Carabid Beetle (Coleoptera: Carabidae) Communities in a Woodland Habitat in Hungary

Carabid beetle (Coleoptera: Carabidae) communities in a woodland habitat in Hungary

Rita Andorké & Ferenc Kadar

Andorko, R. & Kadar, F. 2006: Carabid beetle (Coleoptera: Carabidae) communities in a woodland habitat in Hungary. — Entomol. Fennica 17: 221—228.

Carabid communities were investigated in a woodland area within the long-term framework of the MAB project in Hungary, in 1985—86 and in 1993—94, using pitfall trapping. The structural characteristics ofthe carabid communitites and the habitat preferences of the most abundant species were studied in a beech wood, an oak forest, a transition zone and an ecotone. Altogether 7,636 carabid individuals were collected, representing 39 species. The value of Shannon diversity and the equitability consistently peaked in the transition zone. Further statistical analysis showed that the studied habitats were remarkably differentiated from each other.

R. Andorko, Department of Systematic Zoology and Ecology, Eo'tvo's Lorand University, Faculty ofSciences, Budapest, H—1 1 I 7, Pazmany Peter sétany I/C, Hungary; E—mail.‘ andorko.rita@fi”eemail.lzu; [email protected] F. Kadar, PlantProtection Institute, Hungarian Academy ofSciences, Budapest, H—I525, P. 0. Box 102, Hungary; E—mail.‘ [email protected]

Received 28 December 2005, accepted 18 May 2006

1. Introduction jority of the most abundant species were found only in one or two forest habitats during each In 1982, the Man and the Biosphere (MAB) pro- study year. Thus, at macrohabitat scale, the forest ject was initiated in the Pilis Biosphere Reserve is a patchy environment for carabids (Niemela et (PBR), a woodland region in northern Hungary. al. 1992). At a smaller scale, such as the micro- This project combined general ecological, coeno- habitat scale (within trapping grids), the occur- logical and faunistical studies among others, in- rence ofground beetles has also been shown to be cluding several studies on arthropods (Berczik aggregated (Niemela et al. 1986). Other studies 1984). The examination of carabid species at dif- also suggest that habitat fragmentation at a small ferent spatial and temporal scales was also carried spatial scale may lead to aggregated distributions out. ofground beetles within patches ofsimilar habitat Carabids respond to habitat-related differ- (Luff 1986, Niemela et al. 1986, 1992, Henge- ences at several ecological spatial scales, and rel- veld 1987). atively distinct carabid assemblages characterize The aim ofthe present study was to determine forest communities (Niemela et al. 1992). Kadar whether the composition of ground beetle com- and Szél (1999) studied ground beetle assem- munities and the abundance ofthe most abundant blages at 15 topographically different sites ofthe carabid species vary among different woodland PBR at landscape scale and showed that the ma- habitats in the short-term (one or two years) 222 Andorké & Kdddr ° ENTOMOL. FENNICA Vol. 17

and/or in the long-term (eight and nine years) The canopy layer was slightly open. This hab- temporal scales. We also studied the habitat pref- itat type was not sampled during 1985—86. erences ofthe most abundant carabid species. We examined the differences in composition and di- Carabids were collected using pitfall traps (plas- versity of the ground beetle communities in dif- tic jars of 80 mm diameter, 300 ml volume) con- ferent habitats at the macro-habitat scale. To as- taining 4% formaldehyde as a killing and preserv- sess the relationships between species and their ing agent, and having a metal lid above the traps. habitats, we investigated the presence/absence of The traps were emptied weekly from the end of species and/or their relative densities within the April to early September in two consecutive different habitats. Thus, we were mainly inter- years, in 1985 and 1986 and in 1993 and 1994. ested in how distinctive habitat associations dif- Five traps placed at least 5 m apart were placed in ferent species showed, based on their abundance a row within each area at the same sites each year. in pitfall catches. In 1985 and in 1986, trapping was conducted only at the first three sites.

2. Material and methods 2.2. Statistical analyses 2.1. Study area and sampling Following species identifications, the quantita- The study area was located in the PBR, near the tive carabid data were analysed using hierarchical Danube river, about 50 km from Budapest, north- classification method to explore differences ern Hungary (47°59’N, 18°54’E; 267 m a.s.l.). among the studied habitat types, using Ward’s The study was done in the following four sites, method with Manhattan distances (Podani 1997). treated as macro-habitats: Species diversity in each habitat was estimated using Shannon-Wiener index (H’) (Magurran (1) Beech wood (Melittio—Fagetum). Associated 1988). The average number of individuals, the plant species were the tree Carpinus betulus mean number of species, the mean values of the and, in within-stand patches, the herb Shannon-Wiener indices, and the mean values of Impatiens noli—tangere; the equitabilities ofH’ among the different habi- (2) Oak forest (Querco petreae—Carpinetum) sit- tats were tested using Kruskal—Wallis test and uated on the other side of the stream from the subsequent Mann-Whitney Utest (Sokal & Rohlf beech forest; the altitude was approximately 1981). The Statistica program package was used the same in both forests. for statistical calculations (StatSoft 2000). (3) Transition zone that was about 40 m x 8 m, sit- uated between the previously mentioned two habitats, i.e., bordered on one side by beech 3. Results wood and on the other side by an oak forest, and dominated by Asarum europaeum, 3.1. Species composition Urtica dioica, Fragarl'a vesca and nitrofil weeds. A deep ditch with periodic water-flow Altogether 7636 individuals were caught during separated the transition zone and the oak for- the four years, representing 39 species (Table 1). est. The shrub layer was absent, and the can- The most common species were large or medium opy layer was open. large, whereas most ofthe rare species were small (4) Ecotone with area-wise dimensions similar to to medium large. The most abundant carabid spe- the transition zone, located between the tran- cies was Aptinus bombarda (>37% ofall carabids sition zone and the beech wood. The charac- caught during the four years), being about four teristic plant was Carex pilosa, and the other times more abundantly caught than the second associated plants were the same as in the adj a- and third most abundant species Abax parallele— cent habitats. The shrub layer was moderate, pipedus and Pterostichus melanarius. The fol- dominated by saplings of Carpinus betulus. lowing four most abundant species were repre- ENTOMOL. FENNICA Vol. 17 ° Woodland carabid communities in Hungary 223

Table 1. Data on the carabid community structure in 1985-86 and in 1993-94 in the Pilis Biosphere Reserve, Hungary.

Variable 1985 1986 1993 1994 Total

Number of species 32 35 26 23 39 MeaniSE 10.4i0.92 118710.76 122510.64 12.4:068 Number ofindividuals 1,169 1,249 3,034 2,184 7,636 MeaniSE 77.93:11.43 83.27i9.34 151.7:1494 109.2i12.46 Diversity (H’)iSE 1.42i0.18 1.59i0.12 1.61:0.10 1 9610.07 EquitabilityiSE 0.61i0.06 0.65i0.04 0.65:0.04 0.79:0.02

Table 2. Ground beetle species collected in the Pilis Biosphere Reserve, Hungary, in 1985—86 and in 1993—94 (see also footnote). Frequency = mean relative frequency.

Species Frequency (%) Total catch

Aptinus bombarda (llliger, 1800) 37.32 2,850 Abax parallelepipedus (Piller et Mitterpacher, 1783) 8.00 611 Pterostichus melanarius (llliger, 1789) 7.89 603 Pterostichus melas (Creutzer, 1799) 7.67 586 Platynus assimilis (Paykull, 1790) 7.58 579 Carabus ullrichi Germar, 1824 7.23 552 Carabus scheidleri Panzer, 1799 6.06 463 Pterostichus oblongopunctatus (Fabricius, 1787) 3.84 294 Abax carinatus (Duftschmid, 1812) 3.82 292 Carabus nemoralis M'Liller, 1764 2.38 182 Carabus coriaceus Linnaeus, 1758 1.83 140 Trechus pilisensis Csiki, 1918 1.17 89

Further species collected: Carabus convexus Fabricius, 1775, Mo/ops piceus (Panzer, 1793), Abax ova/is (Duftschmid, 1812), Pterostichus niger (Schaller, 1783), Leistus rufomarginatus (Duftschmid, 1812), Pterostichus ovoideus (Sturm, 1824), Notiophilus rufipes Curtis, 1829, Abax parallelus (Duftschmid, 1812), Harpalus rufipes (De Geer, 1774), Harpalus marginellus Dejean, 1828, Harpalus atratus Latreille, 1804, Cychrus caraboides (Linnaeus, 1758), Amara saphyrea Dejean, 1828, Stomis pumicatus (Panzer, 1796), Platynus dorsalis (Pontoppidan, 1763), Synuchus viva/is (llliger, 1789), Nebria brevicollis (Fabricius, 1792), Carabus intricatus Linnaeus, 1761, Pterostichus strenuus (Panzer, 1797), Calosoma inquisitor (Linnaeus, 1758), Platyderus rufus (Duftschmid, 1812), Bembidion lampros (Herbst, 1784), Licinus hoffmannseggi (Panzer, 1797), Harpalus pumilus Sturm, 1818, Brachinus explodens Duftschmid, 1812, Carabus vio/aceus Linnaeus, 1758, Amara convexior Stephens, 1828.

sented in almost the same number of individuals was hardly represented in the 1985—86 samples, it (6—7%): P. melas, Plalynus assimilis, Carabus was the second most abundant carabid species in ullrichi and C. scheidleri. The mean relative fre- 1994, and the fourth one in 1993, with more than quency and the total number ofindividuals ofthe 200 individuals. more frequent twelve species are given in Table 2. Twenty-two common species occurred during the four examined years. 3.2. Hierarchical classification In 1985—86, A. parallelepipedus was only the sixth (1985) and seventh (1986) most abundant The results of the cluster analysis for the pitfall species with about 40 individuals in each year. In trap catches are given in Figs. 1—2. The clusters 1993—94, it was the second (1993) and the fourth formed three groups of habitat; these were well (1994) most abundant species (227 and 307 indi- separated from each other at macro-scale in 1985 viduals, respectively). In 1986, C. scheidleri was (Fig. 1). However, in 1993, the oak forest and collected in high numbers (332 individuals). In beech wood formed isolated groups, and little contrast, during the other three years this species overlap occurred between the transition zone and occurred only in low numbers. While C. ullrichi the ecotone at a smaller scale (Fig. 2).

ENTOMOL. FENNICA Vol. 17 ° Woodland carabid communities in Hungary 225

Table 3. Data on community structure of carabids of different habitats in 1985 and 1986 (upper part) and in 1993 and 1994 (lower part) in the Pilis Biosphere Reserve, Hungary (mean $ SE; lower-case letters indicate significant (p<0.05) differences by Mann-Whitney U test). The Ecotone habitat was not sampled in 1985—86. Beech = beech forest; Transition = transition zone; Oak = oak forest; Ecotone = ecotone habitat; K—W test = Kruskal-Wallis test (with H2 and p values).

1985 1986 BeechTransition Oak (Ecotone) K—W test Beech Transition Oak (Ecotone) K-W test

Mean no. species 9 14 8.2 H2=8.96; 10.8 15 9.8 H2=9.34; $0.84 a $1.05 b $1.46 a p=0.011 $0.86 a $0.95 b $0.73 a p=0.009

Mean no. individuals 115.2 67.2 51.4 H2=5.46; 76 120.4 53.4 H2=10.52; $24.05 b $11.94 $9.91 a p=0.065 $5.29 a $14.44 b $9.94 a p=0.005 ab Diversity (H') 0.71 2.22 1.31 H2=11.58; 1.06 1.95 1.78 H2=9.92; $0.12 a $0.06 0 $0.17 b p=0.003 $0.08 a $0.09 b $0.13 b p=0.007

Equitability 0.32 0.85 0.66 H2=12.52; 0.45 0.72 0.78 H2=10.82; $0.05 a $0.02 0 $0.05 b p=0.002 $0.04 a $0.02 b $0.04 b p=0.004

1993 1994 Beech Transition Oak Ecotone K—W test Beech Transition Oak Ecotone K-W test

Mean no. species 12 15 9.4 12.6 H2=12.83; 11 14.8 8.8 15 H2=14.49; $0.77 b $1.52 b $0.6 a $0.68 b p=0.005 $0.89 a $0.2 b $0.49 a $0.95 b p=0.002

Mean no. individuals 194.4 126.4 110 176 H2=4.83; 83.2 95.4 77.6 180.6 H2=7.72; $28.85 a $22.62 a $13.29 a $38.99 a p=0.185 $11.3 a $15.22 a $7.86 a $28.02 b p=0.052

Diversity (H') 1.07 2.09 1.46 1.81 H2=15.79; 1.75 2.25 1.65 2.18 H2=13.19; $0.19 a $0.05 0 $0.07 a $0.09 b p=0.001 $0.11 a $0.1 b $0.05 a $0.05 b p=0.004

Equitability 0.43 0.78 0.66 0.72 H2=12.41; 0.74 0.84 0.76 0.81 H2=6.79; $0.08 a $0.02 0 $0.03 b $0.03 bc p=0.006 $0.04 a $0.04 b $0.01 ab $0.02 ab p=0.079

Pterostichus melanarius was mainly found in sition zone. A similar pattern was found for C. the transition zone and in the ecotone during the ullrz'chi that had fewer individuals collected in the four years (Fig. 4), as was P. oblongopunctatus. beech forest. Out of the 332 collected C. The abundance of P. melanarius in the studied scheidleri individuals, 248 were collected in the habitats differed significantly in 1985—86. Based transition zone. on abundance, there was also a significant differ- ence among the four habitats in 1993—94. Pterostichus melas was caught almost exclu- 4. Discussion sively in the oak forest each year (Fig. 5). There were significant differences among the abun- 4.1. Variation within ecological scales dances of P. melas in the studied habitats in 1985—86, and the abundance also significantly In the only published investigation on ground varied among the four habitats in 1993—94. beetle communities of this area (Kadar & Szél Abax parallelepipedus was collected in a 1999), the composition and the occurrence of ca- higher number in 1993—94 (534 individuals), and rabid species were examined between 1982 and only 34 individuals were found in the oak forest. 1984 at the landscape scale. At the scale of forest Many more individuals of P. assimilis were habitat patches, such as at the macro-habitat collected in 1993—94 (479) than in 1985—86 scale, soil moisture appears an important determi- (100), and mainly from the ecotone and the tran- nant ofcarabid distribution (Thiele 1977, Epstein

ENTOMOL. FENNICA Vol. 17 ° Woodland carabid communities in Hungary 227 the composition of the carabid assemblages we with the transition zone and the ecotone, i.e., hab- found, for example by increasing the diversity itats with high level of soil moisture and dense and the equitability during the ten years in each shrub vegetation. It can be found in relatively habitat (Table 3). For example, A. parallele— high numbers in quite different habitats, too; for pipedus increased in 1993—94 probably because example, it is common in agricultural landscapes of the cessation of the forest management. There (Thiele 1977) and riverside woodlands (Baguette were also more individuals in the beech forest 1993). Pterostichus melanarius has also occurred during these two years, but this is an ambiguous in high numbers in various other environment consequence because of the lack of continuous types, such as fields, pastures, abandoned fields, survey between the periods 1985—86 and 1993— and in forests in Finland (Niemela and Halme 94. The “natural” fluctuation of A. parallele— 1992). Although this species is favoured by hu- pipedus population could be another reason be- man activity, our study and that of Niemela and hind this phenomenon. This common species is Spence (1991) also showed that it was not solely well adapted to many environmental conditions, restricted to disturbed habitats. but its distribution might change at a larger scale, Pterostichus melas also showed a clear habi- for example following large-scale forest harvest- tat association, as it was found mainly in the oak ing (Loreau & Nolf 1994). Such changes might forest every year. It is characteristic for different have taken place in our study site, too. Carabus kinds of oak forests such as Querco—Ostryetum scheidleri is associated with disturbed, human- (Contarini 1986) and Querceto—petraeae cerris modified areas (Andorko et al. 2003), which may (Fazekas et al. 1992). Plalynus assimilis was only explain why this species was found in high num- occasionally found in the PBR between 1982 and bers in 1985—86 but was missing in 1993—94. 1984 (Kadar & Szél 1999), whereas in our study Carabus ullrz'chi occurs in more stable forests, it was commonly caught. Similarly to our study, it possibly explaining its distribution in 1993—94. is reported to occur in wet, shady forests (Kleinert 1983). In the present study, Carabus ullrichi was captured from each habitat. It has also been col- 4.4. Habitat associations lected in the PBR between 1982 and 1984, but in low numbers (Kadar & Szél 1999). It occurs in All the abundant species showed associations to wet and shady but also in dry and light forests in particular habitats. The distribution of the most Slovakia (Kleinert 1983). Carabus scheidleri is abundant species, A. bombarda, is well known common in lowlands and hills in mixed forests, (Brandmayr 1974, Casale & Vigna Taglianti and is associated with fields, meadows, gardens, 1983, Pravisani & Torossi 1987, Fazekas et al. hedges, watersides and various types of forests 1992). This species has been found only in the (Hfirka 1996). It has also been found in an aban- Seslerio—Quercetum patch with an extremely doned agricultural field (Andorko et al. 2003). high dominance value (Brandmayr et al. 1980). It Our results emphasize that carabids are ap- is a typical forest-dwelling species in Hungary propriate organisms for ecological studies both at (Kadar & Szél 1993). In our study it was most macro-scale and landscape levels. We conclude abundant in the beech forest, similarly to Magura that carabid beetles are good indicators of site et al. (2002). The second most abundant species type, quality and conservation status (Eyre & in our study, A. parallelepipedus, was also among Rushton 1989). the most abundant species captured by Kadar and Szél (1993, 1999), and Magura et al. (2002). This carabid is an and typically opportunistic species Acknowledgements. The authors are gratefiil to A. Szentesi has a good adaptation ability to various habitat for critical comments and improving the English of the types (Kadar & Szél 1993). That is why it is a manuscript. The authors thank the reviewers for cements on this We thank the dominant species in different habitats (Chemini manuscript. Middle-Danube-Valley Environmental, Nature and Water Inspectorate to approve & Werth 1982, Contarini 1986, Pravislani & the permission for the investigations. This work was sup— Torossi 1987). ported partly by the Hungarian National Scientific Re- In our study, P. melanarius was associated search Foundation (No. OTKA T048434). 228 Andorké & Kdddr ° ENTOMOL. FENNICA Vol. 17

References ground beetles in different habitats ofthe Nagy-szenas Nature Reserve (Coleoptera, Carabidae). 7 Folia 54: 65773. Andorko, R., Kadar, F. & Szekeres, D. 2005: Reproductive Entomologica Hungarica F. & 1999: and oc— characteristics of Carabus scheidleri (Coleoptera, Ca— Kadar, Szel, Gy. Species composition currence 7 of beetles in rabidae) in Hungary. In: Lovei, G. & Toft, S., Euro— ground (Coleoptera, Carabidae) the Pilis Hungary: a trap pean Carabidology 114: 9716. DIAS Report. 401 pp. BiosphereReserve, pitfall 7 Folia 60: 2057212. Baguette, M. 1993: Habitat selection of carabid beetle in study. Entomologica Hungarica deciduous woodlands of southern Belgium. 7 Kleinert, J. 1983: Ground beetles (Coleoptera, Carabidae) Pedobiologia 37: 3657378. in selected habitats of lowland in southeastern Berczik, A. 1984: A Pilis Bioszfera Rezervatum kutatasi Slovakia. 7Biologia 38 (10): 9817989. M. & C.—L. 1994: structure and programj a. (The research programme ofthe Pilis Bio- Loreau, Nolf, Spatial dy—

7 sphere Reserve). 7 Allattani Kozlemények 71: 147 namics of a population ofAbax ater. In: Desender, 16. K. et al. (eds.), Carabid Beetles: Ecology and Evolu— Bommarco, R. & Fagan, W. F. 2002: Influence of crop tion: 1657169. Kluwer Academic Publishers, the edges on movement of generalist predators: a diffu— Netherlands. 492 pp. sion approach. 7 Agricultural and Forest Entomology Luff, M. L. 1986: Aggregation ofsome Carabidae in pitfall

7 In: P. J M. 4: 21730. traps. den Boer, ., Luff, L., Mossakowski, Brandmayr, P. 1974: Auto- und synokologische Unter— D. & Weber, F. (eds.), Carabid beetles, their adapta— suchungen uber die Carabiden zweier Vegetations— tions and dynamics: 3857397. Gustav Fischer, einheiten des Slowenischen Kustenlandes: das Carici Stuttgart. 551 pp. (humilis) 7 Seslerietum juncifoliae und das Seslerio Magura, T., Elek, Z. & Tothmeresz, B. 2002: Impacts of

7 7 (autumnalis) Fagetum (Coleoptera, Carabidae). non—native spruce reforestation on ground beetles. 7 Acta Entomologica Jugoslavica 10: 15710. European Journal of Soil Biology 38: 2917295. Brandmayr, P., Colombetta, G., Drovenik, B., Forti, F., Magurran, A. E. 1988: Ecological diversity and its mea—

L. & T. Z. 1980: Etude multi— 7 Poldini, Brandmayr, surements. University Press, Cambridge. 179 pp. disciplanaire sur l’e'cologie de quelques carabides Murcia, C. 1995: Edge effects in fragmented forests: im—

7 endoges (Coleoptera, Carabidae). Mem. Biospeol. plications for conservation. 7 Trends in Ecology and 7: 85798. [In French.] Evolution 10: 58762. A. & A. 1983: Il Casale, Vigna Taglianti, genere Aptinus Niemela, J. & Halme, E. 1992: Habitat associations of ca- 1810 7 Boll. Mus. Bonelli, (Coleoptera, Carabidae). rabid beetles in fields and forests on the Lland Islands, Reg. Sci. Nat. Torino 1: 21758. SW Finland. 7 Ecography 15: 3711. C. & F. 1982: Censimenti di Carabidi in Chemini, Werth, Niemela, J. & Spence, J. R. 1991: Distribution and abun— tre ambienti forestali di e Favogna di Magre (Provincia dance of an exotic ground—beetle (Carabidae): a test of

Studi 7 Bolzano) (Insecta, Coleoptera, Carabidae). community impact. Oikos 62: 3517359. Trentini Sci. Nat. 59: 2017211. Niemela, J., Haila, Y. & Ranta, E. 1986: Spatial heteroge- Contarini, E. 1986: Eco-profili d’ambiente della Cole— neity ofcarabid beetle dispersion in uniform forests on otterofauna di Il Romagna:1. Querco—Ostryetum 7 the Aland Islands, SW Finland. Annales Zoologici dell’orizzonte submontano. 7 Boll. Ass. Romana Fennici 23: 2897296. Entomol. 41: 1762. Niemela, J., Spence, J. R. & Spence, D. H. 1992: Habitat Desender, K., Ervynck, A. & Tack, G. 1999: Beetle diver— associations and seasonal activity of ground—beetles and historical of woodlands in Flanders. sity ecology in Central Alberta. 7 Cana-

7 (Coleoptera, Carabidae) Belgian Journal of Zoology 129: 1397156. dian Entomologist 124: 5217540. Epstein, M. E. & Kulman, H. M. 1990: Habitat distribution Podani, J. 1997: Bevezetes a tobbvaltozos adat- and seasonal occurrence ofcarabid beetles in east—cen— biologiai feltaras introduction to methods of tral Minnesota. 7 American Midland Naturalist 123: rejtelmeibe. (An multivariate data and evaluat— 2097225. biological processing ing). 7 Scientia Kiado, Budapest. 412 pp. [In Hungar- Eyre, M. D. & Rushton, S. P. 1989: Quantification of conservation criteria using invertebrates. 7 Journal of ian.] L. & G. 1987: Coleotteri carabidi in Applied Ecology 26: 1597171. Pravisani, Torossi, alcune biocenosi dell’Alta Val Torre 7 Atti. Fazekas, J., Kadar, F. & Lovei, G. 1992: Comparison of (Friuli). ground beetle assemblages (Coleoptera, Carabidae) of Mus. Friul. Stor. Nat. 9: 1697200. [In Italian.] P. G. 1995: The status of the science an abandoned apple orchard and the bordering forest. Risser, examining

7 7 Acta Phytopathologica Hungarica 27: 2337238. ecotones. Bioscience 45: 3187325. Hengeveld, R. 1987: Scales ofvariation: Their distinction Sokal, R. R. & Rohlf, F. J. 1981: Biometry. Second edition.

7 and ecological importance. 7 Annales Zoologici Fen- W. H. Freeman and Co., New York. 859 pp. nici 24: 1957202. StatSoft 2000: Statistica for Windows 17111. 7 StatSoft Hurka, K. 1996: Carabidae of the Czech and Slovak Re— Inc., Tulsa, OK. publics. 7 Kabourek, Zlin. 566 pp. Thiele, H.—U. 1977: Carabid beetles in their environment. Kadar, F. & Szel, Gy. 1993: Analysis ofthe distribution of 7 Springer—Verlag, Berlin. 369 pp.