabited during the reproductive peri­ seasonal changes between habitats for reproduction and od). Such species need to be able to escape by flight at for hibernation, straggling individuals leaving hi berna­ any time, for exampl e from water floods after heavy ra in, tion quarters will now and then end up in a s ite, never or from tides in saltmarshes. In such situations, a strong colonised before. Sporadic flight activity of ripe or at positive selection is expected for a permanent functional least inseminated females would strongly increase the flight apparatus. At the end of this line of reasoning are chances for founding new populations. VA N HUIZEN examples of species that need flight for their normal dai ly ( 1990) came to a sim ilar conclusion after dissecting 932 activities (e.g. prey caught during flight or escape fro m beetles belonging to 62 carabid species. Unfortunately, predators in Cicindela or tiger beetles, .. .). this author did not list the species or sampl e sizes, but Maximal proportions of beetles with functional flight gave pooled results, difficult to evaluate. muscles in many species only reach l ow values, which JoH NSON ( 1969) came to the conclusion that, in insects can strongly d iffer between populations. This clearly in general, processes of gonad maturation and degenera­ indicates the complex regulation of flight muscle ex­ tion/regeneration of flight muscles are coupled during pression. Underlying mechanisms can be both autolysis/ morphogenesis. These processes would have been adap­ rebuilding processes and their regul ation but also pure tively brought into phase in di fferent ways according to genetic flight muscle dimorphism. Experimental research the species. De- and regeneration of fl ight muscles are is urgently needed for a better understanding of these part of a labil e system of morphogenes is, but are still processes. Regulation of autolysis and regeneration of insufficiently understood (RANKI 1 et al. , 1986). Flight flight muscles in the li fetime of individuals has hardly muscle development, because of its hidden aspect, has been studied in ground beetles. An important problem been studied insufficiently and the difference betvveen in this respect is that the observed phenotypic variation ontogenetic (direct) and phylogenetic (evolved) fli ght not necessarily reflects the genetic variabi li ty present in muscle dimorphi ·m is vague (JOHNSON, 1969). Variabi­ a population. Ex perimenta l studi es on the regul ation of lity in the synchronisation between egg production and a fli ght muscle fu nctionality of carabids were performed functional flight apparatus is more pronounced when both in Nebria brevicollis (N ELEMANS, 1983, 1987) and in developmental pathways are reversi ble. JOHNSON ( 1969) Pterostichus ob/ongopunclatus (VAN SCHAICK ZI LLESEN 'I Flight muscle development in carabid beetles 29

& BR UNSTING, 1984). Unfortunately, these two species value of ground beetles as model organisms in studies rarely fly and therefore low fractions of individuals with on dispersal and gene flow, population genetics, ecology flight muscles are observed in field situations. Favourable and evolution. conditions during the larval ontogeny of Nebria brevicol­ lis (such as sufficient food supply, short day-length) appear to enhance the expression of functional flight Acknowledgements muscles in the resulting young adults. R esults for Pterostichus oblongopunctatus lead more or less to an This paper would not have been possible without the continuous help, opposite conclusion, but the data seem less convincing. during many years, from many students and colleagues: in particul ar Mark Alderweireldt, Leon Baert, Jean-Pierre Mae!f ait, Marc Pollet and Some small experiments of BOMMARCO ( 1998) suggest Marc Van Kerckvoorde aided in the coll ection of large series of year that wing muscles of Pterostichus cupreus increase in cycle pitfall trap samples in different habitats of our country. Marc Van size with increasing food availability. In a large scale Kerckvoorde in add ition helped by di ssecting a number of series of ground beetles for their flight muscle development. P. Grootaert biometric study on numerous ground b eetle species (RB INSc) and A. Huysseune (RUG) kindly revised previous drafts (DESENDER, I 989b), we have regularly noted that in­ of thi s paper and H. Turin added useful review comments. dividuals possessing fli ght muscles have a larger mean body size as compared to beetles with degenerated flight muscles. This seems to be in agreement with the results References of NELEMANS (1983, 1987) on Nebria brevicollis. In her experiments, beetles raised under more favourable AUKEMA, B. , 1986. Winglength Determination in relation to conditions (higher food supply, sh01t day-length) not Dispersal by Flight in two Wing Dimorphic Species ofCala1hus only developed flight muscles, but were also larger in Bo ELLI (Coleoptera, Carabidae). In: DEN BOE R, P. J., LUFF, size. The somewhat unexpected conclusion drawn from M.L., MOSSAKOWSKI, D., & WEBER, F. (eds.): Carabid Beetles. this observation is that under more favourable environ­ Their adaptations and dynam ics. Stuttgart: Gustav Fischer, mental conditions more beetles would b e capable for pp. 9 1-100. flight. It contradicts the general idea that developing BAGU ETTE, M., 1993. Habitat s election of carabid beetle in flight muscles and performing flight activity would lar­ deciduous woodlands of southern Belgium. Pedobiologia, gely be an adaptive escape reaction in a deteriorating 37:365-378. habitat ( cf. JOHNSON, 1969 for insects in general; VAN BAGUETTE, M., 1987. Spring distri bution of carabid beetles in ZCHAI CK ZILLESEN & BRUNSTING, 1984). Body size of different plant communities of Belgian forests. Acta Phyto­ beetles, on the other hand, is, at least partly, influenced pathologica et Entomologica Hungarica, 22:57-69. by environmental conditions during ontogeny (cf. DESEN­ BASEDOW, Th. & DICK LER, E., 198 1. Untersuchungen tiber die DER, 1989a), which questions to some degree the term Laufkafer in einer Obstanlage anhand von Boden- und Licht­ "adaptive". fa llenfringern (Co l. , Carabidae). Milteilungen der deutschen Clearly, this study area offers a lot of new and original Gesellschajf fiir allgemeine und angewandte Entomologie. subj ects for investigation. First, however, more empirical 3:36-39. data are required, especiall y on other species and in other BOMMA RCO, R., 1998. Stage sensi ti vity to food li mitation for a regions. Such field-collected data, along with those pre­ generalist arthropod predator, Pteroslichus cup reus (Coleop­ tera: Carabidae). Environmental Entomol 27:86 sented in this paper, are a first step towards straightfor­ ogy, 3-869. ward experimental and physiological studies on flight BRIEL, J. , 1962. Quelques observati ons sur l'ecologie des Oplr o­ muscle fu nctionality and its underlying processes. Over­ nus. Entomologiste, 18:7- 12. al l, our results suggest that flight muscle development, BRIGGS, J. B., 1965. Biology of some ground beetles (Col., like other life history traits of ground beetles (such as Carabidae) injurious to strawberries. Bulletin ofentomological habitat preference, life cycle timing, body size, abun­ Research, 56: 79-93. dance, hind wing development, ... ), is part of a s uite of DEN BOE R, P.J. , 1971. On the dispersal power of carabid beetl es coadapted traits (DESENDER, 1986a). and its possible significance. Dispersal and dispersal power of carabid beetles. Miscellaneous Papers In conclusion, a majority of the ground beetle species, Landbouw-Hogeschool Wageningen. 8: I 19-13 7. investigated in this study, shows seasonal variation in the number of specimens w ith functional flight musculature. DEN BOER, P. J. , VAN HUIZEN, T. H. P., DEN BOER-DAANJE, W., AUK EMA, B. & DEN BI EMAN, F. M, 1980. Wing Polymorp In many cases, this variabili ty can be related to other life hism and Dimorphism in Ground Beetles as Stages in an Evolu tion­ history traits. The complexity of the phenomenon is ary Process (Coleoptera: Carabidae). Entomologia Generalis, evident, and causes and effects in many cases are hard 6: 107-1 34. to distinguish, suggesting that flight muscle development DESEN DER, K., 1983. Ecologica l data on C/ivina fossor (Co­ is part of a suite of coadapted traits. The large interspe­ leoptera, Carabidae) from a p asture ecosystem. I. Adult and cific variability in flight muscle f unctionality a nd the larval abundance, seasonal and diu rna l activity. Pedobiologia, more or less pronounced occurrence of an oogenesis­ 25: 157-167. fli ght syndrome is striking. This adds a further dimension DESENDER, K., 1985. Wing polymorphism and reproducti ve to the already very large interspecific diversity in hind bi ology in the halobiont carabid beetle Pogonus chalceus (Mar­ wing development of ground beetles (cf DEN BOER et al., sham) (Coleoptera, Carabidae). Biologisch J aarboek Dodo­ 1980; DESEN DER, 1989b). Eventually, it increases the naea. 53:89-1 00. 30 Konjev DESENDER

DESENDER, K., 1986a. Ecology, di stribution and di spersal !anl agen und Maisbestanden. Verhandlungen SIEEC. I 0:150- power of endangered Carabid beetl es in Belgium. Proceedings 154. of th e Jrd European Congress of Entomology, 3:429-432. KERSTENS, G., 1961. Co leopterologisches vom Lichtfang. En­ DESEN DER, K., 1986b. On the relation between abundance and tomologische Bliiller, 57: 11 9-13 8. flight activity in Carabid beetles from a heavily grazed pasture. LACMA ' E., in press. Carabidae recoltes par un piege asucc ion Journal of Applied Entomology, I 02 :225-231 . a 12 m au-dessus du niveau du sol. Bulletin des Recherches DESE DER , K., 1989a. Heritabi lity of wing development and agronomiques de Cembloux. body size in a Carabid beetle, Pogonus chalceus Marsham, and Li NDERS, E. G.A., TURI N, H. & BRUGGEMA , C.G., 1995. Biol­ its evoluti onary significance. Oecologia (Berlin), 78: 513-520. ogy of th e weevil Trichosirocalus troglodytes and impact on its DESEN DER, K., 1989b. Di spersievermogen en Ecologie van host Plantago lan ceolata . Acta Oecologica, 16:703-7 18. Loopkevers (Coleoptera, Carabidae) in Belgie: een evolutio­ LI NDROT H, C. H. , 1945. Die Fennoskandischen Carabidae: Eine naire benadering. Studiedocumenten van het Koninklijk Bel­ Ti ergeografische Studie. 1. Spezieller Teil. Goteborg: Elanders gisch lnstituut voor Natuul"\ve tenschappen, 135 pp. Boktryckeri Aktiebolag, 630 pp. DESENDER, K., MAE LFAIT , J.-P., D'HULSTE R, M. & VAN HERCKE, Li NDROT H, C. H. , 1946. Inheritance of wi ng dimorphism in L. , 1981. Eco logical and fauna l studies on Co leoptera in agri­ Pterostichus anthracinus Ill. Hereditas, 32:37-40. cultural land. I. Seasonal occurrence of Carabidae in th e grassy MATALIN , A. V. 1994. The strategy of dispersal behaviour in edge of a pasture. Pedobiologia, 22:379-3 84. some Carabidae species of Southeastern Europe. In : DES ENDER, DESEN DER, K. , MAE LFA IT, J.-P. & VANEEC HOUTIE , M, 1986. K., DUFRENE, M. , LOR EAU, M., LUFF, M. L. , & MAELFAIT, J.-P. Allometry and Evo lution of Hind Wing Development in Macro­ (eds): Carabid Beetles: Ecology and Evoluti on. Dordrecht, pterous Carabid Beetles. In: DEN BoER, P. J. , LuFF, M., Mos­ Boston, London: Kluwer Academic Publishers, pp. 183-188. SA KOWSK I, D. , & WEBER, F. (eds.): Carabid Beetles. Their MATA LI N, A. V., 1997. Specific features of life cycle of Pseu­ adaptations and dynamics. Stuttgart: Gustav Fi sher, pp. I 01- doophonus (s. str.) rufipes Deg. (Coleoptera, Carabidae) in 112. Southwest Moldova. fzvestiya Akademii Na uk Seriya Biologi­ DESEN DER, K. , MA ES, D., MAE LFAIT, J.-P. & VAN KERCK­ cheskaya Biology Bulletin, 4:37 1-3 81 (translated from Ru s­ VOORDE, M., 1995. Een gedocumenteerde Rode lij st va n de sian). zandloopkevers en loopkevers van Vlaanderen. Mededelingen MEIJ ER, J. , 1974. A Comparative Study of the Immigration of van het lnstituut voor Natuurbehoud, I: 1-208. Carabids (Coleoptera, Carabidae) in to a New Polder. Oecologia ), 16:185-208. DESENDER, K. , MERTENS, J., D' HULSTER, M. & BERB IERS, Ph. , (Berlin 1984. Di e! activi ty patterns of Carabidae (Coleoptera), Staphy­ MUDA, A. R. B. , TUGWE LL, N. P. & HAIZLIP , M. B. , 198 1. linidae (Coleoptera) and Coll embola in a heavily grazed pas­ Seasonal history and indirect flight muscle degeneration and ture. Revue d '£cologie et de Biologie du Sol, 21:347 -361. regeneration in th e rice water weevi l. Environmental Entomol­ ogy, I 0:685 -690. DESENDE R, K. & POLLET, M. , 1985. Ecological data on C/i vina fossor (Coleoptera; Carabidae) from a pasture ecosystem. 11. NELEMANS, M. N. E. , 1983 . Flight muscle development in the Reproduction, biometry, biomass, wing polymorphism and carabid beetl e Nebria brevicollis. In : BRA DMAYR , P., DEN feeding ecology. Revue d'Ecologie et de Biologie du Sol, BOE R, P. J., & WEBE R, F. (eds): Eco logy of Carabids: the 22:233-246. synthesis of field study and laboratmy experiment. Wagenin­ ge n: Pudoc, pp . 45-52. FAIRBAIRN, D. J. & YADLOWSK I, D. E. , 1997. Coevolution of traits determining mi gratory tendency: correlated response of a NELEMANS, M. N. E., 1987. Poss ibilities for fli ght in the carabid critical enzyme, juvenile hormone esterase, to se lection on beetle Nebria brevicol/is (F.). The importance of food during wi ng morphology. Journal of Evo lutioncuy Biology, I 0:495- larval growth. Oecologia (Berlin). 72:502-509. 513 . RANKIN, M. A. & BURC HSTED, J. C. A. , 1992. The cost of migration in insects. Annual R eview of Entomology, 37:533- GATEHOUSE, A. G. & ZH ANG, X. -X., 1995 . Mi gratory potential 559. in insects: va riation in an uncertain environment. In : DRAKE, V. A. & GATE HOUSE , A. G. (eds.): Insect Mi gration . Cambridge: RANK IN, M. A., HAM PTON , E. N. & SUMMY, K. R. , 1994. University press, pp. 193-242. In vestigations of the oogenesis-flight syndrome in Anthonomus grandis (Coleoptera, Curculionidae) using tethered flight tests. GREENSLADE, P. J. M. & SOUTH WOOD, T. R. E. , 1962. The Journal of Insect Beha vior, 7:795-8 10. Relationship of Flight and Habitat in some Carabidae (Coleop­ tera). Entomologist, 95:86-88. RANK IN, M. A., MACANEL LY, M. L. & BODENHAMEr, J. E. , 1986. The Oogenesis-flight syndrome revi sited. In : DANTH A­ ., The immigration and settl ement of carabids in HAECK, J 19 71. NARA YANA, W. (eds): Insect flight. Di spersal and Migration. the new ljsselmeer-polders. Miscellaneous Papers Lan dbou w­ Berlin: Springer Verl ag, pp. 27-48. Hogeschool Wageningen. 8:33-52. RoFF, D. A., 1986. The evoluti on of wing dimorphism in HARR ISON, R. G. , 1980. Di spersal polymorphi sms in insects. insects. Evolution, 40: I 009-1020. Annual R eview of Ecology and Systematics, II :95- 11 8. SMITH, D. S., 1964. The structure and development of fli ghtless HONEK, A. & PULPAN, 1. , 1983. The fli ght of Carabidae (Co­ Co leoptera: A li ght and electron microscopi c s tudy of th e leoptera) to li ght trap. Vestnick ceskoslovenske S polecnosti wings, thorac ic exoskeleton and rudimentary fli ght muscula­ zoologicke, 4 7:13-26. ture. Joum al of Mo 1phology, 114: I 07-184. JOHNSON, C. G. , 1969. Migration and di spersa l of insects by SOKAL, R. R. & ROHLF, F. J. , 198 1. Bi omet1y . The principl es fli ght. London: Methu en & Co, 763 pp. and practice of stati stics in biological research. San Francisco: KADA R, F. & SZENTK IRALYI , F., 1983. Ana lyse der Lichtfallen­ 2nd ed . W.H. Freeman and Co mpany, 859 pp. fa nge der Laufkafer (Co l. , Carabid ae) in ve rsc hi edenen Apfe- SOLBRECK. C., 1986. Wing and fli ght musc le polymorphism in a Flight muscle development in carabid beetles 'I 3 1 lygaeid bufg, Horvathiolus gibbicollis: detenninants and life in which the occutTence of dispersal by flight of individuals has history consequences. Ecological Entomology, II :435-444. been shown. Entomologische Berichten, Amsterdam, 40:1 66- TADA, S., HONMA, K., KAKIZAKI, M., FUJISAKI, K. & NAKASUJI , 168.

F., 1994. Genetic mode of fli ght muscle dimorphism in a VA I SCHAICK ZI LLESEN, P. G. & BRUNSTING, A. M. H., 1984. Scarabaeid, Heptophylla picea Motschulsky. Applied Entomol­ The Influence of Food Quantity and Photoperiod During the ogy and Zooloy, 29:303-305. Pre-adult Stages on Flight Muscle Development in Adult Phi­ T ADA, S., HONMA, K., KAKIZAKI , M., FUJISAK I, K. & NAKASUJ I, lonthus decorus (Coleoptera: Staphyl inidae) and Pterostichus F. , 1995. Geographic variation of fli ght muscle dimorphism in oblongopunctatus (Coleoptera: Carabidae). Entomologia Gen­ adult fema les of a scarabaeid, Heptophylla picea. Applied En­ eralis, 9:143- 147. tomology and Zoology, 30:501-507. VAN HERCKE, L., MAELFAIT, J.-P. & DESENDER, K., 1980. Beetles TADA, S., TSUTSUM I, S., HATSUKADE, M., HONMA, K., FUJ ISAKI, captured by means of a light trap. Biologisch Jaw·boek K. & NAKASUJI , F., 1993. Sexual difference in flight abilities Dodonaea, 48:153-162. and fli ght muscle dimorphism in female adults of a chafer, WONNACOTT, T. H. & WONNACOTT, R. J. , 1977. Introductory Anomala schonfeldti Ohaus (Coleoptera: Scarabaeidae). Ap­ Statistics (Third Edition). New York: John Wiley & Sons, plied Entomology and Zoology, 28:333-338. 650 pp. TADA, S. , YAMAMOTO, A. & NISH IGAKI, J., 1991. Flight Muscle ZERA , A. J., POTTS, J. & KOBUS, K. , 1998. The physiology of Dimorphism of Female Adults in the Yellowish Elongate Cha­ fer, Heptophylla picea Motschulsky (Coleoptera: Scarabaei­ life-history trade-offs: Experimental analysis of a hormonally­ induced life-history trade-off in Gty llus assimilis. American dae). Applied Entomology and Zoology, 26:5 15-521. Naturalist. 152:7-23. THIELE, H. U. ( 1977) Carabid beetles in their environments. Berlin, Heidelberg, New York: Springer Verlag, 369 pp. ZERA, A. & DEN NO, R. F., 1997. Physiology and ecology of dispersal polymorphism in insects. Annual Review of Entomol­ TI ETZE, F. , 1963. Untersuchungen iiber die Beziehungen zwi­ ogy , 42:207-23 1. chen Fli.igel-reduktion und Ausbildung des Metathorax bei Carabiden unter besonderer Beriicksichtigung der Flugmusku­ ZHANG, J. X., DRUMMOND, F. A. , LI EBMAN, M. & HARTKE, A., latur (Coleoptera: Carabidae). Beitriige zur Entomologie, 1997. Phenology and dispersal of Hwpalus rufipes DeGeer 13:88-167 .. (Coleoptera: Carabidae) in agroecosystems in Maine. Journal of Agricultural En tomology, 14: I 71 -1 86. VAN HUIZEN, T. H. P. , 1990. "Gone with the Wind": Flight Activity of Carabid Beetles in Relation to Wind Direction and ZULKA, K. P., 1994. Carabids in a Central European floodplain: to Reproductive State of Females in Flight. In : STORK, N. E. species distribution and survival during inundations. In: DESEN­ (eds): The Role of Ground Beetles in Ecological and Environ­ DER, K., DUFRENE, M., LOREAU, M., LUFF, M. L., & MAELFAIT, mental Studies. Andover, Hampshire: Intercept, pp. 289-294. J.-P. (eds): Carabid Beetles: Ecology and Evolution. Dordrecht, VAN HUIZEN, T. H. P., 1979. Individual and environmental Boston, London: Kluwer Academic Publishers, pp. 399-405. fac tors determining flight in carabid beetles. Miscellaneous Papers Landbouw-Hogeschool Wageningen, 18: 199-2 12. Konjev DESENDER VAN HUIZEN, T. H. P., 1977. The Signi fica nce ofFii ght Activity Department Entomology in the Life Cycle of Amara plebeja Gyll. (Coleoptera, Carabi­ Royal Belgian Institute ofNatural Sciences dae). Oecologia (Berlin), 29:27-4 1. Vautierstr. 29 B-1000 Brussel VAN HUIZEN, T. H. P. , 1980. Species of Carabidae (Coleoptera) emai l: [email protected]