Tuexenia 31: 301–315. Göttingen 2011.
Vegetation structure, ecological stability, and low-disturbance regime
of abandoned dry grasslands support specific ant assemblages
in Central Slovakia
– Michal Wiezik, Adela Wieziková and Marek Svitok –
Compared to other grassland types across Slovakia, dry grasslands harbour species-rich and
specialised ant communities. High diversity and species richness of ants may be seen as a consequence
of (i) the specific structure of dry grassland vegetation, (ii) long-term ecological stability, and (iii) cur-
rently low or absent management-induced disturbance. With special regard to dry grasslands, we report
on structural characteristics of vegetation and low-disturbance regime, which contribute to ant assem-
blage structure. Our study was carried out in the Štiavnické Vrchy Mts. (Central Slovakia), a region
with a historically well-developed grassland area. We established a set of 25 research plots within south-
orientated grassland habitats representing five different grassland types: wet managed and wet aban-
doned grasslands, mesic managed and mesic abandoned grasslands (Arrhenatherion elatioris), and dry
abandoned grassland habitats (Asplenio-Festucion glaucae). Each habitat type was represented by five
plots. At each plot, a set of ten pitfall traps was used to sample ground-foraging ant assemblages.
Around each trap, structural characteristics of vegetation and microhabitat were assessed. Dry grasslands
were shown to have a specific microhabitat structure, characterised by the presence of a well-developed
moss and lichen layer, exposed bedrock, bare soil, and significantly lower, although species-rich vege -
tation. Besides the specific microhabitat structure, the absence of management may have contributed to
the distinctiveness of these ant assemblages compared to those associated with other grassland categories.
Ant assemblages were more species-rich, and the activity of ants was higher in recently abandoned
grassland habitats. The effect of abandonment was quite opposite for plants, whose species-richness
was, contrary to ants, higher within managed sites.
Zusammenfassung: Vegetationsstruktur, ökologische Stabilität und Störungsregime
geringer Intensität im aufgelassenen Trockengrasland
begünstigen spezifische Ameisengesellschaften in der Mittelslowakei
Im Vergleich mit anderen Graslandtypen der Slowakei beherbergt Trockengrasland artenreiche und
spezialisierte Ameisengesellschaften. Die hohe Diversität sowie der Artenreichtum an Ameisen können
als eine Folge (i) der spezifischen Struktur der Trockengrasland-Vegetation, (ii) langfristiger ökologi-
scher Stabilität und (iii) aktueller geringer oder ganz fehlender Störungsintensität durch Bewirtschaf-
tung angesehen werden. Mit besonderer Berücksichtigung des Trockengraslandes berichten wir über
strukturelle Charakteristika der Vegetation und Störungsregime niedriger Intensität, welche zur Struk-
tur der Ameisengesellschaften beitragen. Unsere Studie wurde in den Schemnitzer Bergen (Štiavnické
Vrchy, Mittelslowakei) durchgeführt, einer Region mit historisch gut entwickelten Graslandflächen. Ein
Set von 25 Untersuchungsflächen (Plots) wurde in südexponierten Graslandhabitaten eingerichtet,
welche fünf unterschiedliche Graslandtypen repräsentieren: Bewirtschaftete und aufgelassene Feucht-
wiesen, bewirtschaftetes und aufgelassenes mesisches Grasland (Arrhenatherion elatioris) sowie aufge-
lassenes Trockengrasland (Asplenio-Festucion glaucae). Jeder Habitattyp ist durch fünf Plots repräsen-
tiert. Auf jedem Plot wurde ein Set von zehn Barber-Fallen (Bodenfallen) benutzt, um am Boden jagen-
de Ameisengesellschaften zu beproben. Rund um jede Falle wurden strukturelle Vegetations- und
Mikrohabitat-Charakteristika erfasst. Es wurde gezeigt, dass Trockengrasland eine spezifische Mikro -
habitatstruktur besitzt, die durch das reichliche Vorhandensein einer Moos- und Flechtenschicht, anste-
hendes Gestein, offenen Boden und eine signifikant niedrigere wiewohl artenreiche Vegetation charak-
terisiert ist. Neben der spezifischen Mikrohabitatstruktur trägt wohl das Fehlen von Bewirtschaftung
zur Besonderheit dieser Ameisengesellschaften im Vergleich mit denjenigen der anderen Graslandkate-
gorien bei. Jüngst aufgelassene Graslandhabitate wiesen artenreichere Ameisengesellschaften und eine
höhere Aktivität der Ameisen auf. Der Effekt der Stilllegung war recht gegensätzlich für Pflanzenarten,
welche, im Gegensatz zu Ameisenarten, mehr Arten in bewirtschafteten Flächen aufwiesen.
Keywords: Carpathians, Formicidae, habitat structure, management, Štiavnické Vrchy Mts.
Dry grasslands represent a characteristic yet extremely threatened and declining feature
of the sub-mountain landscape in Slovakia (DAVID et al. 2007). Having been integrated into
agricultural landscapes of Central Europe for centuries, dry grasslands have experienced a
strong decline and loss of connectivity during the last hundred years (VAN DIJK 1991). This
specific biotope type is considered a remnant of the late Pleistocene steppe formation
(BREDENKAMP et al. 2002) or a land form resulting from long-term agricultural activities in
submontane areas (ELLENBERG 1996). Besides direct devastation including urbanization,
establishment of stone quarries, conversion to farmland and encroachment of invasive
woody species, the change of traditional use and abandonment are among the main risks for
both dry grassland habitats and associated species (WILLEMS 2001, LUOTO et al. 2003,
DEKONINCK et al. 2007, RŪSIŅA & KIEHL 2010).
The abandonment of dry grassland may not cause rapid successional changes and over-
growth similar to the rate recorded from mesic or wet grasslands (ÖCKINGER et al. 2006);
however, the lack of traditional use usually contributes to the loss of open patchy vegetation
patterns. Consequently, although the general open nature of a dry grassland habitat may
remain relatively stable, the building of a dense and compact vegetation layer may have an
adverse effect on the majority of specialised xerophilic species (DEKONINCK et al. 2007).
Traditional or sustaining management is hence being recognised as vital for conservation and
restoration of dry grasslands (BOBBINK & WILLEMS 1993).
Generally, dry grasslands exhibit extreme habitat conditions, mainly due to relatively
dry and nutrient-poor soils. Nevertheless, communities of plants and animals associated
with these habitats are among the most species-rich, harbouring dozens of specialised,
stenotopic and endemic species (DAVID et al. 2007).
Ants (Hymenoptera: Formicidae) are classified as ecosystem engineers (JONES et al. 1994,
FOLGARAIT 1998). Considering their important role in soil engineering, distribution of plant
seeds, and predation (FOLGARAIT 1998), investigations on ant communities are essential for
the assessment of biodiversity change in the cultural landscapes of Europe. Dry grasslands in
Slovakia support a relatively diverse ant fauna with a number of unique species (WIEZIK
2008). However, relatively little is known about the factors influencing dry grassland ant
communities. Vegetation structure is recognised as an important factor affecting ant popula-
tions in temperate habitats (RETANA & CERDÁ 2000), mainly due to the regulation of micro-
climate conditions, which influence, in the form of environmental stress, ant activity and ant
community composition (ANDERSEN 2000). Furthermore, vegetation can directly and indi-
rectly affect availability and quality of food and nesting sites, as well as interaction between
particular ant species (ANDERSEN 2000, RETANA & CERDÁ 2000, LASSAU & HOCHULI 2004).
The aim of this study was to analyse, with special regard to dry grasslands, the variation
in habitat structure along a humidity and management gradient within grassland habitats in
the Štiavnické Vrchy Mts., Central Slovakia, and to identify how these structural and
management patterns are reflected in the structure of associated ant assemblages.
2. Study area
The study was conducted in the Štiavnické Vrchy Mts. in the western Carpathians,
Slovakia. This region has a mean annual precipitation of 895 mm and a mean annual temper-
ature of 7.6 ºC (HLAVAČEK 1985). The soils of the area are of volcanic origin; rich brown
soils overlaying andesite bedrock are most frequent (KONEČNÝ et al. 1998). The region has
been used intensively for mining of metals since the 13th century. Human settlement, min-
ing, and associated intensive timber production left a large portion of this area deforested,
paralleling a general decline of low elevation forests in Slovakia (KORPEĽ 1989) and else-
where in Europe (GILG 2005). The proportion of grassland habitats has hence been artificial-
ly increased and today represents about 30% of the region’s total area (KUNCA et al 2005).
The study was carried out on a set of managed and abandoned grassland habitats in the
Štiavnické Vrchy Mts. (Fig. 1). We selected 25 grasslands at altitudes of 260–655 m a.s.l.
representing grassland categories typically occurring across the study area: wet abandoned
(WA), wet managed (WM), mesic abandoned (MA) and mesic managed (MM) grasslands
(Arrhenatherion elatioris Koch 1926, class Molinio-Arrhenatheretea), and dry abandoned
(DA) grasslands (Asplenio-Festucion glaucae Zolyómi 1936, class Festuco-Brometea). Each
type was represented by five individual plots. Except humid grasslands located in inundation
zones with no apparent sloping, all plots were selected on SE, S, or SW facing slopes. Dry
managed grasslands were not included in the study, as this particular grassland type is cur-
rently absent in the studied region (KUNCA et al. 2005). Wet and mesic abandoned grassland
represented sites that have been abandoned during the last 15 years. A well-preserved open
feature and no or only weak woody species overgrowth (not exceeding 10% of the total
grassland area) were typical for all abandoned sites including dry grasslands that have been
abandoned for at least 30 years. All managed sites were mown annually in early summer;
within the majority of wet managed sites, cattle grazing was conducted during late summer.
Fig. 1: Location of research plots. Grassland categories: I wet abandoned, i wet managed, J mesic
abandoned, j mesic managed, G dry abandoned.
Abb. 1: Lokalisierung der Untersuchungsflächen (Plots). Graslandkategorien: I feucht, aufgelassen;
i feucht, bewirtschaftet; J mesisch, aufgelassen; j mesisch, bewirtschaftet; G trocken, aufgelassen.
3.1. Ant sampling
In each plot, a line of 10 traps, spaced at 3 m intervals, was established. The traps consisted of plastic
cups (diameter 3 cm) buried at soil level and filled with about 50 ml of a 4% solution of formaldehyde.
Each trap was left in place for 7 days before being opened in order to reduce digging-in effects
(GREENSLADE 1973). In total, 250 traps were placed in the field during late May 2008, and each trap
remained open for 7 days. Sampling was repeated in late September 2008. The number of individuals
trapped per plot or grassland category depends on the activity of ants, which is not directly related to
ant abundance; however, ant activity is a widely accepted measure in comparative analyses (BESTELMEYER
et al. 2000, GOMÉZ et al. 2003, OTTONETTI et al. 2006). Ants were identified at the species level accord-
ing to SEIFERT (2007); only workers were considered in the statistical analyses.
3.2. Vegetation and microhabitat measurements
The following vegetation and microhabitat variables were measured: total number of vascular plant
species, maximum vegetation height, mean vegetation height (based on five randomly placed measure-
ments), total vegetation coverage (including the coverage of vascular plants, bryophytes, and lichens),
coverage of particular vegetation layers (e.g. woody species, grasses, herbs, bryophytes + lichens (grow-
ing on soil)), proportion of exposed bedrock and plain soil. All variables were measured in May 2008
before the mowing season, on a plot of 1 m × 1 m placed around each trap used for ant sampling. These
microhabitat variables have been used in previous studies assessing the influence of vegetation on ants
(DAHMS et al. 2005), as they reflect the complexity of vegetation, which is considered a significant stress-
related factor affecting ant communities (ANDERSEN 2000).
3.3. Statistical analyses
Multiple regression analysis was used to examine the effect of vegetation variables (total species rich-
ness, mean vegetation height, total grass cover, total herb cover) on ant species richness. Best-subset
regression based on Mallow’s Cp criterion was employed to build the most parsimonious models (STAT-
SOFT 2004). The primary aim of our study was to separate the effects of humidity and management on
ant assemblages; however, because of immanent correlations in explanatory variables, we fitted separate
linear models for plant characteristics and grassland categories. The separate fitting of models allows for
quantifying the individual and joint effects of humidity and management as the dominant forces
hypothesised to shape ant diversity. Prior to the analyses, seasonal data were pooled over traps.
Assumptions of normality and homoscedasticity were carefully checked using diagnostic plots of resid-
uals. Variables were log transformed when necessary.
Differences in ant assemblages and vegetation characteristics between grassland categories were
analysed according to the hierarchical ANOVA design. Mixed models, which included the fixed effect of
grassland category (for categories cf. section 2) and the random effects of plot, nested within grassland
category, were fitted to the data. A nested ANOVA allows partitioning of residual variance of the pooled
model into the variance explained by the differences among particular plots and pure residual variance.
In order to separate the effects of habitat humidity and management, a hierarchical factorial
ANOVA was employed. A mixed model, which included the fixed effects of management (managed,
abandoned), habitat humidity (wet, mesic), and their interaction and the random effects of plot, nested
within interaction, was fitted to the ant richness and activity and plant characteristics, respectively.
Because of the absence of dry managed grasslands in the studied region, abandoned dry grasslands, and
hence a dryness factor, could not be subjected to this mixed model analysis. For significant effects, rele-
vant pairwise comparisons were made using Tukey’s HSD tests. Assumptions of each model were
screened in diagnostic plots, and data were log transformed when necessary. However, untransformed
data are displayed in figures to facilitate interpretation.
A list of ant species and their activity was compiled for each trap to get a community data matrix.
The same linear models were fitted to the ant assemblage data matrix using distance-based MANOVA
(ANDERSON 2001). These multivariate analyses were based on Bray-Curtis distance (BRAY & CURTIS
1957). Species activities were log transformed to reduce asymmetrical influence of the dominant species.
P values were based on 9999 permutations of raw data using correct permutable units (ANDERSON &
TER BRAAK 2003) in DISTLM package (ANDERSON 2004). Non-metric multidimensional scaling (NMDS;
KRUSKAL 1964) was used to visualise differences in assemblage composition. Relationships between
species and NMDS configuration were examined by Spearman rank correlation.
4.1. Microhabitat structure
The field assessment of microhabitat structure, based on measurements of selected vege-
tation and terrain surface characteristics, showed several unique structural features of dry
grassland habitats. Structural patterns like exposed bedrock, plain soil, and a well-developed
layer of mosses and lichens (Table 1) were recorded exclusively from dry grasslands.
Besides these patterns, marked differences in vegetation height and plant species richness
between the grassland habitats were revealed by the hierarchical ANOVA (Table 2). Dry
grasslands were characterised by the lowest mean vegetation height, being only about one
third to one half of the vegetation height recorded from other grassland categories (Table 1).
All grassland categories, except mesic managed, had significantly lower mean plant species
richness compared to dry grasslands. In general, the lowest plant species richness was
recorded from wet abandoned grasslands. Within all grassland categories (except wet aban-
doned), a moderate dominance of grass cover was typical; woody species occurred almost
exclusively in mesic abandoned grasslands, where their overall coverage exceeded 10%.
We also observed a significant effect of management and its absence reflected in plant
species richness: When comparing managed and abandoned grasslands within the same
humidity category, the abandoned sites were significantly poorer in recorded plant species
Total vegetation [%]
Herb species [%]
Grass species [%]
Woody species [%]
Bryophytes + Lichens [%]
Bare soil [%]
Plant species richness
Max. vegetation height [cm]
Mean vegetation height [cm]
Table 1: Selected microhabitat characteristics of studied grassland categories.
Mean values per 1 m2± SE are displayed.
Tab. 1: Ausgewählte Mikrohabitat-Charakteristika der untersuchten Graslandkategorien.
Es sind Mittelwerte pro 1 m² ± Standardfehler angegeben.
Total herb cover Total grass cover
Source of variance DF
4 4368 2.62ns 1.37ns 6.59**
20 1669 6.94*** 2061 7.29*** 64 13.93*** 1068 7.63***
225 240 283 5 140
Table 2: Results of hierarchical ANOVA on vegetation characteristics.
Degrees of freedom (DF), mean sum of squares (MS), and F statistics with accompanying probabilities
(ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) are displayed for each source of variance. Plot was a nested
random factor within management.
Tab. 2: Ergebnisse der hierarchischen Varianzanalyse (ANOVA) der Vegetations-Charakteristika.
Freiheitsgrade (DF), Mittlere Summe der Quadrate (MS), und F-Statistik mit zugehörigen Signifikanz-
werten (ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) sind für jede Quelle der Varianz angezeigt. Plot war
ein genesteter, zufälliger Faktor innerhalb der Landnutzung.
4.2. Variation in ant assemblages
We collected a total of 6689 workers belonging to 35 different ant species from the 25
research plots. The total number of recorded ant species was highest at dry grasslands and
decreased with increasing habitat humidity, being the lowest at wet managed grasslands
(Table 3). The higher ant species richness of dry grasslands was partially attributed to a
number of species recorded exclusively from this habitat (Ponera testacea, Messor structor,
Plagiolepis spp., Camponotus spp.); however, those differences were not statistically signifi-
cant (Table 4). On the other hand, significant differences were found for total epigaeic activi-
ty. The four most common ant species (Myrmica scabrinodis, M. rubra, Formica pratensis,
and Lasius niger) accounting for 84.9% of the total number of sampled workers were
recorded from wet and mesic grasslands only.
Fig. 2: Significant linear relationship between mean vegetation height and ant species richness (F(1,23) =
5.28; p < 0.05; R2= 0.19). The symbols correspond to the grassland categories (see Table 1).
Abb. 2: Signifikanter linearer Zusammenhang zwischen mittlerer Vegetationshöhe und Artenreichtum
der Ameisen (F(1,23) = 5.28; p < 0.05; R2= 0.19). Die Symbole entsprechen den Graslandkategorien
(siehe Tab. 1).
The assessment of relations between ants and vegetation structure showed a significant
correlation between mean vegetation height and ant species richness (F(1,23)= 5.28; p < 0.05).
Across all grassland sites, the ant species richness was negatively related to the vegetation
height (Fig. 2); this relation, however, was relatively weak, as it explained only about 19%
of the total variance.
Composition of ant assemblages differed significantly among grassland categories (Table
4). Sites from dry grasslands and the remaining grassland categories tended to occupy
different sections of the ordination space, suggesting that the composition of ant assem-
blages differed distinctly between these two groups (Fig. 3). The observed differences in ant
assemblage composition among grassland categories corresponded to changes in propor-
tions of several ant species (Fig. 3).
The dry grasslands were characterised by an increased proportion of Tetramorium
caespitum and Tapinoma ambiguum and by the exclusive occurrence of Tetramorium
moravicum. The remaining grassland categories, except mesic abandoned grasslands,
grouped together to build a centroid characterised by affinity of several common eurytopic
ant species. Mesic abandoned grasslands built a separate centroid, distinguished by an
Table 3: Total epigaeic activity of ants (number of workers trapped per grassland category and 50 traps
during 2 × 7 days of trap exposure) and presence of species at replicated sites (superscript index repre-
senting the number of plots).
ZGe – Zoogeographical element (according to CZECHOWSKI et al. 2002): AP – Amphipalaearctic, CE –
Central-European, E – European, EC – Euro-Caucasian, ES – Euro-Siberian, MD – Mediterranean,
NP – North-Palaearctic, SE – South-European, SP – South-Palaearctic; Ee – Ecological element: e –
eurytopic, p – polytopic, o – oligotopic, s – stenotopic; land use: m – managed, a – abandoned.
Tab. 3: Absolute epigäische Aktivität von Ameisen (Anzahl der gefangenen Arbeiter pro Graslandkate-
gorie und 50 Fallen während 2 × 7 Tagen Fallenexposition).
ZGe – Zoo-Geographisches Element (nach Czechowski et al. 2002): AP – amphipaläarktisch, CE –
mitteleuropäisch, E – europäisch, EC – euro-kaukasisch, ES – euro-sibirisch, MD – mediterran, NP –
nord-paläarktisch, SE – süd-europäisch, SP – süd-paläarktisch; Ee – Ökologisches Element: e – eury-
top, p – polytop, o – oligotop, s – stenotop; Nutzungstyp: m – bewirtschaftet, a – aufgelassen.
Wet Mesic Dry
m a m a a
Ponera testacea (Emery, 1895) MD s - - - - 2 2
Myrmica gallienii Bondroit, 1920
M. lobicornis Nylander, 1846
M. rubra (Linnaeus, 1758)
M. rugulosa Nylander, 1849
M. sabuleti Meinert, 1861
M. scabrinodis Nylander, 1846
M. schencki Viereck, 1903
Messor structor (Latreille, 1798)
Myrmecina graminicola (Latreille, 1802)
Solenopsis fugax (Latreille, 1798)
Leptothorax acervorum (Fabricius, 1793)
Temnothorax interruptus (Schenck, 1852)
Tetramorium caespitum (Linnaeus, 1758)
T. ferox Ruzsky, 1903
T. impurum (Förster, 1850)
T. moravicum Kratochvíl, 1941
Tapinoma ambiguum Emery, 1925
T. erraticum (Latreille, 1798)
Plagiolepis pygmaea (Latreille, 1798)
P. vindobonensis Lomnicki, 1925
Lasius flavus (Fabricius, 1782)
L. jensi Seifert, 1982
L. alienus (Förster, 1850)
L. niger (Linnaeus, 1758)
L. psammophilus Seifert, 1992
Camponotus vagus (Scopoli, 1763)
C. piceus (Leach, 1825)
C. aethiops Latreille, 1798
Formica cunicularia Latreille, 1798
F. gagates Latreille, 1798
F. pratensis Retzius, 1783
F. rufa Linnaeus, 1761
F. rufibarbis Fabricius, 1793
F. sanguinea Latreille, 1798
Species richness 12 13 15 15 23
increased proportion of Myrmica scabrinodis. These results further corroborate our obser-
vation that ant assemblages of dry grasslands differ distinctly from other grassland cate-
gories, bearing features of a separate and specific ant community.
4.3. Impact of humidity and management
We tried to separate the effects of humidity and management on ants and plants through
the hierarchical factorial ANOVA models (Table 5). We found a significant effect of manage-
ment on total species richness of ants in wet and mesic grassland habitats (Fig. 4). The number
of ant species collected per trap in recently abandoned grassland sites was significantly higher
than in annually managed grasslands. This trend occurred along the grassland categories
regardless of habitat humidity. The trend was opposite, although not significant, in plants,
where the number of plant species recorded per square meter was higher at managed sites
(Fig. 4). Contrary to the situation for ants, plant species richness was significantly related to
grassland humidity: Wet grasslands contained ~40% less plant species compared to mesic sites.
Table 4: Results of hierarchical ANOVA and MANOVA on ant assemblage characteristics.
Degrees of freedom (DF), mean sum of squares (MS), and F/pseudo-F statistics with accompanying
probabilities (ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) are displayed for each source of variance. Plot
was a nested random factor within management.
Tab. 4: Ergebnisse der hierarchischen ANOVA und MANOVA der Charakteristika der Ameisengesell-
Freiheitsgrade (DF), Mittlere Summe der Quadrate (MS), und F/pseudo-F-Statistiken mit zugehörigen
Signifikanzwerten (ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) sind für jede Quelle der Varianz angezeigt.
Plot war ein genesteter, zufälliger Faktor innerhalb der Landnutzung.
Ant species richness Total epigaeic activity Assemblage composition
Source of variance DF
Management 4 0.287 2.28ns
Plot (Management) 20 0.126 4.8***
Trap 225 0.026
Fig. 3: NMDS ordination plot of Bray-Curtis dissimilarities in ant assemblages (left). Spearman correla-
tion coefficients of species activities with ordination axes (right). Only species with higher correlation
(> 0.4) are displayed. The symbols correspond to the grassland categories (see Table 1).
Abb. 3: NMDS Ordinationsdiagramm der Bray-Curtis Unähnlichkeiten in Ameisengesellschaften
(links). Spearman Korrelationskoeffizienten der Aktivität der Arten mit den Ordinationsachsen
(rechts). Nur Arten mit höherer Korrelation (> 0.4) sind gezeigt. Die Symbole entsprechen den
Graslandkategorien (siehe Tab. 1).
Table 5: Results of hierarchical factorial ANOVA on ant and plant species richness
Degrees of freedom (DF), mean sum of squares (MS), and F statistics with accompanying probabilities
(ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) are displayed for each source of variance. Plot was a nested
random factor within Humidity Management interaction.
Tab. 5: Ergebnisse der hierarchischen faktoriellen ANOVA des Artenreichtums der Pflanzen und Ameisen
Freiheitsgrade (DF), Mittlere Summe der Quadrate (MS), und F-Statistik mit zugehörigen Signifikanz-
werten (ns> 0.05; * < 0.05; ** < 0.01; *** < 0.001) sind für jede Quelle der Varianz angezeigt. Plot war
ein genesteter, zufälliger Faktor innerhalb der Interaktion Feuchtigkeit × Landnutzung.
Ant species richness
Plant species richness
Source of variance DF
Humidity × Management
Plot (Humidity × Management)
Fig. 4: Comparison of ant species richness (F(1,16) = 8.87; p < 0.01) and plant species richness (n.s.)
across grassland management types in the Štiavnické Vrchy Mts.
Abb. 4: Vergleich des Artenreichtums der Ameisen (F(1,16) = 8.87; p < 0.01) und Pflanzen (n. s.) über
alle Landnutzungstypen des Graslands in den Schemnitzer Bergen (Štiavnické Vrchy).
Our results generally suggest that dry grasslands represent a specific habitat type. Com-
pared to wet and mesic grasslands of the alliance Arrhenatherion elatioris (class Molinio-
Arrhenatheretea), dry grasslands of the alliance Asplenio-Festucion glaucae (class Festuco-
Brometea) were distinct in vegetation structure and consequently in associated ant assem-
blages. The habitat structure was characterised by species-rich, structurally divergent yet
relatively short vegetation and a high proportion of patches without plant cover, as opposed
to the dense and tall vegetation of wet and mesic grasslands. As shown in our study, the
composition of ant assemblages from dry grasslands diverged strongly from assemblages
from wet and mesic grassland habitats. Assemblages from dry grasslands were rich in species
and hosted a number of exclusive Mediterranean zoogeographical elements (Table 3).
Several characteristics of dry grasslands may have contributed to unique ant assemblages,
namely habitat structure, ecosystem stability, and low management-related disturbance.
Even though we tend to interpret the effects of these characteristics separately, they in fact
represent a group of interdependent patterns and processes jointly influencing the associated
ant assemblages. As we did not assess any independent gradients linked to these characteris-
tics in the present study, it is not possible to divide the impacts among them. In the follow-
ing chapters we discuss these characteristics in detail.
5.1. Effect of habitat structure
As shown in our study, ant richness was negatively related to vegetation height. Never-
theless, only 19% of the variation in ant species richness was explained by vegetation height;
hence a major portion of the variation remained unexplained. Several habitat traits, such as
exposed bedrock and bare soil, were restricted to the dry grasslands and hence, as an exclu-
sive structural pattern, were not considered in the statistical analyses. In the environment of
Central Europe, such features are believed to determine the occurrence of the majority of
the ant species recorded from dry grassland habitats (see SEIFERT 2007).
In the majority of invertebrate groups, high complexity habitats support high biodiversity.
A more complex environment increases available microhabitats, food, and shelter, and thus
increases the diversity of associated species (NIEMELÄ 1997, SIMILÄ et al. 2003, LASSAU et al.
2005). In ants, the situation may be different as they tend to show higher affinity for habitats
with less complex vegetation structure (LASSAU & HOCHULI 2004).
From an ant’s perspective, the main advantage of lower vegetation complexity is that it
contributes to a low-stress environment (ANDERSEN 2000). Environmental stress is generally
defined as any factor limiting productivity (GRIME 1979). The primary stressors for ants are,
among others, low temperature and microhabitat structure. Low temperature is considered
to be the primary stressor controlling global patterns of ant productivity and community
structure (ANDERSEN 1995); together, climate and habitat structure determine the degree of
surface insolation (ANDERSEN 2000). Densely covered habitats with structurally complex,
high vegetation forming shaded and cool microclimate are generally poor in ants; on the
other hand, warm and open habitats host the majority of ant species (ANDERSEN 1995). The
structural complexity of the foraging surface exerts a major influence on the ability of ants
to capture food resources. Dense or tall vegetation, for instance, dramatically reduces the
efficiency of locating, retrieving and defending resources (ANDERSEN 2000).
Besides the overall affinity of ants to less shaded habitats, the patchy vegetation structure
and presence of bare substrates within dry grasslands (Fig. 5) play an important role in
structuring ant communities by widening the thermal range of the habitat. In general, vege-
tation alters the habitat microclimate in terms of reducing the diurnal and annual tempera-
ture and humidity variation. Areas that are exposed to wind and direct sunlight are drier and
warmer than those that are protected by canopy cover. Thus, variations in vegetation struc-
ture may either enhance or reduce ant diversity by increasing or reducing habitat hetero-
geneity, microclimate suitability, or the activity of dominant species (RETANA & CERDÁ
2000). The combination of vegetation and bare substrate patches within dry grasslands creates
a mosaic of sunny and shaded microhabitats. A wider thermal range may increase species
diversity, because different species are favoured by different thermal conditions. Gradual
changes in the environment alter competition between species (ANDERSEN 2000). In particu-
lar, behaviourally dominant ant species are heat intolerant, and their external activity is lim-
ited by high temperatures. In areas with high vegetation cover, dominants benefit from
lower temperatures by extending their periods of activity. This results in a decrease in the
abundance of subordinate species (RETANA & CERDÁ 2000). Dominant species reach the
maximum foraging activity at 20–30 °C and lower their activity distinctly at higher tempera-
tures (RETANA & CERDÁ 2000). This allows the subordinate ant species to occupy the habi-
tat during the high-temperature period of a day without taking a risk of encountering the
behaviourally dominant ant species. In open habitats, ant communities are primarily ordered
by thermal variations and hence show few effects of interactions. On the other hand, more
complex habitats, where thermal variation is limited by lush vegetation, usually favour dom-
inant ant species (ANDERSEN 2000, RETANA & CERDÁ 2000).
This temperature-driven niche partitioning in ants has been well documented from
Mediterranean and steppe regions (HEATWOLE & MUIR 1989, RETANA & CERDÁ 2000).
However, in tropical and boreal areas, the response of dominant and subordinate species to
temperature is similar (ANDERSEN 1992); thus, dominant species may also dominate open
habitats (SAMWAYS 1990). This, however, is not the case in the dry grasslands of our study.
High richness and increased proportion of Mediterranean ant species (Table 3) suggest high
temperature and strong thermal variation at assessed habitats. It is possible that the microcli-
mate at the dry grasslands formed the current ant assemblages in a way observed in the
Mediterranean. This would further explain why abundant ant species, which dominated a
wide range of wet and mesic grassland habitats, were virtually absent from dry grasslands.
Fig. 5: Dry abandoned grasslands
(Asplenio-Festucion glaucae, class
Festuco-Brometea) in the
Štiavnické Vrchy Mts. host unique
ant assemblages even after decades
(Photo: M. Wiezik 2008).
Abb. 5: Aufgelassene Trocken-
rasen des Verbandes Asplenio-
Festucion glaucae (Klasse Festuco-
Brometea) in den Schemnitzer
Bergen (Štiavnické Vrchy)
beherbergen besondere Ameisen-
gesellschaften selbst Jahrzehnte
nach der Nutzungsaufgabe
(Foto: M. Wiezik 2008).
5.2. Effect of ecosystem stability
High species richness may increase ecosystem stability (TILMAN et al. 1996) and, con-
versely, long-term ecosystem stability may promote and maintain biodiversity (FJELDSÅ et
al. 1997). The dry grasslands in our study could be characterised by such long-term ecosys-
tem stability. In contrast to wet and mesic grasslands, whose open structure is maintained
(and has been created) by direct human intervention, dry grasslands tend to undergo less
rapid successional changes after abandonment. In fact, the extreme soil and humidity condi-
tions typical for dry grasslands may prevent excessive overgrowth after abandonment for
several decades. The management of dry grasslands has traditionally been moderate to rela-
tively extensive; however, its role in the preservation of open dry grassland structure is
recognised as vital (DEKONINCK et al. 2007, SCHRAUTZER et al. 2009). Several authors
suggest that a part of dry grasslands may be considered as natural, hence management-inde-
pendent (VERA 2000, BREDENKAMP et al. 2002, LOŽEK 2007); however, even in this case the
grazing of large mammals (both wild and domestic) represents a significant disturbance
factor. Although the overall open character of extreme dry grasslands may remain relatively
stable, changes in plant composition (mainly a decrease in light-demanding dry grassland
species) were reported after abandonment (MOOG et al. 2002). In general, when compared to
other grassland categories, dry grasslands have higher ecological stability, especially in
relation to secondary succession.
In our study, a large proportion of species richness in dry grasslands was due to steno-
topic steppe specialists (such as Ponera testacea, Messor structor, Tetramorium ferox, and
Camponotus piceus). In Central Europe, these species are characterised by limited occur-
rence within steppe and forest-steppe habitats (CZECHOWSKI et al. 2002, CSŐSZ & SEIFERT
2003, SEIFERT 2007). The dominance of these highly specialised species within dry grasslands
and, on the other hand, their absence from semi-natural habitats of wet and mesic grasslands
refer to a distinct community, paralleling the differences between the alliance Arrhenaterion
elatioris (moist and mesic grasslands) and the alliance Asplenio-Festucion glaucae (dry aban-
doned grasslands) and higher rank plant syntaxa: the class Molinio-Arhenatheretea compris-
ing fertilised meadows and pastures, and the class Festuco-Brometea comprising dry and
semi-dry grasslands and basiphilous oligotrophic grasslands.
5.3. Effect of low disturbance
Dry grasslands in our study included abandoned sites only. As currently managed dry
grasslands are not present in the studied region (KUNCA et al. 2005), abandonment could not
be compared to management for the dry sites. However, comparison with wet and mesic
sites allows for insight into the disturbance-related variation.
Disturbance is defined as a factor removing biomass (GRIME 1979). In Central Europe,
management in the form of mowing and grazing represents the main disturbance regime for
the majority of grasslands (MORRIS 2000). Although management is generally recognised as
a principle tool for protection and restoration of dry grasslands, target species have to be
recognised, and consequently, specific management treatments have to be applied, as the
response of different groups of plants and animals to a particular management may differ
(MORRIS 2000, SCHRAUZER et al. 2009). As shown by ENGLISCH et al. (2005), ants and vascu-
lar plants of Central European dry grasslands respond to habitat structure in different ways.
The results of our study demonstrate similar differences in response to management, sug-
gesting that annual mowing is a suitable practice for preservation of rich plant communities,
whereas the same treatment tends to decrease the species richness of associated ants. The
positive effect of cutting on plant diversity is mainly due to its support of annual and bienni-
al species, while dominant perennial species are rather impaired by the treatment (BONANOMI
et al. 2006). By preventing woody encroachment, mowing maintains the structure and com-
position of grassland vegetation, and hence, from the perspective of the grassland plant com-
munity, is regarded as vital and positive. From the perspective of invertebrates, however, the
effect of mowing is considerably different (MORRIS 2000). Cutting is non-selective and
reduces the vegetation on a site level to a uniform height. It strongly reduces topographical
features in the sward such as anthills and tussocks; by removing plant biomass it reduces the
food resources and shelter, and by exposing the soil surface it rapidly alters the micro -
climate. Hence, repeated annual cutting does appear to result in impoverishment of grass-
land fauna (MORRIS 1990, 2000). The response of ants to cutting is even more pronounced,
as sessile ant colonies have only limited opportunities to take evasive action. Besides the
mentioned negative effects, direct destruction of anthills and reduction of flowering plants
and phytophagous insects during cutting contribute to increased management-related
disturbance of ant communities.
The overall decrease in disturbance is considered the main reason why the majority of
arthropod groups experience rejuvenation during the cessation period between two manage-
ment treatments or in the early stage of grassland abandonment (MORRIS 2000). After aban-
donment, this regenerative effect of management absence appears to be only temporary and
is most pronounced during the initial period of secondary succession, after which the
increasing encroachment of woody species distinctly alters the structure and composition of
abandoned grasslands. As shown by WIEZIKOVÁ et al. (2010), the negative effect of sec-
ondary succession on grassland ant assemblages initially appeared after woody species
encroachment exceeded 10–30% of the grassland area, which, depending on various envi-
ronmental variables and management history, occurred several years to several decades after
grassland abandonment. As discussed earlier in this paper, shrub encroachment may be
strongly limited by the extreme soil and humidity conditions in dry grasslands. Thus, even
though the dry grassland habitats in the Štiavnické Vrchy Mts. have been abandoned for a
longer period, the rich and specialised ant assemblages found there may have benefited from
the combined effects of open structure and low disturbance persisting at this specific grass-
The study was funded by research grants nos. 1/0026/08, 1/0529/09, and 1/0557/10 of the Slovak Grant
Agency for Science (VEGA). We would like to acknowledge Anne Dovčiak and Aiko Huckauf for
English language editing. The manuscript was greatly improved by the comments of Camilla Wellstein
and two anonymous reviewers.
ANDERSEN, A. N. (1992): Regulation of “momentary” diversity by dominant species in exceptionally
rich ant communities of the Australian seasonal tropics. – Am. Nat. 140: 401–420. Chicago.
– (1995): A classification of Australian ant communities, based on functional groups which parallel plant
life-forms in relation to stress and disturbance. – J. Biogeogr. 22: 15–29.
– (2000): Global ecology of rainforest ants. Functional groups in relation to environmental stress and
disturbance. – In: AGOSTI, D., MAJER, J., ALONSO, L. E. & SCHULTZ, T. R. (Eds.): Ants: Standard
methods for measuring and monitoring biodiversity: 25–34. Smithsonian Institution Press, Washing-
ANDERSON, M. J. (2001): A new method for non-parametric multivariate analysis of variance. – Austral
Ecol. 26: 32–46.
– (2004): DISTLM v.5: a FORTRAN computer program to calculate a distance-based multivariate
analysis for a linear model. – Department of Statistics, University of Auckland, Auckland: 10 pp.
– & Ter Braak, C. J. F. (2003): Permutation tests for multi-factorial analysis of variance. – J. Statist.
Comput. Simul. 73: 85–113.
BESTELMEYER, B. T., AGOSTI, D., ALONSO, L. E., BRANDAO, C. R. F., BROWN, W. L., DELABIE, J. H. C.
& SILVESTRE, R. (2000): Field techniques for the study of ground-dwelling ants. – In: AGOSTI, D.,
MAJER, J., ALONSO, L. E. & SCHULTZ, T. R. (Eds.): Ants: Standard methods for measuring and moni-
toring biodiversity: 122–144. Smithsonian Institution Press, Washington, D.C.
BOBBINK, R. & WILLEMS, J. H. (1993): Restoration management of abandoned chalk grassland in the
Netherlands. – Biodivers. Conserv. 2: 616–626. Dordrecht.
BONANOMI, G., CAPORASO, S. & ALLEGREZZA, M. (2006): Short-term effects of nitrogen enrichment,
litter removal and cutting on a Mediterranean grassland. – Acta Oecol. 30: 419–425.
BRAY, J. R. & CURTIS, J. T. (1957): An ordination of upland forest communities of southern Wisconsin. –
Ecol. Monogr. 27: 325–349. Washington, DC.
BREDENKAMP, G. J., SPADA, F. & KAZMIERCZAK, E. (2002): On the origin of northern and southern
hemisphere grasslands. – Plant Ecol. 163: 209–229. Dordrecht.
CSŐSZ, S. & SEIFERT, B. (2003): Ponera testacea Emery, 1895 stat.n. – a sister species of P. coarctata
(Latreille, 1812) (Hymenoptera, Formicidae). – Acta Zool. Acad. Sci. Hung. 49: 201–214. Budapest.
CZECHOWSKI, W., RADCHENKO, A. & CZECHOWSKA, W. (2002): The ants (Hymneoptera, Formicidae) of
Poland. – Museum and Institute of Zoology PAS, Warszawa: 200 pp.
DAHMS, H., WELLSTEIN, C., WOLTERS, V. & DAUBER, J. (2005): Effects of management practices on ant
species richness and community composition in grasslands (Hymenoptera: Formicidae). –
Myrmekol. Nachr. 7: 9–16. Wien.
DAVID, S., KALIVODA, H., KALIVODOVÁ E. & ŠTEFFEK, J. (2007): Xerotermné biotopy Slovenska. –
Biosféra, Bratislava: 74 pp.
DEKONINCK, W., DE KONINCK, H., BAUGNÉE, J.-Y. & MAELFAIT J.-P. (2007): Ant biodiversity conser-
vation in Belgian calcareous grasslands: active management is vital. – Belg. J. Zool. 137: 137–146.
ELLENBERG, H. (1996): Vegetation Mitteleuropas mit den Alpen in ökologischer, dynamischer und his-
torischer Sicht. 5th ed. – Ulmer Verlag, Stuttgart: 1096 pp.
ENGLISCH T., STEINER, F. M., SCHLICK-STEINER B., C. (2005): Fine-scale grassland assemblage analysis
in Central Europe: ants tell another story than plants (Hymenoptera: Formicidae, Spermatophyta). –
Myrmecol. Nachr. 7: 61–67. Wien.
FJELDSÅ, J., EHRLICH, D., LAMBIN, E. & PRINS, E. (1997): Are biodiversity ‘hotspots’ correlated with
current ecoclimatic stability? A pilot study using the NOAA-AVHRR remote sensing data. – Bio-
divers. Conserv. 6: 401–427. Dordrecht.
FOLGARAIT, P. J. (1998): Ant biodiversity and its relationship to ecosystem functioning: a review. – Bio-
divers. Conserv. 7: 1221–1244. Dordrecht.
GILG, O. (2005): Old-growth forests: characteristics, conservation and monitoring. – ATEN Tech. Rep.
74. ATEN, Montpellier, France.
GÓMEZ, C., CASELLAS, D., OLIVERAS, J. & BAS, J. M. (2003): Structure of ground-foraging ant assem-
blages in relation to land-use change in the northwestern Mediterranean region. – Biodivers. Con-
serv. 12: 135–146. Dordrecht.
GREENSLADE, P. J. M. (1973): Sampling ants with pitfall traps: Digging-in effects. – Insectes Soc. 20:
GRIME, J. P. (1979): Plant strategies and vegetation processes. – Wiley, Chichester: 456 pp.
HEATWOLE, H. & MUIR, R. (1989): Seasonal and daily activity of ants in the pre-Saharan steppe of
Tunisia. – J. Arid. Environ. 16: 49–67.
HLAVAČEK, A. (1985): Flóra CHKO Štiavnické vrchy. – UŠOP, Liptovský Mikuláš: 776 pp.
JONES, C. G., LAWTON, J. H. & SHACHAK, M. (1994): Organisms as ecosystem engineers. – Oikos 69:
373 – 386. Copenhagen.
KONEČNÝ, V., LEXA, J., HALOUZKA, R., HÓK, J., VOZÁR, J., DUBLAN, L., NAGY, A., ŠIMON, L., HAVRILA,
M., IVANIČKA, J., HOJSTRIČOVÁ, V., MIHÁLIKOVÁ, A., VOZÁROVÁ, A., KONEČNÝ, P., KOVÁČIKOVÁ,
M., FILO, M., MARCIN, D., KLUKANOVÁ, A., LIŠČÁK, P. & ŽÁKOVÁ, E. (1998): Vysvetlivky ku geo-
logickej mape Štiavnických vrchov a Pohronského Inovca (štiavnický stratovulkán) I. diel, 1: 50 000.
– GSSR, Bratislava: 248 pp.
KORPEĽ, Š. (1989): Pralesy Slovenska. – Veda, Bratislava: 340 pp.
KRUSKAL, J. B. (1964): Multidimensional scaling by optimizing goodness of fit to a nonmetric hypo -
thesis. – Psychometrika 29: 1–27. New York.
KUNCA, V., ŠTEFFEK, J., OLAH, B., GAVLAS, V. & WIEZIK, M. (2005): Dynamika ekosystémov Štiavnick-
ých vrchov (zhodnotenie z pohľadu zmien využitia krajiny, štruktúry vybraných zoocenóz a stability
lesných ekosystémov). – Technical University in Zvolen, Zvolen: 103 pp.
LASSAU, S. A. & HOCHULI, D. F. (2004): Effects of habitat complexity on ant assemblages. Ecography
–, –, CASSIS, G. & REID C. A. M. (2005): Effect of habitat complexity on forest beetle diversity: do func-
tional groups respond consistently? – Divers. Distrib. 11: 73–82. Oxford.
LOŽEK, V. (2007): Zrcadlo minulosti. Česká a slovenská krajina v kvartéru. – Dokořán, Praha: 198 pp.
LUOTO, M., PYKÄLA, J. & KUUSAARI, M. (2003): Decline of landscape-scale habitat and species diversity
after the end of cattle grazing. – J. Nat. Conserv. 11: 171–178. Jena.
MOOG, D., POSCHLOD, P., KAHMEN, S. & SCHREIBER, K. F. (2002): Comparison of species composition
between different grassland management treatments after 25 years. – Appl. Veg. Sci. 5: 99–106. Uppsala.
MORRIS, M.G. (1990): The Hemiptera of two sown calcareous grasslands. III. Comparisons with the
Auchenorhynchafaunas of other grasslands. – J. Appl. Ecol. 27: 394–409. Oxford.
– (2000): The effects of structure and its dynamics on the ecology and conservation of arthropods in
British grasslands. – Biol. Conserv. 95: 129–142. Amsterdam.
NIEMELÄ, J. (1997): Invertebrates and boreal forest management. – Conserv. Biol. 11: 601–610. Boston.
ÖCKINGER, E., ERIKSSON, A.K. & SMITH, H. G. (2006): Effects of grassland abandonment, restoration
and management on butterflies and vascular plants. – Biol. Conserv. 133: 291–300. Amsterdam.
OTTONETTI, L., TUCCI, L. & SANTINI, G. (2006): Recolonization patterns of ants in a rehabilitated lig-
nite mine in Central Italy: Potential for the use of Mediterranean ants as indicators of restoration
processes. – Restor. Ecol. 14: 60–66.
RETANA, J. & CERDÁ, X. (2000): Patterns of diversity and composition of mediterranean ground ant
communities: tracking spatial and temporal variability in the thermal environment. – Oecologia 123:
RŪSI?A, S. & KIEHL, K. (2010): Long-term changes in species diversity in abandoned calcareous grass-
lands in Latvia. – Tuexenia 30: 467–486, Göttingen.
SAMWAYS, M. J. (1990): Species temporal variability: epigaeic ant assemblages and management for abun-
dance and scarcity. – Oecologia 84: 482–490. Berlin.
SCHRAUTZER, J., JANSEN, D., BREUER, M. & NELLE O. (2009): Succession and management of calcareous
dry grasslands in the Northern Franconian Jura, Germany. – Tuexenia 29: 339–351. Göttingen.
SEIFERT, B. (2007): Die Ameisen Mittel- und Nordeuropas. – Lutra, Görlitz/Tauer: 368 pp.
SIMILÄ, M., KOUKI, J. & MARTIKAINEN, P. (2003): Saproxylic beetles in managed and seminatural Scots
pine forest: quality of dead wood matters. – For. Ecol. Manage. 174: 365–381.
STATSOFT, INC. (2004): STATISTICA (data analysis software system), version 7.
TILMAN, D., WEDIN, D. & KNOPS, J. (1996): Productivity and sustainability influenced by biodiversity
in grassland ecosystems. – Nature 379: 718–720.
VAN DIJK, G. (1991): The status of semi-natural grasslands in Europe. – In: GORIUP, P. D., BATTEN, L. A.
& NORTON, J. A. (Eds): The conservation of lowland dry grassland birds in Europe: 15–36. Peter-
borough: Joint Nature Conservation Committee.
VERA, F. W. M. (2000): Grazing Ecology and Forest History. – CABI Publishing, Wallingford, Oxon:
WIEZIK, M. (2008): Ant communities (Hymenoptera: Formicidae) of forest-steppe habitats at southern
and south-western slopes of Plešivecká planina Plateau at Slovesnký kras Karst. – Nat. Carpat. 49:
WILLEMS, J. H. (2001): Problems, approaches, and results in restoration of Dutch calcareous grassland
during the last 30 years. – Restor. Ecol. 9: 147–154. Cambridge, Mass.
WIEZIKOVÁ, A., WIEZIK, M. & SVITOK M. (2010): Spoločenstvá mravcov pod vplyvom vybraných
stresových a disturbančných faktorov v podmienkach trvalých trávnych porastov podhorských
oblastí Slovenska. – Technical University in Zvolen, Zvolen: 99 pp.
Faculty of Ecology and Environmental Science, Technical University in Zvolen
T. G. Masaryka 24
Zvolen 960 53, SLOVAKIA
Co-ordinating editor: Camilla Wellstein
Manuscript received: 15.10.2010; accepted: 31.01.2011.