scieee Open visual document viewer

Body size inequality of carabids along an urbanisation gradient

Magura, Tibor; Tóthmérész, Béla; Lövei, Gábor L.

Full text

UNCORRECTED PROOF Basic and Applied Ecology ](]]]])]]]—]]] Body size inequali y o ca abids along an u banisa ion g adien Un e schiede de Ko¨ pe g o¨ße bei Ca abiden en lang eines U banisa ionsg adien en Tibo Magu a a,  ,Be´la To´ hme´ e´sz b ,Ga´bo L. Lo¨ ei c a Ho oba´gy Na ional Pa k Di ec o a e, H-4002 Deb ecen, POB. 216, Hunga y b Depa men o Ecology, Uni e si y o Deb ecen, H-4010 Deb ecen, POB. 71, Hunga y c Depa men o In eg a ed Pes Managemen , Danish Ins i u e o Ag icul u al Sciences, Flakkebje g Resea ch Cen e, DK-4200 Slagelse, Denma k Recei ed 3 Ma ch 2005; accep ed 8 Augus 2005 KEYWORDS G ound bee les; Dis u bance; Globene p ojec ; Skewness; Gini coe icien ; Lo enz asymme y coe icien Summa y Analysis o size inequali y can shed ligh on coexis ence mechanisms and help o in e p e pa e ns in assemblages. We es ed se e al measu es o hei powe o e alua e changes in ca abid body size along an u banisa ion g adien (ci y pa k–subu ban a ea– u al), ep esen ing dec easing in ensi ies o human dis u - bance. Ca abids we e collec ed by pi all aps o e wo ull ac i i y pe iods in lowland oak o es pa ches in and nea he ci y o Deb ecen, Eas e n Hunga y. The a e age alue o skewness was la ges in he u ban a eas compa ed o he subu ban and u al ones, indica ing ha small indi iduals we e mo e p ominen in he u ban a eas. The Gini coe icien also dec eased om u ban owa ds u al a eas, sugges ing ha inequali y in body size o he ca abid assemblages dec eased along he g adien . Howe e , nei he o hese ends was signi ican . The Lo enz asymme y coe icien was signi ican ly highe in u al a eas compa ed o subu ban and u ban a eas indica ing ha he e was a signi ican di e ence in inequali y and/ o asymme y o body size ac oss he g adien . This di e ence was p ima ily due o mo e indi iduals wi h la ge body size in u al a ea. We sugges ha he obse ed a ia ion in ca abid body size along he g adien is ela ed o habi a al e a ion caused by u banisa ion. &2005 Published by Else ie GmbH on behal o Gesellscha u¨ O ¨kologie. 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 ARTICLE IN PRESS www.else ie .de/baae 3B2 8:06a=wðDec 5 2003Þ:51c XML: e :5:0:1BAAE : 50051 P od:Type:FTP pp:111ðcol: ig::NILÞ ED:Ashim PAGN:anu SCAN:kp ashan h 1439-1791/$ - see on ma e &2005 Published by Else ie GmbH on behal o Gesellscha u¨ O ¨kologie. doi:10.1016/j.baae.2005.08.005  Co esponding au ho . Tel.: +36 52 529920; ax: +36 52 529940. E-mail add esses: magu [email protected] (T. Magu a), [email p o ec ed] (B. To´ hme´ e´sz), gabo [email p o ec ed] (G.L. Lo¨ ei). UNCORRECTED PROOF Zusammen assung Die Analyse on G o¨ßenun e schieden kann die Koexis enzmechanismen beleuch en und hel en, Mus e on Ansammlungen zu in e p e ie en. Wi un e such en e schiedene Pa ame e au ih Po enzial die Ve a¨nde ungen de Ko¨ pe g o¨ße bei Ca abiden en lang eines U banisa ionsg adien en (S ad pa k–Vo s ad gebie –la¨nd- lich) zu e kla¨ en, de eine abnehmende In ensi a¨ de an h opogenen S o¨ ung ep a¨sen ie e. Die Ca abiden wu den in Boden allen u¨be zwei olls a¨ndige ak i e Pe ioden in Tie landeichenwa¨lde n und in de Na¨he de S ad Deb ecen im o¨s lichen Unga n gesammel . De du chschni liche We de Schie e wa in den s a¨d ischen Gebie en im Ve gleich zu den o s a¨d ischen und la¨ndlichen am g o¨ß en und wies da au hin, dass kleine Indi iduen in den s a¨d ischen Gebie en meh he o a en. De Gini- Koe izien e inge e sich eben alls on den s a¨d ischen zu den la¨ndlichen Gebie en und la¨ß e mu en, dass die Un e schiede in de Ko¨ pe g o¨ße in den Ca abiden-Ansammlungen en lang des G adien en abnahmen. Keine diese T ends wa jedoch signi ikan . De Lo enz-Asymme ie-Koe izien wa in den la¨ndlichen Gebie en im Ve gleich zu den o s a¨d ischen und s a¨d ischen Gebie en signi ikan g o¨ße und wies da au hin, dass es einen signi ikan en Un e schied in den Ko¨ pe g o¨ßeun e schieden und/ode in de Asymme ie de Ko¨ pe g o¨ße en lang des G adien en gab. Diese Un e schied wa o allem da au zu u¨ckzu u¨h en, dass es in den la¨ndlichen Be eichen meh Indi iduen mi eine g o¨ße en Ko¨ pe g o¨ße gab. Wi e mu en, dass die beobach e e Va ia ion in de Ko¨ pe g o¨ße de Ca abiden en lang des G adien en mi de du ch die U banisa ion e u sach en Ve a¨nde ung de Habi a e e bunden is . &2005 Published by Else ie GmbH on behal o Gesellscha u¨ O ¨kologie. In oduc ion Rela ionships be ween body size and he s uc- u e o animal assemblages ha e been he ocus o much a en ion in ecological s udies. Body size is co ela ed wi h many aspec s o li e his o y ( ep oduc ion a e, dispe sal, de elopmen ime, e c.) (Pe e s, 1983). Body size also has a signi ican impac on ecological in e ac ions, esou ce use, and mo e indi ec ly, he pe iod o ac i i y, habi a sui abili y, and nume ous o he pa ame e s (Pe e s, 1983). A change in body size, ei he in indi idual species, o in he size dis ibu ion o he species p esen in a habi a is also a pa ame e po en ially indica ing di e en ypes o en i onmen al s ess (McGeoch, 1998). A numbe o an h opogenic ac i i ies, including a ming, o es y and u banisa ion, ha e a signi i- can impac on he en i onmen and c ea e pa ch- wo ks o modi ied land ypes ha exhibi simila pa e ns h oughou he wo ld (Poschlod, Bakke , & Kahmen, 2005;Ul ich & Buszko, 2004). Global u banisa ion has caused he loss o as amoun s o habi a and caused majo modi ica ions o he en i onmen al condi ions (Ta ainen, Ma kkola, & S omme , 2003). Howe e , li le is known on whe he o no hese changes a ec biodi e si y in simila ways ac oss he globe (Niemela¨ e al., 2000). In 1998, an in e na ional collabo a i e e o o sea ch o gene alisa ions in u banisa ion im- pac s on biodi e si y was ini ia ed. The p ojec , called Globene (Niemela¨ e al., 2000), examines u ban–subu ban– u al g adien s, using a common me hodology and a ge in e eb a e axon (g ound bee les; Fam. Ca abidae) (Niemela¨ e al., 2000). This axon was selec ed, because ca abids a e especially use ul ecological indica o s o s udy en i onmen al impac s, being sensi i e o habi a modi ica ions and en i onmen al changes, abun- dan and su icien ly a iable bo h axonomically and ecologically (Lo¨ ei & Sunde land, 1996). The esul s published so a ocussed mainly on he changes o ca abid assemblage composi ion along he g adien (Niemela¨ e al., 2002), wi h some conside a ion o he e ec s o u banisa ion on body size (Ishi ani, Ko ze, & Niemela¨, 2003;Magu a, To´ hme´ e´sz, & Molna´ , 2004). Va ia ion in body size has adi ionally been desc ibed and analysed using he skewness o he size dis ibu ion, o o he s a is ics de i ed om he s a is ical momen s o he dis ibu ion (Sokal & Rohl , 1995). Recen ly, he ocus has shi ed owa d an emphasis on inequali y in size. Se e al measu es o inequali y, de eloped o use in economics (Sen, 1973), ha e been used o analyse a ia ion in size wi hin assemblages. These measu es use he Lo enz cu e (Lo enz, 1905), whe e indi iduals a e anked by size, and he cumula i e p opo ion o s udy 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 ARTICLE IN PRESS BAAE : 50051 T. Magu a e al.2 UNCORRECTED PROOF objec s is plo ed agains he cumula i e p opo - ion o hei o al size on he y-axis. I all indi iduals a e equal in size, he Lo enz cu e is a diagonal line, called he ‘‘line o equali y’’ (Fig. 1). Inequali y causes he line o un below his line, and he g ea e he inequali y among he s udy objec s, he lowe he cu e uns below he line o equali y. One app oach o quan i y his is he Gini coe icien (Dixon, Weine , Mi chell-Olds, & Wood- ley, 1987;Gini, 1912;Sen, 1973). Howe e , he Gini coe icien is only ela ed o he size (a ea) and no he shape o he cu e. Thus, he Gini coe icien does no con ain all he in o ma ion in he Lo enz cu e. Di e en Lo enz cu es can ha e he same Gini coe icien (Damgaa d & Weine , 2000;Shum- way & Koide, 1995;Weine & Solb ig, 1984). The e o e, Damgaa d and Weine (2000), o cha - ac e ise he shape o he Lo enz cu e, p oposed a so-called ‘‘Lo enz asymme y coe icien ’’. This coe icien cha ac e ises an impo an aspec o he shape o a Lo enz cu e: i shows which size classes con ibu e mos o he o al inequali y o he assemblage (Damgaa d & Weine , 2000). The new index was illus a ed by an example om plan ecology, bu we seek o ex end i s use o he analysis o size ela ionships in animal assem- blages. In his s udy, we used pi all da a, collec ed ac oss an u ban–subu ban– u al g adien o e 2 yea s, o analyse he body size inequali y o g ound bee le (Ca abidae) assemblages. Using measu es desc ibing asymme y and/o inequali y o body size pa e n in ca abid assemblages, a hypo hesis, sugges ed by Szyszko (1983),G ay (1989) and Blake, Fos e , Ey e, and Lu (1994) was es ed. Acco ding o his ‘‘dec easing body size hypo h- esis’’, smalle ca abids should be ound in habi a s wi h highe dis u bance le els han in hose wi h lowe dis u bance. In ou case, he hypo hesis p edic s ha he mean ca abid body size should dec ease om he u al o he u ban a ea. We ound ha he Lo enz asymme y index was he mos powe ul me hod o de ec ends in size along he g adien . Ma e ial and me hods Cha ac e ising he body size dis ibu ions The ollowing measu es we e used o desc ibe he asymme y and/o inequali y o body size pa e n in ca abid assemblages. (1) Skewness. The asymme y o a uni a ia e con inuous dis ibu ion is commonly measu ed by he classical skewness coe icien (Sokal & Rohl , 1995), which is de ined as g¼Sn i¼1ðxi¯ xÞ3 ns3, whe e nis he numbe o indi iduals, x i is he body size o indi iduals i,¯ xis he mean body size and sis he s anda d de ia ion o body size. A symme ic dis ibu ion has ze o skewness, i.e. g¼0. An asymme ic dis ibu ion wi h a longe le ail has nega i e skewness (in ou case: la ge indi iduals a e dominan ), while a posi i e gindica es skewness o he igh (smalle indi iduals a e dominan ). (2) Medcouple. Since he skewness es ima o is based on he i s h ee momen s o he da a se , i is s ongly in luenced by he p esence o ou lie s; hus, a obus measu e o skewness, he medcouple (B ys, Hube , & S uy , 2004) was also used. I has a 25% b eakdown alue and a bounded in luence unc ion. The possible alues o medcouple ange om 1 o1. Fo no a ional con enience, he elemen s o he da a se a e so ed such ha x½1px½2ppx½n. Le med(X n ) deno e he median o he da a se X n , de ined as medðXnÞ¼ ðx½n=2þx½n=2þ1Þ=2i nis e en; x½nþ1=2i nis odd: ( The medcouple is de ined as MCn¼med hðx½i;x½jÞ;x½ipmedðXnÞpx½j  , whe e he ke nel unc ion his de ined by 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 020406080100 0 20 40 60 80 100 popula ion C popula ion B popula ion A line o equali y Cumula i e pe cen age o body size Cumula i e p e cen a g e o indi iduals Figu e 1. Lo enz cu es o h ee hypo he ical popula- ions. All popula ions ha e he same Gini coe icien , bu di e en Lo enz asymme y coe icien s. In he case o he popula ion A, he Lo enz asymme y coe icien (S)is la ge han one (S41), in he case o popula ion B, So1, while size dis ibu ion in popula ion Cis symme ic, leading o S¼1. Body size inequali y 3 UNCORRECTED PROOF hðx½i;x½jÞ¼ðx½imedðXnÞÞ  ðmedðXnÞx½jÞ x½ix½j o all x½iax½j.B ys e al. (2004) also desc ibed a as algo i hm o compu e he alue o he medcouple. (3) Gini coe icien . A adi ional g aphical ap- p oach o measu e inequali y in size dis ibu ion is he Lo enz cu e (Sen, 1973;Weine & Solb ig, 1984). Indi iduals a e anked by size and he cumula i e p opo ion o indi iduals is plo ed agains he co esponding cumula i e p opo ion o hei o al size. When all indi i- duals a e o he same size, he Lo enz cu e is a s aigh diagonal line, called he line o equal- i y. I he e is any inequali y in size, he Lo enz cu e uns below he line o equali y (Fig. 1). The o al amoun o size inequali y can be quan i ied by he Gini coe icien (Gini, 1912), which is he a io be ween he a ea enclosed by he line o equali y and he Lo enz cu e, and he o al iangula a ea unde he line o equali y. The Gini index o agg ega ion is based on o de ed da a by inc easing body size as ollows (Dixon e al., 1987): G¼Pn i¼1ð2in1Þx½i n2¯ x, whe e nis he numbe o indi iduals, x½iis he o de ed body size o indi iduals iand ¯ xis he mean body size. The Gini coe icien calcula ed by he abo e equa ion should be mul iplied wi h n=ðn1Þ o ob ain an unbiased es ima e (Glas- se , 1962). The Gini coe icien anges om a minimum alue o ze o, i all indi iduals ha e he same body size, o a maximum o one in a hypo he- ical assemblage in which e e y indi idual excep one has a size o ze o. Howe e , i has been demons a ed (Damgaa d & Weine , 2000; Shumway & Koide, 1995;Weine & Solb ig, 1984) ha di e en Lo enz cu es (assemblages wi h di e en inequali y in size) can ha e he same Gini coe icien (example on Fig. 1). (4) Lo enz asymme y coe icien . To complemen he abo e-men ioned Gini coe icien , Dam- gaa d and Weine (2000) p oposed he Lo enz asymme y coe icien , o quan i y he asym- me y o he Lo enz cu e. The coe icien (S) can be calcula ed om he o de ed body size da a using he ollowing equa ions (Damgaa d & Weine , 2000): S¼Fð^ xÞþLð^ xÞ¼mþd nþLmþdx0mþ1 Ln , whe e d¼¯ xx0m x0mþ1x0m and ¯ xis he mean body size, mis he numbe o indi iduals wi h a body size less han ¯ x,L m is he cumula i e body size o indi iduals wi h a body size less han ¯ x, and L n is he cumula i e body size o all indi iduals. When S¼1, he Lo enz cu e o he assemblage is symme ic, while o he S alues ep esen asymme ic Lo enz cu es. When S41, mos o he inequali y wi hin he assemblage is due o he la ges indi iduals, which disp opo iona ely con- ibu es o he cumula i e body size (mass) o he assemblage. When So1, he inequali y demon- s a ed in he assemblage is due p ima ily o he ela i ely la ge numbe o small indi iduals (Fig. 1; Damgaa d & Weine , 2000). S udy a ea and sampling me hods G ound bee les we e s udied along an u ban–su- bu ban– u al g adien in Deb ecen (Eas e n Hun- ga y), he second la ges ci y o he coun y (Magu a e al., 2004). The u ban, subu ban and u al sampling a eas we e all pa o a once- con inuous o es (Nagye d+ o Fo es Rese e) bo - de ing he ci y. All a eas we e si ua ed in con in- uous pa ches o old o es (4100 yea s) domina ed by English oak (Que cus obu ). The ypical, na i e o es associa ion o he sampling si es was Con alla io-Que ce um. The c i e ion o dis in- guishing sampling a eas (u ban, subu ban, u al) was he a io o he buil -up a ea o he na u al habi a s. The a ea o he buil -up en i onmen and he na u al habi a s was measu ed by he A cView GIS p og am using an ae ial pho og aph. In he u ban a ea he buil -up a ea exceeded 60%, in he subu ban a ea i was app oxima ely 30%, while in he u al a ea he buil -up a ea was 0%. The o es agmen s in he u ban a ea we e pa ks, whe e se e al pa hs wi h asphal su aces had been c ea ed and he sh ub laye was s ongly hinned. In he subu ban a ea allen ees we e emo ed. The e we e occasional, low-in ensi y o es y man- agemen ope a ions in he u al si e. Dis ance be ween he sampling a eas (u ban, subu ban, u al) was a leas 1 km, as p esc ibed by he gene al me hodology o he Globene p ojec (Niemela¨ e al., 2000). Fou si es, a leas 50 m om each o he (in o de o achie e independence, see Digweed, Cu ie, Ca´ camo, & Spence, 1995), we e selec ed wi hin each sampling a ea. Ca abids we e collec ed a each o he ou si es o he h ee sampling a eas 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 T. Magu a e al.4 UNCORRECTED PROOF using pi all aps o e wo ull ac i i y pe iods in 2001 and 2002. Ten aps we e placed andomly a leas 10 m apa a each si e. This esul ed in a o al o 120 aps sca e ed along he u ban– u al g adien (3 a ea 4 si es 10 aps). Each pi all ap was a leas 50 m om he nea es o es edge, in o de o a oid edge e ec s (Molna´ , Magu a, To´ hme´ e´sz, & Elek, 2001). Fu he in o ma ion on ap design, placemen , mode o ope a ion as well as he gene al desc ip ion o he collec ed assem- blages a e gi en by Magu a e al. (2004). Fo he p esen pape , we only used body size, collec ed om he li e a u e (Hu ˚ ka, 1996), and he numbe o indi iduals in he ca ch. Fo species whe e minimum and maximum sizes we e gi en, we used he mid- ange alue (see Table 1). To es o di e ences in he measu es desc ibing asymme y and/o inequali y o body size pa e n in ca abid assemblages among he h ee sampling a eas, epea ed measu es analyses o a iance (ANOVA) we e pe o med (Sokal & Rohl , 1995). When he ANOVA e ealed a signi ican di e ence be ween he means, LSD (leas signi ican di e - ence) es s we e pe o med o mul iple compa - isons among means. The analyses we e ca ied ou using he R package (R De elopmen Co e Team, 2004) and he SPSS-PC p og am. Resul s In bo h yea s, alues o skewness we e la ges in he u ban a eas and smalles in he subu ban ones indica ing ha mo e small indi iduals we e p esen in he u ban a eas han in ei he o he o he wo (Fig. 2A). The di e ences, howe e , we e no s a is ically signi ican (Table 2). Simila esul s we e ob ained wi h he obus measu e o skew- ness, he medcouple. The alues we e highes in he u ban a eas and lowes in he subu ban ones (Fig. 2B). These di e ences we e conside ed signi ican (Table 2). The Gini coe icien was highes in he u ban a eas, and dec eased owa ds he u al a eas, sugges ing ha body size inequali y o ca abid assemblages was la ges in he u ban pa ks, and dec eased along he u banisa ion g a- dien (Fig. 2C). He e he yea  ea men in e ac- ion was signi ican , bu nei he o he componen ac o s was (Table 2). The Lo enz coe icien s had alues S41 o all si ua ions, indica ing he impo ance o la ge indi iduals o he shape o he Lo enz cu e. In u ban a eas, he alue was e y close o S¼1 in bo h yea s. This is a cha ac e is ic o a nea ly symme ic Lo enz cu e. In bo h yea s, he u al a eas had he highes S alues, and he subu ban a eas had in e media e ones (Fig. 2D). The di e ences in he Lo enz asymme y coe icien s among he s udied a eas we e signi ican (Table 2). The alue o his coe icien was signi ican ly highe in he u al a eas compa ed o he subu ban and u ban a eas (di e ences be ween hese wo las a eas we e no s a is ically signi ican ). The e o e, he signi ican di e ence in he shape o he Lo enz cu es was caused p ima ily by a highe numbe o indi iduals wi h la ge body size in he u al a ea s. he o he wo a eas unde highe deg ee o u banisa ion. Discussion Analysing inequali y has a longe his o y in plan han animal s udies. To desc ibe inequali y in plan size, se e al s udies (C eed, Kain, & No on, 1998; Di ommaso & Wa son, 1997;Zammi & Zedle , 1993) used skewness de i ed om he s a is ical momen s o he size dis ibu ion. Se e al o he pape s used he Gini coe icien o measu e inequali y in plan size o biomass (Hanley & G o es, 2002;He, Ma, B own, & Lynch, 2005;Leiss &Mu¨lle -Scha¨ e , 2001;Rams ad & Hes ma k, 2001; Shumway & Koide, 1995;Wilson & Gu e i ch, 1995). Fo plan s, Damgaa d and Weine (2000) calcula ed he Lo enz asymme y coe icien o da a om Shumway and Koide (1995) o in e p e he e ec o myco hizae and plan densi y on he numbe o capsules p oduced by Abu ilon heo- ph as i (Fam. Mal aceae) indi iduals. They we e able o show ha he epo ed inequali y in he numbe o capsules when he plan s con ained myco hizae was caused by he inc eased impo - ance o indi iduals wi h high capsule p oduc ion (Damgaa d & Weine , 2000). This, howe e , emains he only example o using he p oposed index. Skewness is he only me hod used o e alua e inequali y and/o asymme y in size dis ibu ion o animal popula ions o assemblages (Gomez & Espadale , 2000;G ego y, 2000;Knou , 2004; Kozlowski & Gawelczyk, 2002;No o ny & Kindl- mann, 1996;Poulin & Mo and, 1997). In he p esen pape , we ex ended he ange o me hods applied o analyse he inequali y o animal body size dis ibu ion using wo o he pa ame e s ( he Gini coe icien and he Lo enz asymme y coe icien ) and showed ha he Lo enz asymme y coe icien , S, is a powe ul me hod o s udies desc ibing and in e p e ing a ia ions in body size. Published s udies in he in e na ional Globene p ojec cha ac e ised changes in he ca abid body size dis ibu ion along he u ban–subu ban– u al 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 Body size inequali y 5 UNCORRECTED PROOF 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 Table 1. A lis o he ca abid species, hei body size, and he numbe o indi iduals collec ed in he u ban, subu ban and u al sampling a eas du ing he wo s udy yea s nea Deb ecen, Eas e n Hunga y Species Body size (mm) U ban Subu ban Ru al Agonum lugens 9.0 2 0 0 Ama a an hobia 6.2 10 0 0 Ama a communis 6.5 9 0 0 Ama a consula is 8.3 0 1 2 Ama a con exio 7.7 113 26 47 Ama a amilia is 6.3 16 5 4 Ama a lucida 5.6 1 0 0 Ama a o a a 8.6 10 0 0 Ama a saphy ea 8.8 10 16 39 Ama a simila a 8.6 2 2 2 Anchomenus do salis 6.6 1 0 0 Anisodac ylus nemo i agus 8.9 52 0 0 Anisodac ylus signa us 11.8 1 1 0 Asaphidion la ipes 4.3 2 0 0 Badis e bulla us 5.2 11 1 0 Badis e lace osus 6.3 5 15 1 Badis e me idionalis 6.7 8 2 0 Bembidion lamp os 3.4 38 0 3 Cala hus e a us 9.5 2 0 1 Cala hus uscipes 11.1 26 0 3 Cala hus melanocephalus 7.1 1 0 0 Calosoma inquisi o 20.0 0 0 10 Ca abus con exus 17.0 1 107 124 Ca abus g anula us 19.0 6 1 6 Ca abus ull ichi 27.0 1 0 0 Ca abus iolaceus 28.0 75 78 237 Cli ina osso 5.9 3 0 0 Diach omus ge manus 8.4 1 0 0 Ha palus dis inguendus 9.5 0 1 0 Ha palus la us 9.1 14 1 24 Ha palus lu eico nis 7.1 5 20 1 Ha palus a dus 9.4 104 86 69 Ha palus xan hopus winkle i 7.1 21 10 3 Leis us e ugineus 6.8 0 0 1 Licinus dep essus 10.4 7 5 1 No iophilus bigu a us 4.9 2 0 0 No iophilus palus is 5.1 5 2 6 No iophilus u ipes 5.3 38 14 7 Ophonus ni idulus 9.6 1 2 42 Ophonus schaube ge ianus 8.8 0 0 1 Oxypselaphus obscu us 5.5 0 1 1 Panagaeus bipus ula us 7.2 7 5 0 Pla yde us u us 6.3 76 41 79 Poecilus cup eus 11.8 3 0 0 Poecilus e sicolo 10.5 1 0 0 Pseudoophonus g iseus 10.1 2 0 1 Pseudoophonus u ipes 13.1 10 26 19 P e os ichus an h acinus 11.2 5 3 0 P e os ichus mace 12.9 1 0 0 P e os ichus melana ius 15.7 58 3 1 P e os ichus melas 14.9 2 0 3 P e os ichus mino 7.6 0 1 0 P e os ichus nige 18.4 22 15 23 P e os ichus oblongopunc a us 11.5 117 454 1505 P e os ichus o oideus 7.1 1 0 0 P e os ichus s enuus 6.0 27 52 11 T. Magu a e al.6 UNCORRECTED PROOF g adien using ei he he dis ibu ion among di e - en , a bi a y size classes (Ala uikka, Ko ze, Ma - einen, & Niemela¨, 2002;Ishi ani e al., 2003)o he mean body size o he species weigh ed by hei espec i e abundance (Gaublomme, Dhuy e e , Ve dyck, & Desende , 2005;Magu a e al., 2004; Niemela¨ e al., 2002). In Finland, Ala uikka e al. (2002) in es iga ing he changes o ca abid body size ac oss an u banisa ion g adien concluded ha medium- o la ge-sized ca abid indi iduals we e mo e likely o be collec ed in he u al si es han in u ban o es agmen s. In Japan, he e a e no la ge and only ew medium-sized specialis species in he u ban en i onmen , while many specimens o medium-sized and some la ge-sized specialis spe- cies occu in he subu ban and u al si es (Ishi ani e al., 2003). Mean ca abid body size changed signi ican ly om small alues in he u ban a ea o la ge ones in bo h subu ban and u al a eas in Bulga ia (Niemela¨ e al., 2002), Hunga y (Magu a e al., 2004) and Belgium (Gaublomme e al., 2005). The e was a ma ginally signi ican change in he same di ec ion along he same g adien in Finland, bu none in Canada (Niemela¨ e al., 2002). Howe e , no only body size o he ca abid assemblages may change ac oss an u banisa ion g adien ; he e could be changes among di e en popula ions o he same species. The body size o Ca abus nemo alis dec eased signi ican ly om he u al su oundings o Hambu g, Ge many, owa ds he ci y cen e (Welle & Ganzho n, 2004). The p esen s udy, using a mo e sophis ica ed me hod ( he Lo enz asymme y coe icien ), no only p o ed he exis ence o a signi ican change in inequali y o ca abid body size ac oss he u ban–- subu ban– u al g adien , bu indica ed ha his di e ence was p ima ily due o an inc ease in he con ibu ion o indi iduals wi h la ge body size in he u al a ea. The mean body size o g ound bee les also inc eased (Magu a e al., 2004), bu 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 Table 1. (con inued ) Species Body size (mm) U ban Subu ban Ru al S omis pumica us 7.5 1 15 27 Synuchus i alis 7.2 7 6 146 T echus quad is ia us 3.5 0 1 1 Species sequence is alphabe ical. U ban Subu ban Ru al U ban Subu ban Ru al U ban Subu ban Ru al U ban Subu ban Ru al 0 1 2 3 (D) (C) (B) (A) Skewness -0.2 0.0 0.2 0.4 Robus skewness 2001 2002 0.0 0.1 0.2 0.3 Gini coe icien 0.0 0.5 1.0 1.5 Lo enz asymme y coe icien Figu e 2. A e age alues (7S.E.) o he skewness (A), he obus skewness (B), he Gini coe icien (C) and he Lo enz asymme y coe icien (D) o he u ban, subu ban and u al ca abid assemblages in he wo s udy yea s. Body size inequali y 7 UNCORRECTED PROOF his change can esul om a dec ease in he impo ance o small species, om he inc ease in medium-sized o la ge species, o a combina ion o hese. By e alua ing he mean body size, we canno dis inguish among hese possibili ies. The Lo enz asymme y coe icien allowed us o demons a e which o hese heo e ical possibili ies was espon- sible o he obse ed e ec . The la ge ca abid body size in he less dis u bed a ea ( u al a ea) and he smalle body size in he mode a ely o highly dis u bed a eas (subu ban and u ban a eas) could be explained by he hypo hesis pos ula ed by Szyszko (1983),G ay (1989) and Blake e al. (1994). Szyszko (1983), s udying he egene a ion o pine plan a ions a e clea -cu ing in Poland, sugges ed and used he mean indi idual biomass (MIB) index. This index is simply calcula ed as he a io o he o al esh body mass o he ca ch in a ap, di ided by he numbe o ca abid indi iduals caugh . Szyszko (1983) showed ha as egene a ion in he plan a ion p oceeds, he a e age alue o he MIB index also inc eases. Mean body size is posi i ely ela ed o body mass, and hus he conclusion is ha he mean body size in ca abid assemblages will also inc ease. G ay (1989) hypo hesised ha he mean body size o species should dec ease om undis u bed owa ds dis u bed habi a s. Ca abid assemblages o di e en ly managed g asslands ga e suppo o his hypo hesis (Blake e al., 1994). Highly dis u bed a eas suppo ca abid 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 Table 2. The esul s o epea ed-measu es ANOVA o he alues desc ibing asymme y and/o inequali y o body size pa e n in ca abid assemblages Sou ce SS d MS FpLSD es Skewness Tes s o wi hin-subjec s con as s Yea 0.001 1 0.001 0.008 0.931 Yea A ea 0.306 2 0.153 0.890 0.444 E o 1.547 9 0.172 Tes s o be ween-subjec s e ec s A ea 2.473 2 1.237 1.235 0.336 E o 9.011 9 1.001 Robus skewness Tes s o wi hin-subjec s con as s Yea 0.058 1 0.058 1.099 0.322 Yea A ea 0.020 2 0.010 0.192 0.828 E o 0.479 9 0.053 Tes s o be ween-subjec s e ec s A ea 0.424 2 0.212 3.868 0.061 U4S E o 0.493 9 0.055 Gini coe icien Tes s o wi hin-subjec s con as s Yea 0.0001 1 0.0001 0.142 0.715 Yea A ea 0.008 2 0.004 5.073 0.033 E o 0.007 9 0.0007 Tes s o be ween-subjec s e ec s A ea 0.014 2 0.007 1.974 0.195 E o 0.031 9 0.003 Lo enz asymme y coe icien Tes s o wi hin-subjec s con as s Yea 0.003 1 0.003 0.044 0.839 Yea A ea 0.058 2 0.029 0.469 0.640 E o 0.558 9 0.062 Tes s o be ween-subjec s e ec s A ea 0.932 2 0.466 7.315 0.013 U¼SoR E o 0.573 9 0.064 Yea ¼ he e ec o s udy yea (2001 and 2002), A ea ¼ he u ban, subu ban and u al sampling a eas. Resul s o he LSD es indica e which a ea(s) di e (s) signi ican ly (po0:05) om he o he s; o example U¼SoRindica es ha he measu ed alue was signi ican ly highe in he u al a ea han in he u ban and subu ban a ea ( hese wo a eas, howe e , we e no di e en ). T. Magu a e al.8 UNCORRECTED PROOF assemblages wi h species o smalle a e age body size han do less dis u bed si es (Blake e al., 1994; G andchamp, Niemela¨, & Ko ze, 2000;Holliday, 1991;Magu a, Elek, & To´ hme´ e´sz, 2002;Ribe a, Dole´dec, Downie, & Fos e , 2001;S ˇus ek, 1987). The causes o his can be mani old. Ca abids ha e g ound-li ing la ae ha a e weakly chi inised, limi ed in mobili y, and hus mo e sensi i e o changing condi ions han adul s (Lo¨ ei & Sunde - land, 1996). Dis u bance will equen ly c ea e un a ou able condi ions o g ound bee le adul s as well as la ae, when hei densi ies dec ease (Tho bek & Bilde, 2004) and species may become locally ex inc . Small-sized ca abid species may su e less mo ali y du ing such dis u bance e en s. Thei densi ies a e also usually highe han ha o la ge-sized species (Lu , 2002), so hey ha e a lowe p obabili y o local ex inc ion. Small species a e mo e o en winged han a e la ge-sized species (Ribe a e al., 2001). Consequen ly, small species a e mo e agile han la ge species and can colonise dis u bed and uns able a eas mo e easily (Thiele, 1977). Smalle species may also need ewe esou ces and/o may de elop as e han la ge species (Pe e s, 1983). In ca abids, la ge species ha e longe la al pe iods, making hem mo e ulne able o dis u bance e en s (Ko ze & O’Ha a, 2003). Small species can use he small ‘‘windows o sui abili y’’ o su i e in he dis u bed habi a . Lo¨ ei and Sunde land (1996) also call a en ion o he impo ance o la ae o explain ends in adul s. Along he s udied u banisa ion g adien , he deg ee o dis u bance is highe in he u ban (pa ed pa hs, hinned sh ub laye , in ensi e landscape managemen ) and in he subu ban a ea (manage- men o mode a e in ensi y, e.g. allen ees a e emo ed) han in he u al a ea ( a e occasions o in e en ion, low in ensi y managemen ). Dis u - bance caused by u banisa ion appea s o elimina e a ou able mic osi es o o es species wi h la ge body size and c ea e al e ed, ela i ely homoge- neous mic o-habi a s in aded by small-sized spe- cies capable o lying. All hese habi a al e a ions accompanied by u banisa ion con ibu ed o he obse ed a ia ion in ca abid body size ac oss he u ban–subu ban– u al g adien . Using he Lo enz asymme y coe icien (Dam- gaa d & Weine , 2000), we we e able o mo e comple ely analyse he size dis ibu ions o he g ound bee le assemblages along he u banisa ion g adien . This index has p o en o be mo e powe - ul han mo e adi ional me hods such as skew- ness, obus skewness (medcouple), o he Gini coe icien . The biological in e p e a ion o he index is no p oblema ic, and we sugges ha i is a use ul ool o u u e s udies o size/biomass dis ibu ion in animal assemblages. Acknowledgemen s The Globene s udy in Hunga y was suppo ed by he Ho oba´gy Na ional Pa k Di ec o a e. T. Magu a is a Bolyai Resea ch Fellow o he Hunga ian Academy o Sciences. We a e especially g a e ul o Ti ada Molna´ o ield and labo a o y assis- ance. We hank wo anonymous e iewe s o hei help ul commen s on he manusc ip . Re e ences Ala uikka, D. M., Ko ze, D. J., Ma einen, K., & Niemela¨, J. (2002). Ca abid and spide assemblages along an u ban o u al g adien in Sou he n Finland. Jou nal o Insec Conse a ion,6, 195–206. Blake, S., Fos e , G. N., Ey e, M. D., & Lu , M. L. (1994). E ec s o habi a ype and g assland managemen p ac ices on he body size dis ibu ion o ca abid bee les. Pedobiologia,38, 502–512. B ys, G., Hube , A., & S uy , A. (2004). A obus measu e o skewness. Jou nal o Compu a ional & G aphical S a is ics,13, 996–1017. C eed, J. C., Kain, J. M., & No on, T. A. (1998). An expe imen al e alua ion o densi y and plan size in wo la ge b own seaweeds. Jou nal o Phycology,34, 39–52. Damgaa d, C., & Weine , J. (2000). Desc ibing inequali y in plan size o ecundi y. Ecology,81, 1139–1142. Digweed, S. C., Cu ie, C. R., Ca´ camo, H. A., & Spence, J. R. (1995). Digging ou he digging-in e ec o pi all aps: in luences o deple ion and dis u bance on ca ches o g ound bee les (Coleop e a: Ca abidae). Pedobiologia,39, 561–576. Di ommaso, A., & Wa son, A. K. (1997). E ec o he ungal pa hogen, Colle o ichum coccodes,onAbu i- lon heoph as i heigh hie a chy de elopmen . Jou - nal o Applied Ecology,34, 518–529. Dixon, P. M., Weine , J., Mi chell-Olds, T., & Woodley, R. (1987). Boo s apping he Gini coe icien o inequal- i y. Ecology,6, 1548–1551. Gaublomme, E., Dhuy e e , H., Ve dyck, P., & Desende , K. (2005). E ec s o u banisa ion on ca abid bee les in old beech o es s. In: Lo¨ ei, G. L., & To , S. (Eds.), Eu opean ca abidology 2003. P oceedings o he 11 h Eu opean ca abidologis s’ mee ing, DIAS Repo , Vol. 114, pp. 111–123. Gini, C. (1912). Va iabili a e mu abili a. In E. Pizze i, & T. Sal emini (Eds.), Memo ie di me odologica s a is- ica (pp. 211–382). Rome, I aly: E edi Vi gilio Veschi. Glasse , G. J. (1962). Va iance o mulas o he mean di e ence and coe icien o concen a ion. Jou nal o he Ame ican S a is ical Associa ion,57, 648–654. 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 ARTICLE IN PRESS BAAE : 50051 Body size inequali y 9