장음표시 사용
91쪽
The ordination of the habitats on the basis of ant species composition was achi eved by car-rying out a correspondeiace analySis, the symmetric weighting option was chosen for the analysis GAUCA l 982). The NuCOS a TOTHMEREsZ l 993a) and SYN-TAX PODANI l 993) packages were u sed for computations. Diversi ty profites were employed sor scale-dependent diversi ty characterigation. while the Renyi's diversity index family was used because it is known to provide good resulis for communities of ait si Zes TOTHMERESZ l 995). Diversi ty profites were thera computed using the Di Vord package TOTHMERESZ l 993h). Using a modifled form of the niche position-breadth analysis MACNALLY 9 DooLAN l 986 GALLE et ah l 998) we estimated the values of two parameters, called position sa) and breadth sp)obtained hom multi variate data ora the ani composition for the various forests types. The range of assemblages Occupied by a species in the ordination factor space is called species 's breadth sp) and the distance of the species range centre Dom the origo, i .e. the cominia ity specifici ty of the species in question is the position sa).
The diversity ordering using the Renyi's parameter silowed that the ant assemblages of juniper forests are more diverse than the other habitats Fig. l). The
Tahle 2. Indices of species diversi ty for ant assem lages in seven forest habitats 24 sampled forests). The values represent the maximum and minimum Values in each case. Species diversi ty indices Forest type
92쪽
oak and Russian olive habitats could not be ordered according to their ant species diversity however, as the diversity profites intersected each other. This proves that the diversity indices we used order the assemblages different ly. The dominance diversity curves of the different forests are very Similar, with the exception os native poplar fore sis. On this basis we can assume that thestructure of the ani assem blages in the poplar foresis differs hom the other habi
resented J9. TR of the total number of individua is collected. Although we mund a bimodat distribution in the species numberis ite number histogram Fig. 3); but after we had grouped the occupied si tes in other ways this bimodali ty was lost. No correlation was mund het ween the locat densities and the number of occupied sites r m O. ld. p 0. l), therefore, this set os data donot support PIANS K s sl 982) core and satellite species theory. In accordance withthese findings, a continuous transition was observed belween frequent and rares pegies in the values of the produci os density. number of occupied siles and thereciprocat of the between-site aggregation of the distribution Table 3. Fig. 4). It
Fig. 1. Diversi ty ordering of the stud ted ant assemblages by the Renyi's diversity index function. Curves with higher values suggest higher diversi ty of the tested ant assemblages. Intersecting curves means that differences in diversity cannot be established. ForestS: Poplar, Oak, Juniper, Pine, Russian olive. Black locust, Hybrid poplar
94쪽
ANT ASSEMBLAGES ASSOCIATED UITH LO LAND FORESI S
Tahle 3. Information content of the regionat distribution for the studied ant species. The rate of aggregation was computed by the index os dispersion, I m variance mean. based On the number os occupied traps in ali habitats. c 'im Number of in dividuals 'reciprocat of the between-site agregationos the distribution
0. 06848.28 28 Myrmica Sabuleti
95쪽
the rate os aggregation are two sides of the information content of the distributionos the species in question whicli can he measti red with Shannon-Wiener function. Plotting this function against log densi ty of the species, a significant correlation is obtained r α 0. 39, p ς 0. 02); but the segregation belween common andrare species appearS to be very slight in this case too Fig. 5). The delailed ant composition of the forest is gi ven in Appendix l. Differen-ces in the ani composition was mund dependent on the forest types. Ants were numeri catly the most abundant in poplar habitais and the least in black locust habitat S. The native poplar and oak forest habitats could he characteri red with F. fusca whicli is the most typical forest ant species in the Great Hungarian Plain. Formica s. str. Species 'red wOOd antS F. rusci, F. truncorum, F. PolyCterim OC-cured also in the planted pine woods, as they be long to the original fauna os pine fore sis. Can Onotus ViagMS, whicli is a thermophilous and xerotolerant ant, is re- No of habitata
Fig. 3. Number of species plotted against the numher os occupied si tes
96쪽
Fig. 4. Rank sequence of species performed on the basis of their 'common ness . a produci of the locat density. number os occupied si tes and distribution os evennessstricted to the open forest sites se. g. , poplar, juniper). The aiat assemblages Ofjuni per woods consist of both thermophilotis forest and grassiand an is, o wing tot he open character of these habitats. In the introduced forests, including yotingpine plantations the dominant ant species are grassiand faunal elements L. S
Fig. S. The information content of species distribution H) as a stinction of the locat densi ty. Dolsabove the stashed line casa be regarded as 'commonV and those bolow the line as 'rare
97쪽
Higer, L. PSammOPhiluS, L. Viat Glieri HS, L. alteriuS, T. CGemitum, T. Gmbigilum etc.), whicli are poOr fractions os different grassiand ani communities. On the basis of the Mann-Whitney test ZAR l 996). performed to comparethe nuriaber of individuals in the different forests, few significant values remat nedas ter employing Bon ferron s correction. We found that belween native forestsi recently introduced ones, F. fuScia was significantly more abundant in the native
Fig. 6. Ordination diagram os the siles hased ora the composition of the ani assemblages using correspondence analysis. Filled circles: native forests. Fili ed triangles: introduceu foresis
98쪽
foresis and T. caemitum in the recent ly introduced ones. Bet ween nati velo id introduced fore sis: F. sanguinea is more abundant in the old introduced fore sis, and when we compared old introduced forest strecently introduced forests: L. ny n-deri was more abundant in the old introduced ones Table d). The correspondeiace analysis of the 24 different forests, based on their species composition Fig. 6), showed only a very Slight overi ap bet ween the assem hi ages of the native foresis and those of the introduced ones, suggesting different successionat pathways of these forests. The position of juni per wood is belween the two Other typCS. The analysis of communi ty- specifici ty i. e. the position sta) and breadth sp). of the thii ty si x ant species resulted in a hight y significant negative correlationbetween these two parameters r m -0. 68 p ς 0. 00 l. Fig. P). The two extremes in Fig. P are the speciali sis and the generali sis in the gi ven set of the communities. The speciali sis, like F. Polyclerici, F. PratenSiS and M. rubria, haVe a large position Value and are Ocupying more extreme types of communities. Whereas thegro up of generali StS aso L. Caemitum. L. uriis SciatuS, F. Cliniculiarici, Incl M. Sci- Dialeti have a large breadth value and are present in many kinds of assem blages.
Fig. T. The position and breadth plane. The location of each species in the plane is stlown 36 antspecies). A large position value indicates that an ant species is a communi ty specialist. whereas a large breadth value shows that a species is present in a wide range os communities
99쪽
Our furvey shows that a wide variety of ant assemblages occurred in seven forest types in the solitheria part of the Great Hungarian Plain. Similar studies have been conducted in sonae other East-European countries as weli. CZECHOW-
in Transsylvani an forests of Romani a. Our findings indicate that the number of species and diversi ty of ant assemblages in forests of the solitheria part of the Great Hungarian Plain are similar or higher than those in the neigh bouring re
Examining the species richness, we found that poplar forests showed the
As regards to eveni ess we found that Oak tree sites h ad the lowest values 0. 2J 0.ββ) and juni per tree siles the highest 0. 64-0. J9). Inirmation on the diversityos Hungarian forest anis is quite scarce, and preci se definition for the gradient of succession for the forests we examined was therefore not possibi e . CONNEI L l9 8). dealing with variations in diversi ty with in locat areas, tested the intermediate disturbance hypothesis. whicli suggesis that the highest diversity is maintained at intermediate scales of disturbance. He then applied the hypothesis totropical communities Such as ratia foresis and coral reess. PUNTILLA et ah sl 99 l).working on the coloni Zation of clearcut foresis in the solitheria Pinnisti talga found that during succession an increase of species richness was followed by a decrease probab ly because of increased stiade of the canopy cover. TOUYAM A l99J); studying Lu PhyllouS foresis in different de vel opmental stages insolith-western Japan, found that the myrmecolauna decreases in diversi ty as an effect of disturbance. In successionat sand clune areas of the Hungarian Plain diversi ty in creases in the mature stages of ant communi ty succession hut there is a decline at the final phase GALLE l 998). This is most likely a consequence of the hi gli densi ty of the red wood ant F. Polyclena) in Such areas. In Our SurVeyjuni per forests of the Kiskunsag Nationat Park showed the highest degree of antili Versi ty where anthropogenic disturbance and degradation were very low. The transitionat character of these sites could also assure the maintonance of iis diver-Sity. In comparison. ant diversity was the lowest in Russian olive foresis localedin the vicini ty of SZeged, where anthropogenic disturbance was qui te high and graZing waS eVident. Three basic leve is can usual ly be distinguished in the competition hierarchyos horeat ants VEPSALAINEN & PISARS KI l 982). The group of highesi levet competitiors. the territorials in Central and Northeria Europe, includes ants, whichali be long to the Formicinae subfami ly the subgenera Formica S. Str. L. and
100쪽
Ptoformica MULL. Formica Servi formica) cinerea MAYR., L. fuliginoSMS)J. These are relative ly abundant species in foresis and forest boundaries. As a
havi oural competiti ve) dominance. The effecis of competitive interactions maybe difficult to infer and this makes more difficult to define possibie ant hierarchies in the investigated habitats. Examining the order of diversi ty for the foresis in question we mund that L. Pla thorax appeared to he the more abundant species in the juni per forest localed in Bugac. SEI FERT l99 l) defines L. Pla thoriax as a sibi ing species of L. niger. Occurring in forests. bogs and fens. Much higher L. Platythortia densitiesare normalty seen in areas with less extreme moisture. Although the Boc saforests were fati ly iso lated and not show any sigia os degradation, L. Plat moriaXwaS n Ot present there. T. iambiguum and T. Cesemitum had Very similar abundances there. The absence of aggressive species is evident and L. Pla thoraX was numeri catly dominant hecati se of this. According to SAVOLAINEN and VEPSALAINEN l 988), when territorial species are absent; lower- levet species are sit inedhigher up in the hierarchy. Abundance of dominant anis can reduce the fitness of
forests as well. Territorial expansion os superior competitors drive weaker species to extinction or compei them to change their nesting sites. But in communities lacking territorial species the hierarchical relations are much less pre
In black locust forests L. nylancteri, a typical forest ant, be longing to the native aiat communi ty, is one of the most dense species. Original native forest mOStly OakS were replaced by black locust and sonae remnants of the originalfauna could survive there. In spite of their low species number, black locust forests are loca ted second in the diversi ty order. this could be explained by the faci that they are en viron mental ly stressing habitatS e. g. no aphidS extrememicroclimate), keeping the density of the populations evenly low. Pine forests showed resulis consistent with those Dona SAVOLAINEN and VEPSALAINEN l 989). insolar that F. Sanguinea and F. rusci are at the top of the