The effect of two ant species Lasius niger and Lasius flavus on soil ...
European Journal of Soil Biology 42 (2006) 158–165 http://france.elsevier.com/direct/ejsobi
Original article
The influence of different vegetation patches on the spatial distribution of nests and the epigeic activity of ants (Lasius niger) on a spoil dump after brown coal mining (Czech Republic) Michal Holec a,b,c,*, Jan Frouz a, Richard Pokorný c b c
a Institute of Soil Biology, Academy of Sciences of the Czech Republic, Na Sádkách 7, 337005 České Budějovice, Czech Republic Faculty of Biological Sciences, University of South Bohemia in České Budějovice, Na sádkách 7, 370005 České Budějovice, Czech Republic Faculty of Environmental Studies, University of J.E. Purkyně in Ústí nad Labem, Králova výšina 7, 400 96 Ústí nad Labem, Czech Republic
Received 3 January 2005; accepted 23 December 2005 Available online 03 February 2006
Abstract A study was carried out during 2001 on mine tailings in NW Bohemia aimed at describing the spatial patterns of nests distribution and epigeic activity of ants in relation to the vegetation mosaic. Lasius niger was the most abundant species of ant and its nest mounds were significantly more numerous in patches with sparse vegetation than inside dense Calamagrostis epigejos vegetation; this was particularly true for small and medium-sized nests. Small and medium nests also occurred more frequently near the edges of a given patch than in the center. Large and medium nests were randomly distributed in the area, whereas small nests had an aggregated distribution. Pitfall trapping reveal significantly higher activity of L. niger workers in tall and dense vegetation stands in comparison with low and sparse vegetation. This pattern was particularly pronounced during the peak of foraging activity in summer and was not so significant in spring or autumn. We expect that ant preferentially forage in shaded habitats during the summer months when bare soil may be too hot. The results indicated that nesting and foraging may differ in their microclimatic requirements and the formation of vegetation mosaics may be important to changes in the ant population during succession. © 2006 Elsevier SAS. All rights reserved. Keywords: Ants; Dumps; Tailings; Coal mining; Nest distribution; Vegetation pattern; Lasius niger
1. Introduction Ants interact with plants in many different ways. Ants are important seed eaters and dispersers which
* Corresponding
author. E-mail address: [email protected] (M. Holec).
1164-5563/$ - see front matter © 2006 Elsevier SAS. All rights reserved. doi:10.1016/j.ejsobi.2005.12.005
can be important for dispersion and distribution of certain namely myrmecochorous plant [9,17]. Ants are also considered as soil engineers [18,21] and their nest building activities significantly change soil conditions such as soil texture, water regime, nutrient supply, biotic properties [4,6,10,13,15,22,26,30]. In addition of these ants nests are important source of small-scale environment heterogeneity [6,8,24].
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
On the other hand vegetation significantly affects ants communities [2]. In particular vegetation structure is one of the most important factor affecting ants populations. Vegetation is a major regulator of microclimatic conditions [23], mainly temperature, which is important due to its influence on metabolic activity, development rate, and water loss and on activity in relation to the search for food [16,25,32]. Vegetation also directly or indirectly affects food sources for many ant species. Moreover, vegetation also influences soil properties, which play important role in the ecology of ants [22, 26]. During succession develops vegetation mosaic consisting from mixture of patches of vegetation typical for younger and more advanced succession stages. This mosaic form habitat heterogeneity that may be important for animals as some animals prefer different habitats for different activities [10]. In this study we focused on the effects of different vegetation types on ants populations on the spoil dumps left after brown coal mining. In particular we compared effect of dense vegetation cover dominated by expansive perennial rhizomatous grass Calamagrostis epigejos and more open vegetation patches overgrown by more diverse but sparse vegetation typical for younger succession stages. C. epigejos is important competitor of later successional stages on dumps and successfully expanding also to other disturbed sites (forest clearings, infrequently mowed or degraded meadows, dumps etc.). Prach [28] considered habitats with this species are arrested successional stages or subclimax [31]. In aspect of nature protection it is important species not only for its well spreading but also because difficulty decomposed litter is accumulated and colonization and persistence of other biota, including ants, may be limited. The general hypothesis is that two activities, nest building and foraging, of dominant species Lasius niger could differ in relation to different vegetation mosaic. In particular we tested what is effect of vegetation on distribution of nest, what is effect of nest distribution on epigeic activity and finally what is effect of vegetation mosaic and nest distribution in distribution of epigeic activity. 2. Materials and methods 2.1. Study area The study was carried out on the “Velká podkrušnohorská” dump in the Sokolov coal mining district (The Czech Republic). The altitude of the Sokolov region varies from 450 to 550 m a.s.l, the mean annual tem-
159
perature is 7–8 °C and annual precipitation 600– 650 mm. Study site was situated proximately 3 km north-west from Sokolov town—50°14′29.32″N, 12° 41′58.69″E. The tailings studied were created after open-cast brown coal mining and are formed of tertiary Miocene clay material. Further details about the area can be found in Frouz et al. [12]. Ants were studied on reclaimed sites, which had been leveled by earth moving machinery about 5 years ago, but total age (from heaping) of selected plots was proximately 20 years. Studied area was covered by a mosaic of bare soil and islands of vegetated patches, more closely described in following section (Section 2.2) and on pictures (Fig. 1). 2.2. Sampling and analysis Soil mound nests of L. niger were sampled in 50 × 50 m quadrates, during May 2001. The diameter of each nest was measured on the soil surface and the position of the nest was recorded as an xy co-ordinate, using the perpendicular distances from the east and south margins of the study area as measured by a leveling theodolite. Three size categories of nest mounds were recognized and their positions were recorded on a map (Fig. 1a): 1. Small mounds (diameter about 0.1 m). 2. Medium mounds (diameter usually about 0.3 m). 3. Large mounds (diameter usually about from 0.5 to 1 m). The nests were also classified according to their vitality as: 1. Living nests. or 2. Decaying or dead nests. Within living nest intensive and easy observable nest building activity was recorded. Moreover, nests on the heaps were usually only sparse vegetated (apart from nests from surrounding meadows) and this feature was maintained where nest were active. On the contrary, in dead or decaying nests, was either abandoned or occupied just by few workers apparently not adequate to nest size, no building activity was observed and nest mounds were covered by vegetation. This assessment was also recorded on the map (Fig. 1b). At the same time, the vegetation was also mapped and four basic vegetation types were recognized (Fig. 1): 1 – Dense Calamagrostis—floristically very poor habitat with vegetation strongly dominated by well developed clumps of the grass C. epigejos. Here, vegetation cover is about 95%, height approximately 100 cm and there is a dense litter layer accumulated on the soil surface. 2 – Sparse Calamagrostis—C. epigejos is still the dominant plant species, but it is not so dense (vegetation cover 70%) and there is only weak litter accumulation. 3. – Herbaceous vegetation—continuous, low
160
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
Fig. 1. Vegetation pattern and occurrence of L. niger nest mounds in the study quadrate (2500 m2). 1 – Dense vegetation with C. epigejos. 2 – Sparse Calamagrostis. 3 – Herbaceous vegetation. 4 – Open habitat. A) Distribution of nest mounds of different sizes: small medium big mounds . B) Distribution of nest mounds of different vitality: Δ – live, + dead or decaying mounds.
herbaceous vegetation covers the soil surface and species such as Avenella flexuosa, Poa pratensis, Lotus cornitulatus, Trifolium pratense are common. 4. – Open habitat—mosses and lichens cover most of the soil surface and vascular plants (Centaurea renana, Fragaria spp., Achilleum sp., L. cornitulatus, Poa compresa) cover only up to 40% of the soil surface. Herbaceous vegetation is sparse and floristically more diverse. To observe ant activity in the different vegetation types 36 pitfall traps (8 cm in diameter and 13 cm height), filled with 4% formaldehyde, and were randomly installed in 50 × 50 m quadrate; 17 traps were located within open habitat patches, 19 traps were positioned inside patches with dominant dense Calamagrostis. The traps were set out in March and emptied regularly after 2.5 months period, in May, August and November 2001. The ants were identified using the keys of Seifert [29] and Czechowski et al. [5]. 2.3. Data analysis The distribution of ant nests among individual habitats was compared with a random distribution pattern by the χ2-test. The expected, random distribution was calculated from the total number of nests and the proportions of each habitat in the whole area. The locations
of ant nests and the areas covered by individual vegetation types were mapped using Arc View, Gis 3.2. To describe the spatial distribution of nest mounds, an index of dispersion [20] was used in the form s2/x, where s2 is the variance and x is the mean number nest per unit area. To calculate the index of dispersion, 60 “sampling plots” of three different sizes (3 × 3, 6 × 6 and 9 × 9) were randomly selected on the map. Generation of random positions was performed in Microsoft Excel by calculating the distance of the south east corner of the sample plots from the south east corner of the area investigated using random number generator of Microsoft Excel. Since the numbers generated, by random number generator were between 0 and 1, we multiplied each number by 50 (the size of each side of our study plot) to give numbers between 0 and 50 which could then be used, in pairs, as co-ordinates on the x and y axes of the plot to locate random positions within the quadrate. Plots that overlapped the margin of the study area were not used. In these sampling plots, nests were counted and these data were used to calculate an index of dispersion. The index of dispersion reaches a value of about 1 for a random distribution, lower values for regular distributions and higher values for aggregated distributions [27]. To test if the index of dispersion significantly differs from 1, the criteria (s2/x) (n – 1) (where n
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
is number of samples, in our case 60) were compared with critical χ2 values for n – 1 degrees of freedom [9].
3. Results
Numbers of ants trapped in individual habitats were also compared with an expected random distribution, by the χ2-test. The numbers of ants trapped during each sampling interval were compared by one-way ANOVA. To answer the question whether the ants caught by the pitfall traps depended on the vegetation parameters closely surrounding the trap, or on the density of the ant population near the trap, the ant densities and vegetation parameters in the area surrounding the traps were correlated with the catches in individual traps. Around each trap, three circular areas were marked out, with a trap in the center and diameters of 1, 2 and 4 m. In these areas, cover of individual vegetation types, average vegetation cover, average vegetation height, number of ant nests and ant density were counted. Average vegetation cover was calculated as weight average of plant cover in individual vegetation types (sum of plant cover in individual vegetation type multiplied by the relative proportion of this vegetation type expressed as a decimal number). The same approach was used to calculate the average height. The density of ants in the surroundings of the trap was estimated by counting the nests in each category, and then multiplying the number of nests by the average number of workers per nest in a given category (1200 workers for a small nest, 4500 workers for a medium nest and 11,000 for a large nest). These numbers were based on nest excavation and counting the ants [14].
3.1. Nest distribution
To answer the question of whether the nests were located randomly within individual patches, or if they tended to be located more centrally or peripherally, we measured the distance between an individual nest and the nearest margin of the patch. These values were compared with the distances between 50 randomly selected points and the closest patch margin, using the ttest. All Analyses were made using Microsoft Excel and Statistica 5.5.
161
Most of the sampled quadrate areas were covered by dense Calamagrostis (1033.6 m2) or by open habitat (942.3 m2) (Fig. 1). All nests found in the area investigated belong to one species, L. niger. A total of 181 nest mounds were found in the quadrate sampled; 60 nests were small, 104 medium and 17 large nests (Fig. 3 and Table 1). The majority of nests were active (Fig. 1 and Table 1). The distribution of ant mounds among individual vegetation types differed significantly from the theoretical, random distribution (χ2-test, P < 0.05) for all size categories. Looking at small and mediumsized nests, as well as at total nest numbers, L. niger nests were more numerous in open habitats, dominated by sparse herbaceous vegetation (Table 1), whereas patches with dense Calamagrostis were only sporadically occupied by ant mounds. In contrast, large mounds were more frequent in dense Calamagrostis vegetation than expected from the random distribution and more rare in other vegetation patches. Moreover, the proportion of dead nests was higher in the dense Calamagrostis and lower in other vegetation types than expected (χ2, P < 0.05) (Table 1). Comparison of the nest locations with randomly selected points indicated that small and medium nests tended to be significantly closer to the patch margin than randomly chosen points (t-test, P < 0.05) whereas, in the case of large nests, no such significant difference was found. The spatial distribution of ant nests differs with nest size and the size of the sample plots (Table 2). In general, smaller nests tend to have a more aggregated distribution than large nests and larger sampling plots tend to result in a more aggregated distribution than smaller plots (Table 2). Looking at individual nest categories, only small nests showed a value of the dispersion index
Table 1 Frequency of L. niger nest mounds in individual vegetation types (1–4) in a quadrant of 2500 m2. Nest distribution among individual vegetation types was compared with a random distribution. * indicates that the nest distribution among individual vegetation types (in a particular column) differs from random (χ2, P < 0.05). In the columns where nest distribution differs significantly from a random distribution, < indicates vegetation types that harbor fewer nests that expected if the distribution was random and > indicates categories with more nests than expected Vegetation type Dense Calamagrostis Sparse Calamagrostis Herbaceous Open Total
Area (m2) 1033.6 284.2 239.9 942.3 2500.0
Small 7< 10 < 15 < 28 > 60 *
Nests of various size Middle Large 21 < 11 13 > 0 13 > 1 57 > 5 104 * 17
Live 27 < 23 > 29 > 86 > 165 *
Vitality Abandoned 12 0 0 4 16
Total 39 < 23 < 29 < 90 > 181*
162
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
Table 2 Index of dispersion values of all nests of L. niger and individual size categories of the nests in various sizes of sampling plots; > 1 indicates that the index is significantly higher than 1 (χ2, P < 0.05), other values did not differ significantly from 1 Plot Size 3×3 m 6×6 m 9×9 m
Small 1.531 > 1 2.070 > 1 2.159 > 1
Nest category Medium 1.251 1.264 1.286
All nest Large 0.917 1.332 1.000
1.512 > 1 1.784 > 1 1.813 > 1
significantly (χ2, P < 0.05) greater than 1, which determines the aggregated distribution. In the other two size categories, the dispersion indexes of medium and large nests did not differ significantly from 1 and their distribution can thus be assumed to be random. Small nests, however, represent the majority of L. niger nests found in the area investigated and, as a consequence, the dispersion index for all L. niger nests was also greater than 1, which indicates an aggregated distribution overall. 3.2. Epigeic activity Using 36 pitfall traps located in a quadrate of 2500 m2, a total of 5930 individual workers, belonging to six species were recorded (Table 3). L. niger (3399 individuals) was the dominant species. In total, 108 individual females belonging to four species were recorded (Table 3) among which L. niger was also dominant (87 individuals). Altogether, the number of females of all species was significantly higher in traps located within dense Calamagrostis patches, than in traps located in open habitat patches (t-test, P < 0.05). For workers, the same pattern was statistically significant for L. niger (t-test, P < 0.05), Myrmica scabrinodis (t-test, P < 0.05) and M. rubra (t-test, P < 0.05) (Table 3).
Fig. 2. Average number of ants caught by pitfall traps within different vegetation patches (open, C. epigejos and other vegetated habitats) during the observed period. Bars represent standard deviations.
We focused on the epigeic activity of the most dominant species L. niger, which was more active during summer than during spring or autumn (Fig. 2) and more ants were trapped in the dense Calamagrostis patches than in open habitat patches during summer (t-test, P < 0.05); no such difference was found in spring or autumn. Epigeic activity may be affected by many factors, such as microclimate, food availability, distribution of ant nests, etc. In this study we focused on two main factors, distribution of the nesting ant population around the trap, given by the number and size of ant nests, and vegetation parameters: cover of dominant vegetation types (open vegetation and dense Calamagrostis), average vegetation cover and average vegetation height. During summer, no correlation was found between numbers of ants nesting near the trap and the catch of workers in the trap. On the other hand, a significant correlation was found between trap catch and vegetation parameters near the nest (Table 4). Catch per trap positively correlated with vegetation cover, vegeta-
Table 3 Average numbers ± S.D. of worker and female ants caught by pitfall traps within dense Calamagrostis (A) (N = 19) and open habitat (B) (N = 17); A > B – ant numbers recorded within vegetation type A are significantly higher than numbers trapped in vegetation type B (t-test, P < 0.05)] Species
Habitat type B Workers 200 ± 159 104 ± 116 – – 5±9 5 ± 10 3±4 2±3 1±4 1±1 2±3 0±0 1±2 0±1 212 ± 155 112 ± 122 A
Lasius niger Lasius fuliginosus Formica cunicularia Formica rufibarbis Formica fusca Myrmica scabrinodis Myrmica rubra Totaly ants
A A>B – ns ns ns A>B A>B A>B
3±3 0±0 – – – 0±0 0±0 3±3
Habitat type B Females 1±2 0±0 – – 0±0 0±0 1±2
A>B ns– – – – ns ns A>B
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165 Table 4 Correlation coefficients between numbers of L. niger workers caught in pitfall traps and selected parameters in various areas around the trap. Only significant (P < 0.05) correlation coefficients are mentioned Sampling period/parameter 1m I Sampling period All parameters II Sampling period Ant density (individuals per area) Open habitat cover (%) Dense Calamagrostis cover (%) Vegetation cover (%) Vegetation height (cm) III Sampling period All parameters Whole year Ant density (individuals per area) Open habitat cover (%) Dense Calamagrostis cover (%) Vegetation cover (%) Vegetation height (cm)
Study area diameter 2m 4m
ns
ns
ns
ns ns 0.420 0.404 0.415
ns –0.370 0.441 0.398 0.446
ns –0.350 0.480 0.496 0.493
ns
ns
ns
ns ns 0.412 0.395 0.406
ns –0.370 0.434 0.390 0.438
ns –0.340 0.476 0.485 0.484
tion height and with cover of dense Calamagrostis, and negatively with the area of open habitat. These trends were similar for all sizes of area near the trap considered (Table 4). Thus the number of L. niger workers found in pitfall traps during summer depended more on vegetation parameters in the surroundings of the nest than on numbers of ants nesting in the vicinity of the trap. No significant correlation between these parameters and ant epigeic activity was found in spring or autumn. Thus, the pattern of activity in summer, in fact, determines the correlation between these parameters in the catch of a whole year. 4. Discussion 4.1. Ant community The ant community at the sites investigated was formed almost exclusively of one species, L. niger. Other species found were so rare that it is unlikely that interspecific competition plays an important role in the observed pattern of L. niger nests on this site. 4.2. L. niger nests Ants are, in general, thermophilous animals for whom higher temperatures are important to high brood production and population growth [2,3]. The open (unvegetated) patches are more easily warmed by insolation, which may ensure higher nest temperatures. This
163
is the most probable reason for their preference for building nests in sparse vegetation. This is particularly important for small nests, which may be easily shaded by dense and tall vegetation. On the other hand, the large nests were most likely established earlier, before tall and dense vegetation had overgrown them. It is also more likely that they are able to survive because a large nest, with a diameter of 0.5–1 m, can form an opening in dense vegetation, which is big enough to ensure sufficient exposure to the sun. Small and medium nests were more often situated on the edges of vegetation patches. This is in agreement with the observation of Dauber and Wolters [7], who found that L. niger prefer edges of patches in an agricultural landscape. Similarly, Braschler and Baur [1] observed that ants tend to be more concentrated on the edges of artificial patches made by experimental fragmentation of meadow habitat. There may be several reasons why ants prefer the margins of a patch: it enables them to forage in both habitats and to choose their habitat according to the more frequent food sources or better microclimatic conditions. Nest building at the boundary between tall and low vegetation may also be beneficial for microclimatic reasons. This position may ensure good insolation and, at the same time, shelter from wind [11]. Small nests were found to have an aggregated distribution, whereas large nests have a random distribution. A similar pattern was observed by Korb and Linsenmair [19], who studied the distribution of fungus feeding termites in Africa. These authors suggested that the distribution of young nests is aggregated because they tend to concentrate in patches with suitable microclimate. On the other hand, older nests have more regular distributions due to competition between individual nests. Similar mechanisms may also be operating in our case. Only a small number of big nests were found, which may indicate high mortality during their growth and competition pressure may be one of the factors which affect nest extinction. Alternatively, we suggest that, in our case, successions changes of vegetation may also affect more random distribution of old nests and aggregated distribution of young nests. After plot leveling by earth moving machinery, the plot may be quite homogeneous which may cause more random distribution of nests, which develop in the plot at that time. Later on, when vegetation mosaic develops, distribution of habitat suitable for young nest establishment become more patchy and this may influence the aggregated distribution of young nests.
164
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
4.3. L. niger epigeic activity We have found no effect of ant nest distribution on ant epigeic activity. On the other hand, the character of the vegetation surrounding the traps did significantly affect ant epigeic activity. Higher ant activity inside patches with dense vegetation may be related to food searching or to better microclimatic conditions. The fact that L. niger prefer dense Calamagrostis vegetation mainly during summer months characterized by higher temperature, may indicate that the preference of this habitat corresponds with avoidance of hot bare surfaces. In addition to the effect of temperature, ants are sensitive to, and try to avoid, ultraviolet radiation [2]. L. niger is partly dependent on an invertebrate diet; their prey, mainly other insects, is likely to be more frequent in vegetated patches where they are sheltered from the hotter environment of bare soil. A temperature versus food hypothesis is thus difficult to test with the present dataset. Greater female activity inside dense Calamagrostis patches seems to be in contradiction to the fact that most of the young nests were found in open patches. We suggest that microclimatic condition make migration under a shelter of dense vegetation more attractive than in true open patches and that dense vegetation perhaps offers some shelter from predators. These results have indicated that both nesting and foraging may differ in their microclimatic requirements and the formation of a vegetation mosaic may be important in changes to an ant population during succession. Acknowledgements This study was supported by the Czech Science Foundation (grant No. 526/01/1055), Grant Agency Academy of Sciences of the Czech Republic (grant No. 1QS600660505) and by Sokolovská uhelná a.s. coal mining company. We also thank the Sokolovská uhelná for the research permit and technical information about the plot history.
[5] [6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18] [19]
[20] [21]
References [22] [1]
[2] [3] [4]
B. Braschler, B. Baur, Effects of experimental small-scale grassland fragmentation on spatial distribution, density, and persistence of ant nests, Ecol. Entomol. 28 (2003) 651–658. M.V. Brian, Ants, Collins, London, 1977. M.V. Brian, A.D. Brian, Insolation and ant population in the west of Scotland, Trans. R. Soc. Lond. 102 (1951) 303–330. Z. Czerwinski, H. Jakubczyk, J. Pętal, Influence of ant hills on the meadow soil, Pedobiologia (Jena) 11 (1971) 277–288.
[23]
[24]
W. Czechowski, A. Radchenko, W. Czechowska, The ants of Poland. MIZ PAN, Warszawa, 2002. J. Dauber, V. Wolters, Microbial activity and functional diversity in the mounds of three different ant species, Soil Biol. Biochem. 32 (2000) 93–99. J. Dauber, V. Wolters, Edge effects on ant community structure and species richness in an agricultural landscape, Biodivers. Conserv. 13 (2004) 901–915. W.R.J. Dean, S.J. Milton, S. Klotz, The role of ants nestmounds in maintaining small-scale patchiness in dry grasslands in central Germany, Biodivers. Conserv. 6 (1997) 1293–1307. P. Dostal, Effect of three mound-building ant species on the formation of soil seed bank in mountain grassland, Flora 200 (2) (2005) 148–158. P. Dostál, M. Březnová, V. Kozlíčková, T. Herben, P. Kovář, Ant-induced soil modification and its effect on plant belowground biomass, Pedobiologia (Jena) 49 (2) (2005) 127–137. J. Frouz, M.G. Paoletti, Spatial distribution of different life stages of one Dipteran community along hedgerow and field margin, Landsc. Urban Plan. 48 (2000) 19–29. J. Frouz, B. Keplin, V. Pižl, K. Tajovský, J. Starý, A. Lukešová, A. Nováková, V. Balík, L. Háněl, J. Materna, C. Düker, J. Chalupský, J. Rusek, T. Heinkele, Soil biota and upper soil layer development in two contrasting post mining chronosequences, Ecol. Eng. 17 (2001) 275–284. J. Frouz, M. Holec, J. Kalčík, The effect of Lasius niger (Hymenoptera, Formicidae) ant nest on selected soil chemical properties, Pedobiologia (Jena) 47 (2003) 205–212. M. Holec, J. Frouz, Ant (Hymenoptera: Formicidae) communities in reclaimed and unreclaimed brown coal mining spoil dumps in the Czech Republic, Pedobiologia (Jena) 49 (2005) 345–357. N.R. Hulugalle, Effect of ant hills on soil physical properties of a Vertisol, Pedobiologia (Jena) 39 (1995) 34–41. T.F. Jensen, I. Hoolm-Jensen, Energetic cost of running in workers of three ant species, Formica fusca L., Formica rufa L. and Camponotus herculaneus L. (Hymenoptera, Formicidae), J. Comp. Physiol. 137 (1980) 151–156. R.A. Johnson, Seed-harvester ants (Hymenoptera: Formicidae) of North America: an overview of ecology and biogeography, Sociobiology 36 (2000) 83–122. C.G. Jones, J.H. Lawton, M. Shachak, Organisms as ecosystem engineers, Oikos 69 (1994) 373–386. J. Korb, K.E. Linsenmair, The causes of spatial patterning of mounds of a fungus-cultivating termite: results from nearestneighbour analysis and ecological studies, Oecologia 127 (2001) 324–333. C.J. Krebs, Ecological Methodology, Harper Collins Publisher, New York, 1987. P. Lavelle, Faunal activities and soil processes: adaptive strategies determine ecosystem function, Advances in Ecological Research 27 (1997) 93–132. L.A. Lobry de Bruyn, Ants as bioindicators of soil function in rural environments, Agric. Ecosyst. Environ. 74 (1999) 411– 425. F. López, J.M. Serrano, F.J. Acosta, Temperature–vegetation structure interaction: the effect on the activity of the ant Messor barbarus (L.), Vegetatio 100 (1992) 119–128. M.A. McGinley, S.S. Dhillion, J.C. Neumann, Environmental heterogeneity and seedling establishment: ant–plant–microbe interaction, Funct. Ecol. 8 (1994) 607–615.
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165 [25] M.G. Nielsen, Respiratory rates of ants from different climatic areas, J. Ins. Phisiol. 32 (1986) 125–131. [26] J.M. Petal, The role of ants in ecosystems, in: M.V. Brian (Ed.), Production Ecology of Ants and Termites, Cambridge University Press, London, 1978. [27] E.C. Pielou, Mathematical Ecology, Wiley, New York, 1977. [28] K. Prach, Succession of vegetation on Dumps from Strip Coal Mining, N.W. Bohemia, Czechoslovakia. Folia Geobot. Phytotax. 22 (1987) 339–354.
165
[29] B. Seifert, Ameisen: beobachten, bestimen, Naturbuch-Verl, Augsburg, 1966. [30] D. Wang, Spatial and temporal patterns of ant burrow openings as affected by soil properties and agricultural practices, Pedobiologia (Jena) 40 (1996) 201–211. [31] R.H. Whittaker, Climax concepts and recognition, in: R. Knapp (Ed.), Vegetation Dynamics, The Hague, 1974, pp. 139–154. [32] W.G. Whitford, C.A. Kay, A.M. Schumacher, Water loss in Chihuahuan desert ants, Phisiol. Zool. 48 (1976) 390–397.
Original article
The influence of different vegetation patches on the spatial distribution of nests and the epigeic activity of ants (Lasius niger) on a spoil dump after brown coal mining (Czech Republic) Michal Holec a,b,c,*, Jan Frouz a, Richard Pokorný c b c
a Institute of Soil Biology, Academy of Sciences of the Czech Republic, Na Sádkách 7, 337005 České Budějovice, Czech Republic Faculty of Biological Sciences, University of South Bohemia in České Budějovice, Na sádkách 7, 370005 České Budějovice, Czech Republic Faculty of Environmental Studies, University of J.E. Purkyně in Ústí nad Labem, Králova výšina 7, 400 96 Ústí nad Labem, Czech Republic
Received 3 January 2005; accepted 23 December 2005 Available online 03 February 2006
Abstract A study was carried out during 2001 on mine tailings in NW Bohemia aimed at describing the spatial patterns of nests distribution and epigeic activity of ants in relation to the vegetation mosaic. Lasius niger was the most abundant species of ant and its nest mounds were significantly more numerous in patches with sparse vegetation than inside dense Calamagrostis epigejos vegetation; this was particularly true for small and medium-sized nests. Small and medium nests also occurred more frequently near the edges of a given patch than in the center. Large and medium nests were randomly distributed in the area, whereas small nests had an aggregated distribution. Pitfall trapping reveal significantly higher activity of L. niger workers in tall and dense vegetation stands in comparison with low and sparse vegetation. This pattern was particularly pronounced during the peak of foraging activity in summer and was not so significant in spring or autumn. We expect that ant preferentially forage in shaded habitats during the summer months when bare soil may be too hot. The results indicated that nesting and foraging may differ in their microclimatic requirements and the formation of vegetation mosaics may be important to changes in the ant population during succession. © 2006 Elsevier SAS. All rights reserved. Keywords: Ants; Dumps; Tailings; Coal mining; Nest distribution; Vegetation pattern; Lasius niger
1. Introduction Ants interact with plants in many different ways. Ants are important seed eaters and dispersers which
* Corresponding
author. E-mail address: [email protected] (M. Holec).
1164-5563/$ - see front matter © 2006 Elsevier SAS. All rights reserved. doi:10.1016/j.ejsobi.2005.12.005
can be important for dispersion and distribution of certain namely myrmecochorous plant [9,17]. Ants are also considered as soil engineers [18,21] and their nest building activities significantly change soil conditions such as soil texture, water regime, nutrient supply, biotic properties [4,6,10,13,15,22,26,30]. In addition of these ants nests are important source of small-scale environment heterogeneity [6,8,24].
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
On the other hand vegetation significantly affects ants communities [2]. In particular vegetation structure is one of the most important factor affecting ants populations. Vegetation is a major regulator of microclimatic conditions [23], mainly temperature, which is important due to its influence on metabolic activity, development rate, and water loss and on activity in relation to the search for food [16,25,32]. Vegetation also directly or indirectly affects food sources for many ant species. Moreover, vegetation also influences soil properties, which play important role in the ecology of ants [22, 26]. During succession develops vegetation mosaic consisting from mixture of patches of vegetation typical for younger and more advanced succession stages. This mosaic form habitat heterogeneity that may be important for animals as some animals prefer different habitats for different activities [10]. In this study we focused on the effects of different vegetation types on ants populations on the spoil dumps left after brown coal mining. In particular we compared effect of dense vegetation cover dominated by expansive perennial rhizomatous grass Calamagrostis epigejos and more open vegetation patches overgrown by more diverse but sparse vegetation typical for younger succession stages. C. epigejos is important competitor of later successional stages on dumps and successfully expanding also to other disturbed sites (forest clearings, infrequently mowed or degraded meadows, dumps etc.). Prach [28] considered habitats with this species are arrested successional stages or subclimax [31]. In aspect of nature protection it is important species not only for its well spreading but also because difficulty decomposed litter is accumulated and colonization and persistence of other biota, including ants, may be limited. The general hypothesis is that two activities, nest building and foraging, of dominant species Lasius niger could differ in relation to different vegetation mosaic. In particular we tested what is effect of vegetation on distribution of nest, what is effect of nest distribution on epigeic activity and finally what is effect of vegetation mosaic and nest distribution in distribution of epigeic activity. 2. Materials and methods 2.1. Study area The study was carried out on the “Velká podkrušnohorská” dump in the Sokolov coal mining district (The Czech Republic). The altitude of the Sokolov region varies from 450 to 550 m a.s.l, the mean annual tem-
159
perature is 7–8 °C and annual precipitation 600– 650 mm. Study site was situated proximately 3 km north-west from Sokolov town—50°14′29.32″N, 12° 41′58.69″E. The tailings studied were created after open-cast brown coal mining and are formed of tertiary Miocene clay material. Further details about the area can be found in Frouz et al. [12]. Ants were studied on reclaimed sites, which had been leveled by earth moving machinery about 5 years ago, but total age (from heaping) of selected plots was proximately 20 years. Studied area was covered by a mosaic of bare soil and islands of vegetated patches, more closely described in following section (Section 2.2) and on pictures (Fig. 1). 2.2. Sampling and analysis Soil mound nests of L. niger were sampled in 50 × 50 m quadrates, during May 2001. The diameter of each nest was measured on the soil surface and the position of the nest was recorded as an xy co-ordinate, using the perpendicular distances from the east and south margins of the study area as measured by a leveling theodolite. Three size categories of nest mounds were recognized and their positions were recorded on a map (Fig. 1a): 1. Small mounds (diameter about 0.1 m). 2. Medium mounds (diameter usually about 0.3 m). 3. Large mounds (diameter usually about from 0.5 to 1 m). The nests were also classified according to their vitality as: 1. Living nests. or 2. Decaying or dead nests. Within living nest intensive and easy observable nest building activity was recorded. Moreover, nests on the heaps were usually only sparse vegetated (apart from nests from surrounding meadows) and this feature was maintained where nest were active. On the contrary, in dead or decaying nests, was either abandoned or occupied just by few workers apparently not adequate to nest size, no building activity was observed and nest mounds were covered by vegetation. This assessment was also recorded on the map (Fig. 1b). At the same time, the vegetation was also mapped and four basic vegetation types were recognized (Fig. 1): 1 – Dense Calamagrostis—floristically very poor habitat with vegetation strongly dominated by well developed clumps of the grass C. epigejos. Here, vegetation cover is about 95%, height approximately 100 cm and there is a dense litter layer accumulated on the soil surface. 2 – Sparse Calamagrostis—C. epigejos is still the dominant plant species, but it is not so dense (vegetation cover 70%) and there is only weak litter accumulation. 3. – Herbaceous vegetation—continuous, low
160
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
Fig. 1. Vegetation pattern and occurrence of L. niger nest mounds in the study quadrate (2500 m2). 1 – Dense vegetation with C. epigejos. 2 – Sparse Calamagrostis. 3 – Herbaceous vegetation. 4 – Open habitat. A) Distribution of nest mounds of different sizes: small medium big mounds . B) Distribution of nest mounds of different vitality: Δ – live, + dead or decaying mounds.
herbaceous vegetation covers the soil surface and species such as Avenella flexuosa, Poa pratensis, Lotus cornitulatus, Trifolium pratense are common. 4. – Open habitat—mosses and lichens cover most of the soil surface and vascular plants (Centaurea renana, Fragaria spp., Achilleum sp., L. cornitulatus, Poa compresa) cover only up to 40% of the soil surface. Herbaceous vegetation is sparse and floristically more diverse. To observe ant activity in the different vegetation types 36 pitfall traps (8 cm in diameter and 13 cm height), filled with 4% formaldehyde, and were randomly installed in 50 × 50 m quadrate; 17 traps were located within open habitat patches, 19 traps were positioned inside patches with dominant dense Calamagrostis. The traps were set out in March and emptied regularly after 2.5 months period, in May, August and November 2001. The ants were identified using the keys of Seifert [29] and Czechowski et al. [5]. 2.3. Data analysis The distribution of ant nests among individual habitats was compared with a random distribution pattern by the χ2-test. The expected, random distribution was calculated from the total number of nests and the proportions of each habitat in the whole area. The locations
of ant nests and the areas covered by individual vegetation types were mapped using Arc View, Gis 3.2. To describe the spatial distribution of nest mounds, an index of dispersion [20] was used in the form s2/x, where s2 is the variance and x is the mean number nest per unit area. To calculate the index of dispersion, 60 “sampling plots” of three different sizes (3 × 3, 6 × 6 and 9 × 9) were randomly selected on the map. Generation of random positions was performed in Microsoft Excel by calculating the distance of the south east corner of the sample plots from the south east corner of the area investigated using random number generator of Microsoft Excel. Since the numbers generated, by random number generator were between 0 and 1, we multiplied each number by 50 (the size of each side of our study plot) to give numbers between 0 and 50 which could then be used, in pairs, as co-ordinates on the x and y axes of the plot to locate random positions within the quadrate. Plots that overlapped the margin of the study area were not used. In these sampling plots, nests were counted and these data were used to calculate an index of dispersion. The index of dispersion reaches a value of about 1 for a random distribution, lower values for regular distributions and higher values for aggregated distributions [27]. To test if the index of dispersion significantly differs from 1, the criteria (s2/x) (n – 1) (where n
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
is number of samples, in our case 60) were compared with critical χ2 values for n – 1 degrees of freedom [9].
3. Results
Numbers of ants trapped in individual habitats were also compared with an expected random distribution, by the χ2-test. The numbers of ants trapped during each sampling interval were compared by one-way ANOVA. To answer the question whether the ants caught by the pitfall traps depended on the vegetation parameters closely surrounding the trap, or on the density of the ant population near the trap, the ant densities and vegetation parameters in the area surrounding the traps were correlated with the catches in individual traps. Around each trap, three circular areas were marked out, with a trap in the center and diameters of 1, 2 and 4 m. In these areas, cover of individual vegetation types, average vegetation cover, average vegetation height, number of ant nests and ant density were counted. Average vegetation cover was calculated as weight average of plant cover in individual vegetation types (sum of plant cover in individual vegetation type multiplied by the relative proportion of this vegetation type expressed as a decimal number). The same approach was used to calculate the average height. The density of ants in the surroundings of the trap was estimated by counting the nests in each category, and then multiplying the number of nests by the average number of workers per nest in a given category (1200 workers for a small nest, 4500 workers for a medium nest and 11,000 for a large nest). These numbers were based on nest excavation and counting the ants [14].
3.1. Nest distribution
To answer the question of whether the nests were located randomly within individual patches, or if they tended to be located more centrally or peripherally, we measured the distance between an individual nest and the nearest margin of the patch. These values were compared with the distances between 50 randomly selected points and the closest patch margin, using the ttest. All Analyses were made using Microsoft Excel and Statistica 5.5.
161
Most of the sampled quadrate areas were covered by dense Calamagrostis (1033.6 m2) or by open habitat (942.3 m2) (Fig. 1). All nests found in the area investigated belong to one species, L. niger. A total of 181 nest mounds were found in the quadrate sampled; 60 nests were small, 104 medium and 17 large nests (Fig. 3 and Table 1). The majority of nests were active (Fig. 1 and Table 1). The distribution of ant mounds among individual vegetation types differed significantly from the theoretical, random distribution (χ2-test, P < 0.05) for all size categories. Looking at small and mediumsized nests, as well as at total nest numbers, L. niger nests were more numerous in open habitats, dominated by sparse herbaceous vegetation (Table 1), whereas patches with dense Calamagrostis were only sporadically occupied by ant mounds. In contrast, large mounds were more frequent in dense Calamagrostis vegetation than expected from the random distribution and more rare in other vegetation patches. Moreover, the proportion of dead nests was higher in the dense Calamagrostis and lower in other vegetation types than expected (χ2, P < 0.05) (Table 1). Comparison of the nest locations with randomly selected points indicated that small and medium nests tended to be significantly closer to the patch margin than randomly chosen points (t-test, P < 0.05) whereas, in the case of large nests, no such significant difference was found. The spatial distribution of ant nests differs with nest size and the size of the sample plots (Table 2). In general, smaller nests tend to have a more aggregated distribution than large nests and larger sampling plots tend to result in a more aggregated distribution than smaller plots (Table 2). Looking at individual nest categories, only small nests showed a value of the dispersion index
Table 1 Frequency of L. niger nest mounds in individual vegetation types (1–4) in a quadrant of 2500 m2. Nest distribution among individual vegetation types was compared with a random distribution. * indicates that the nest distribution among individual vegetation types (in a particular column) differs from random (χ2, P < 0.05). In the columns where nest distribution differs significantly from a random distribution, < indicates vegetation types that harbor fewer nests that expected if the distribution was random and > indicates categories with more nests than expected Vegetation type Dense Calamagrostis Sparse Calamagrostis Herbaceous Open Total
Area (m2) 1033.6 284.2 239.9 942.3 2500.0
Small 7< 10 < 15 < 28 > 60 *
Nests of various size Middle Large 21 < 11 13 > 0 13 > 1 57 > 5 104 * 17
Live 27 < 23 > 29 > 86 > 165 *
Vitality Abandoned 12 0 0 4 16
Total 39 < 23 < 29 < 90 > 181*
162
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
Table 2 Index of dispersion values of all nests of L. niger and individual size categories of the nests in various sizes of sampling plots; > 1 indicates that the index is significantly higher than 1 (χ2, P < 0.05), other values did not differ significantly from 1 Plot Size 3×3 m 6×6 m 9×9 m
Small 1.531 > 1 2.070 > 1 2.159 > 1
Nest category Medium 1.251 1.264 1.286
All nest Large 0.917 1.332 1.000
1.512 > 1 1.784 > 1 1.813 > 1
significantly (χ2, P < 0.05) greater than 1, which determines the aggregated distribution. In the other two size categories, the dispersion indexes of medium and large nests did not differ significantly from 1 and their distribution can thus be assumed to be random. Small nests, however, represent the majority of L. niger nests found in the area investigated and, as a consequence, the dispersion index for all L. niger nests was also greater than 1, which indicates an aggregated distribution overall. 3.2. Epigeic activity Using 36 pitfall traps located in a quadrate of 2500 m2, a total of 5930 individual workers, belonging to six species were recorded (Table 3). L. niger (3399 individuals) was the dominant species. In total, 108 individual females belonging to four species were recorded (Table 3) among which L. niger was also dominant (87 individuals). Altogether, the number of females of all species was significantly higher in traps located within dense Calamagrostis patches, than in traps located in open habitat patches (t-test, P < 0.05). For workers, the same pattern was statistically significant for L. niger (t-test, P < 0.05), Myrmica scabrinodis (t-test, P < 0.05) and M. rubra (t-test, P < 0.05) (Table 3).
Fig. 2. Average number of ants caught by pitfall traps within different vegetation patches (open, C. epigejos and other vegetated habitats) during the observed period. Bars represent standard deviations.
We focused on the epigeic activity of the most dominant species L. niger, which was more active during summer than during spring or autumn (Fig. 2) and more ants were trapped in the dense Calamagrostis patches than in open habitat patches during summer (t-test, P < 0.05); no such difference was found in spring or autumn. Epigeic activity may be affected by many factors, such as microclimate, food availability, distribution of ant nests, etc. In this study we focused on two main factors, distribution of the nesting ant population around the trap, given by the number and size of ant nests, and vegetation parameters: cover of dominant vegetation types (open vegetation and dense Calamagrostis), average vegetation cover and average vegetation height. During summer, no correlation was found between numbers of ants nesting near the trap and the catch of workers in the trap. On the other hand, a significant correlation was found between trap catch and vegetation parameters near the nest (Table 4). Catch per trap positively correlated with vegetation cover, vegeta-
Table 3 Average numbers ± S.D. of worker and female ants caught by pitfall traps within dense Calamagrostis (A) (N = 19) and open habitat (B) (N = 17); A > B – ant numbers recorded within vegetation type A are significantly higher than numbers trapped in vegetation type B (t-test, P < 0.05)] Species
Habitat type B Workers 200 ± 159 104 ± 116 – – 5±9 5 ± 10 3±4 2±3 1±4 1±1 2±3 0±0 1±2 0±1 212 ± 155 112 ± 122 A
Lasius niger Lasius fuliginosus Formica cunicularia Formica rufibarbis Formica fusca Myrmica scabrinodis Myrmica rubra Totaly ants
A A>B – ns ns ns A>B A>B A>B
3±3 0±0 – – – 0±0 0±0 3±3
Habitat type B Females 1±2 0±0 – – 0±0 0±0 1±2
A>B ns– – – – ns ns A>B
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165 Table 4 Correlation coefficients between numbers of L. niger workers caught in pitfall traps and selected parameters in various areas around the trap. Only significant (P < 0.05) correlation coefficients are mentioned Sampling period/parameter 1m I Sampling period All parameters II Sampling period Ant density (individuals per area) Open habitat cover (%) Dense Calamagrostis cover (%) Vegetation cover (%) Vegetation height (cm) III Sampling period All parameters Whole year Ant density (individuals per area) Open habitat cover (%) Dense Calamagrostis cover (%) Vegetation cover (%) Vegetation height (cm)
Study area diameter 2m 4m
ns
ns
ns
ns ns 0.420 0.404 0.415
ns –0.370 0.441 0.398 0.446
ns –0.350 0.480 0.496 0.493
ns
ns
ns
ns ns 0.412 0.395 0.406
ns –0.370 0.434 0.390 0.438
ns –0.340 0.476 0.485 0.484
tion height and with cover of dense Calamagrostis, and negatively with the area of open habitat. These trends were similar for all sizes of area near the trap considered (Table 4). Thus the number of L. niger workers found in pitfall traps during summer depended more on vegetation parameters in the surroundings of the nest than on numbers of ants nesting in the vicinity of the trap. No significant correlation between these parameters and ant epigeic activity was found in spring or autumn. Thus, the pattern of activity in summer, in fact, determines the correlation between these parameters in the catch of a whole year. 4. Discussion 4.1. Ant community The ant community at the sites investigated was formed almost exclusively of one species, L. niger. Other species found were so rare that it is unlikely that interspecific competition plays an important role in the observed pattern of L. niger nests on this site. 4.2. L. niger nests Ants are, in general, thermophilous animals for whom higher temperatures are important to high brood production and population growth [2,3]. The open (unvegetated) patches are more easily warmed by insolation, which may ensure higher nest temperatures. This
163
is the most probable reason for their preference for building nests in sparse vegetation. This is particularly important for small nests, which may be easily shaded by dense and tall vegetation. On the other hand, the large nests were most likely established earlier, before tall and dense vegetation had overgrown them. It is also more likely that they are able to survive because a large nest, with a diameter of 0.5–1 m, can form an opening in dense vegetation, which is big enough to ensure sufficient exposure to the sun. Small and medium nests were more often situated on the edges of vegetation patches. This is in agreement with the observation of Dauber and Wolters [7], who found that L. niger prefer edges of patches in an agricultural landscape. Similarly, Braschler and Baur [1] observed that ants tend to be more concentrated on the edges of artificial patches made by experimental fragmentation of meadow habitat. There may be several reasons why ants prefer the margins of a patch: it enables them to forage in both habitats and to choose their habitat according to the more frequent food sources or better microclimatic conditions. Nest building at the boundary between tall and low vegetation may also be beneficial for microclimatic reasons. This position may ensure good insolation and, at the same time, shelter from wind [11]. Small nests were found to have an aggregated distribution, whereas large nests have a random distribution. A similar pattern was observed by Korb and Linsenmair [19], who studied the distribution of fungus feeding termites in Africa. These authors suggested that the distribution of young nests is aggregated because they tend to concentrate in patches with suitable microclimate. On the other hand, older nests have more regular distributions due to competition between individual nests. Similar mechanisms may also be operating in our case. Only a small number of big nests were found, which may indicate high mortality during their growth and competition pressure may be one of the factors which affect nest extinction. Alternatively, we suggest that, in our case, successions changes of vegetation may also affect more random distribution of old nests and aggregated distribution of young nests. After plot leveling by earth moving machinery, the plot may be quite homogeneous which may cause more random distribution of nests, which develop in the plot at that time. Later on, when vegetation mosaic develops, distribution of habitat suitable for young nest establishment become more patchy and this may influence the aggregated distribution of young nests.
164
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165
4.3. L. niger epigeic activity We have found no effect of ant nest distribution on ant epigeic activity. On the other hand, the character of the vegetation surrounding the traps did significantly affect ant epigeic activity. Higher ant activity inside patches with dense vegetation may be related to food searching or to better microclimatic conditions. The fact that L. niger prefer dense Calamagrostis vegetation mainly during summer months characterized by higher temperature, may indicate that the preference of this habitat corresponds with avoidance of hot bare surfaces. In addition to the effect of temperature, ants are sensitive to, and try to avoid, ultraviolet radiation [2]. L. niger is partly dependent on an invertebrate diet; their prey, mainly other insects, is likely to be more frequent in vegetated patches where they are sheltered from the hotter environment of bare soil. A temperature versus food hypothesis is thus difficult to test with the present dataset. Greater female activity inside dense Calamagrostis patches seems to be in contradiction to the fact that most of the young nests were found in open patches. We suggest that microclimatic condition make migration under a shelter of dense vegetation more attractive than in true open patches and that dense vegetation perhaps offers some shelter from predators. These results have indicated that both nesting and foraging may differ in their microclimatic requirements and the formation of a vegetation mosaic may be important in changes to an ant population during succession. Acknowledgements This study was supported by the Czech Science Foundation (grant No. 526/01/1055), Grant Agency Academy of Sciences of the Czech Republic (grant No. 1QS600660505) and by Sokolovská uhelná a.s. coal mining company. We also thank the Sokolovská uhelná for the research permit and technical information about the plot history.
[5] [6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18] [19]
[20] [21]
References [22] [1]
[2] [3] [4]
B. Braschler, B. Baur, Effects of experimental small-scale grassland fragmentation on spatial distribution, density, and persistence of ant nests, Ecol. Entomol. 28 (2003) 651–658. M.V. Brian, Ants, Collins, London, 1977. M.V. Brian, A.D. Brian, Insolation and ant population in the west of Scotland, Trans. R. Soc. Lond. 102 (1951) 303–330. Z. Czerwinski, H. Jakubczyk, J. Pętal, Influence of ant hills on the meadow soil, Pedobiologia (Jena) 11 (1971) 277–288.
[23]
[24]
W. Czechowski, A. Radchenko, W. Czechowska, The ants of Poland. MIZ PAN, Warszawa, 2002. J. Dauber, V. Wolters, Microbial activity and functional diversity in the mounds of three different ant species, Soil Biol. Biochem. 32 (2000) 93–99. J. Dauber, V. Wolters, Edge effects on ant community structure and species richness in an agricultural landscape, Biodivers. Conserv. 13 (2004) 901–915. W.R.J. Dean, S.J. Milton, S. Klotz, The role of ants nestmounds in maintaining small-scale patchiness in dry grasslands in central Germany, Biodivers. Conserv. 6 (1997) 1293–1307. P. Dostal, Effect of three mound-building ant species on the formation of soil seed bank in mountain grassland, Flora 200 (2) (2005) 148–158. P. Dostál, M. Březnová, V. Kozlíčková, T. Herben, P. Kovář, Ant-induced soil modification and its effect on plant belowground biomass, Pedobiologia (Jena) 49 (2) (2005) 127–137. J. Frouz, M.G. Paoletti, Spatial distribution of different life stages of one Dipteran community along hedgerow and field margin, Landsc. Urban Plan. 48 (2000) 19–29. J. Frouz, B. Keplin, V. Pižl, K. Tajovský, J. Starý, A. Lukešová, A. Nováková, V. Balík, L. Háněl, J. Materna, C. Düker, J. Chalupský, J. Rusek, T. Heinkele, Soil biota and upper soil layer development in two contrasting post mining chronosequences, Ecol. Eng. 17 (2001) 275–284. J. Frouz, M. Holec, J. Kalčík, The effect of Lasius niger (Hymenoptera, Formicidae) ant nest on selected soil chemical properties, Pedobiologia (Jena) 47 (2003) 205–212. M. Holec, J. Frouz, Ant (Hymenoptera: Formicidae) communities in reclaimed and unreclaimed brown coal mining spoil dumps in the Czech Republic, Pedobiologia (Jena) 49 (2005) 345–357. N.R. Hulugalle, Effect of ant hills on soil physical properties of a Vertisol, Pedobiologia (Jena) 39 (1995) 34–41. T.F. Jensen, I. Hoolm-Jensen, Energetic cost of running in workers of three ant species, Formica fusca L., Formica rufa L. and Camponotus herculaneus L. (Hymenoptera, Formicidae), J. Comp. Physiol. 137 (1980) 151–156. R.A. Johnson, Seed-harvester ants (Hymenoptera: Formicidae) of North America: an overview of ecology and biogeography, Sociobiology 36 (2000) 83–122. C.G. Jones, J.H. Lawton, M. Shachak, Organisms as ecosystem engineers, Oikos 69 (1994) 373–386. J. Korb, K.E. Linsenmair, The causes of spatial patterning of mounds of a fungus-cultivating termite: results from nearestneighbour analysis and ecological studies, Oecologia 127 (2001) 324–333. C.J. Krebs, Ecological Methodology, Harper Collins Publisher, New York, 1987. P. Lavelle, Faunal activities and soil processes: adaptive strategies determine ecosystem function, Advances in Ecological Research 27 (1997) 93–132. L.A. Lobry de Bruyn, Ants as bioindicators of soil function in rural environments, Agric. Ecosyst. Environ. 74 (1999) 411– 425. F. López, J.M. Serrano, F.J. Acosta, Temperature–vegetation structure interaction: the effect on the activity of the ant Messor barbarus (L.), Vegetatio 100 (1992) 119–128. M.A. McGinley, S.S. Dhillion, J.C. Neumann, Environmental heterogeneity and seedling establishment: ant–plant–microbe interaction, Funct. Ecol. 8 (1994) 607–615.
M. Holec et al. / European Journal of Soil Biology 42 (2006) 158–165 [25] M.G. Nielsen, Respiratory rates of ants from different climatic areas, J. Ins. Phisiol. 32 (1986) 125–131. [26] J.M. Petal, The role of ants in ecosystems, in: M.V. Brian (Ed.), Production Ecology of Ants and Termites, Cambridge University Press, London, 1978. [27] E.C. Pielou, Mathematical Ecology, Wiley, New York, 1977. [28] K. Prach, Succession of vegetation on Dumps from Strip Coal Mining, N.W. Bohemia, Czechoslovakia. Folia Geobot. Phytotax. 22 (1987) 339–354.
165
[29] B. Seifert, Ameisen: beobachten, bestimen, Naturbuch-Verl, Augsburg, 1966. [30] D. Wang, Spatial and temporal patterns of ant burrow openings as affected by soil properties and agricultural practices, Pedobiologia (Jena) 40 (1996) 201–211. [31] R.H. Whittaker, Climax concepts and recognition, in: R. Knapp (Ed.), Vegetation Dynamics, The Hague, 1974, pp. 139–154. [32] W.G. Whitford, C.A. Kay, A.M. Schumacher, Water loss in Chihuahuan desert ants, Phisiol. Zool. 48 (1976) 390–397.
