Applied Soil Ecology 33 (2006) 67–78 www.elsevier.com/locate/apsoil Soil mesofaunal responses to post-mining restoration treatments Pilar Andrés a,*, Eduardo Mateos b a Center for Ecological Research and Forestry Applications (CREAF), Autonomous University of Barcelona, 08193 Bellaterra, Barcelona, Spain b Department of Animal Biology of the Faculty of Biology, University of Barcelona, Diagonal 645, 08028 Barcelona, Spain Received 19 January 2005; received in revised form 8 August 2005; accepted 24 August 2005 Abstract As soil destruction is one of the most conspicuous environmental impacts of opencast mining, mining companies are compelled to restore the exploited areas and to develop monitoring programs to test the effectiveness of the restoration treatments. Soil physical and chemical parameters are usually used as soil quality indicators, but bioindicators are more promising as they evaluate the global soil capability to support its ecological functions. In this work, we used soil mesofaunal bioindicators to evaluate four post-mining restoration treatments (soil spreading, soil spreading + grass and herb sowing, soil spreading + tree planting and soil spreading + sowing + planting). Twelve years after restoration none of the treatments had achieved the restoration of the preoperational forest soil conditions. Soil spreading was the least effective treatment. Soil spreading + sowing and soil spreading + sowing + planting generated grassland soil conditions. Soil spreading + planting induced an incipient forest soil structure. Number of taxa, collembolan and oribatid species diversity and communites structure were the most sensitive mesofaunal parameters to evaluate the soil restoration treatments. # 2005 Elsevier B.V. All rights reserved. Keywords: Soil mesofauna; Soil quality; Bioindicators; Restoration; Mining 1. Introduction Landscape degradation is one of the most negative consequences of opencast mining. As social concern for landscape quality maintenance is growing in Europe, laws compelling mining companies to carry out postmining land restoration or rehabilitation are being implemented by environmental authorities. An evaluation of the success of minesite restoration is usually required prior to declaring the mine’s closure. After mining, the ecosystem may recover spontaneously if the resultant mineral substrate and the * Corresponding author. Tel.: +34 93 581 46 80; fax: +34 93 581 41 51. E-mail address: pilar.andres@uab.es (P. Andrés). 0929-1393/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.apsoil.2005.08.007 environmental conditions are adequate, but in most cases, deficient physical, chemical and biological soil conditions (e.g. unbalanced granulometry, low organic matter content, soil biota shortage) or isolation from colonization sources impede initiation of secondary succession or slow it down to a level incompatible with the social requirement for rapid solutions (Ash et al., 1994; Bradshaw, 1997). In these cases, a wide range of restoration techniques, of different intensities and costs, may be applied depending on resource availability and on ecosystem resilience. Commonly, restoration works include combinations of the following treatments: (a) in situ improvement of the physical and chemical substrate conditions; (b) top-soiling and later improvement and spreading of the recovered soil; (c) soil stabilization and protection with herbaceous species and (d) shrubs and 68 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 tree planting (Alcañiz et al., 1996; Green et al., 1999; Shopper, 1993; Vallejo et al., 1999). To evaluate the success of restoration methods, two tools are needed: a clearly defined reference state and a robust and easily measurable indicator or indicators system. When land restoration is the goal, undisturbed neighboring ecosystems may be used as reference sites. Land units supporting the designated land use, over a similar geological substratum and sharing climatic conditions, may be useful when the goal is rehabilitation or reclamation. Even though ecological restoration implies recreating the entire ecosystem structure, function and diversity, restoration projects often are only evaluated based on the reestablishment of the pre-disturbance dominant vegetation (Longcore, 2003). In the Mediterranean areas, secondary plant succession is slow, with an early dominance of annuals that are progressively substituted by forbs, grasses and woody species. Percentage of woody plants and plant cover are expected to increase with time after restoration. Nevertheless, the frequent coexistence in time of plants belonging to different life forms may question the value of their frequency distribution as a maturity indicator, and recent studies (Bonet and Pausas, 2004) conclude that plant cover in semi-arid Mediterranean systems is not dependent on plot age. Plant species diversity is expected to exhibit a peak attributable to immigration during the early period of 15 years since restoration, but plant diversity correlates more with the vegetation characteristics of the neighbouring undisturbed forest than with the restoration age (UB/CEIB-ESAB/ CREAF, 2001). Therefore, several environmental compartments must be monitored, with soil being the most strategic, given its well known key function in supporting the whole ecosystem (Bezdicek, 1996). The ‘‘post-restoration’’ soil quality is commonly monitored by using soil physical and chemical indicators, such as soil erodibility (Ojeda et al., 2003; Sort and Alcañiz, 1996), soil porosity and structure (Sort and Alcañiz, 1999a,b) or ability to support vegetation (Jorba and Andrés, 2000; Moreno, 2000; Zas and Alonso, 2002). More recently, landscape indicators, such as the ‘‘complexity habitat index’’ (Ludwig et al., 2003) are being calculated. National and international monitoring programs have proposed soil health bioindicators, including microbial biomass and soil respiration (Ortiz, 1998), N mineralization, microbial diversity and others related to the soil faunal communities (Andrés, 1999; Schloter et al., 2003) but there is no agreement about which soil biota bioindicators are most indicative of the final success of restoration or of the suitability of the successional process in a restored mine (Ekschmitt et al., 2003). Soil invertebrates may be appropriate tools in indicating the degree to which soil is affected by human activities (Kimberling et al., 2001; Ruf, 1998;) or land use intensification (Ponge et al., 2003) as they are sensitive to the anthropogenic disturbance, well correlated with soil functions and are helpful for elucidating ecosystem processes (Doran and Zeiss, 2000). Physical, chemical or landscape derived indicators may be less expensive than bioindicators in timeeffort or expertise (Ekschmitt et al., 2003), but they do not necessarily reflect the soil’s capability to maintain its essential ecological functions, due to their sensitivity to the time of sampling in relation to certain management and environmental events (van Bruggen and Semenov, 2000). In this work, we used soil mesofaunal bioindicators to evaluate the success of four post-mining restoration treatments in recreating the biological forest soil conditions. 2. Materials and methods This experiment is part of a multidisciplinary study conducted in 2000–2001 to use a set of vegetal and soil physical, chemical and biological indicators to evaluate the quality of 12 limestone quarries restored in Catalonia from 1987 to 1997 by different methods. During this period, four main restoration methods were applied by the mining companies: soil spreading over the mineral layer (soil spreading treatment), soil spreading followed by sowing of herbaceous species (sowing treatment), soil spreading followed by planting of shrubs and trees (planting treatment) and soil spreading followed by sowing and planting (mixed treatment). Lolium perenne (ryegrass), Festuca arundinacea (tall fescue), Dactylis glomerata (orchardgrass), Medicago sativa (lucerne) and Onobrichys sp. were the grasses and herbs usually sown. Pinus halepensis (Aleppo pine), Pinus pinea (stone pine) and Quercus ilex (holm oak) were the most commonly used trees, and Spartium junceum (Spanish broom), Pistacia lentiscus (mastic tree), and Quercus coccifera (kermes oak) were the most commonly planted shrubs. The experimental design for soil faunal bioindicators included 10 plots (two plots per treatment and two more for reference). The eight treatment plots, each 300 m2 in size, were located in four limestone quarries of the Garraf region of Barcelona (Catalonia, Spain) restored between 1988 and 1992. The two reference plots were selected in the unexploited forest areas adjacent to two 69 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 1 General plot characteristicsa Plots Plots code Quarry HC PD Sowing 1 Sowing 2 Mixed 1 Mixed 2 Planting 1 Planting 2 Spreading 1 Spreading 2 Reference 1 Reference 2 A1 A2 B1 B2 C1 C2 D1 D2 E1 E2 Santa Margarida Vallcara Santa Margarida Vallcarca Corral del Carro Vallcarca Santa Margarida El Tirano Santa Margarida Garraf 92.6  6.5 99.7  0.6 68  8 55.7  17 91.3  6.5 1.7  2.1 59  10.4 85.3  20.4 Data unavailable Data unavailable 1.23  0.15 0.63  0.17 0.6  0.07 1.48  0.16 1.65  0.61 0.36  0.08 1.72  0.47 1.39  0.52 Data unavailable Data unavailable HC, herbaceous cover in % (mean  S.E.); PD, plant diversity by the Shannon index in bits (mean  S.E.). a Sources: UB/CEIB-ESAB/CREAF, 2001. underlying upper 5 cm of the mineral layer was sampled with a cylindrical borer of 5 cm diameter and 5 cm depth. Samples were immediately transported to the laboratory for mesofaunal extraction in BerleseTullgren funnels (2 mm diameter mesh) for 15 days under constant light and temperature conditions. Soil animals were first classified into wide taxonomic groups. Mites were later determined at the suborder level. Mesostigmatid and prostigmatid mites were determined at family level and Cryptostigmata (oribatids) and collembolans at species level. This classification was considered the most efficient to determine the trophic structure of the arthropod community and the species diversity of oribatids and collembolans. Given the contagious distribution and the high micro-spatial heterogeneity of the soil fauna, we considered soil samples as independent replicates for the purpose of evaluating within-treatment variability. For each plot, we calculated the mesofaunal density, the number of taxa, the relative abundance of of the quarries. Distances between the four quarries were 3–18 km, and the geological substratum and climate were the same for all of them, with an average annual temperature of 14–16 8C and a mean annual precipitation of 525–575 mm. The selection of the experimental plots had to be adapted to the quarries available. As the quarries were restored by different mining companies, applying different restoration plans, none of the quarries available for monitoring included areas representative of all four treatments. However, we chose plots of similar age, climate, soil class and geological substrata from four quarries in the same homogeneous area, which included the four treatment types. The final allocation of the plots, together with their general characteristics, is summarized in Tables 1 and 2. Sampling was conducted in May 2000, when the biological activity in Mediterranean soils is high. We sampled at random at six points per plot. At each sampling point, the whole depth of the organic layer was taken with a square 10  10 section sampling tool. The Table 2 General plots soil characteristicsa Plots TC GR OM CC pH WC R Sowing 1 Sowing 2 Mixed 1 Mixed 2 Planting 1 Planting 2 Spreading 1 Spreading 2 Reference 1 Reference 2 Silt clay loam Loam Loam Loam Loam Silt loam Loam Loam Clay loam Loam 28.1 36.3 22.9 55.3 22.6 47.1 15.9 71.2 24.1 37.5 1.74 2.65 1.05 0.58 1.41 1.25 0.52 0.64 8.27 8.94 31.8 26.9 23.1 19.7 21 21.6 20.8 21.8 27.4 25.3 8.5 8.4 8.5 8.5 8.1 8.2 8.6 8.7 8.1 8.3 50.1 0.64 49.1 0.25 40.9 0.51 43.7 0.88 41.8 1.07 TC, soil textural class; GR, percentage of gravel; OM, percentage of soil organic matter; CC, percentage of clay; pH, pH measured in water; WC, mean water content at saturation as %; R, mean soil respiration in g CO2 m 2 h 1. a Sources: UB/CEIB-ESAB/CREAF, 2001. 70 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 collembolans and Acari, the Cryptostigmata/Prostigmata ratio, the vertical distribution and the trophic structure of the mesofauna. Also, we calculated the percentage of the four acarine suborders with respect to the total number of Acari. For collembolans and oribatids, we measured the species diversity by means of the Shannon index. The trophic structure of the mesofaunal community was characterized by calculating the percentage of predators. The vertical distribution of the mesofauna was calculated in terms of percentage of animals inhabiting the upper soil organic layer. Results were tested by means of nested-ANOVA with two levels (level 1, treatments; level 2, plots within treatments). To characterize the structure of the oribatid and collembolan communities, we compared treatments and undisturbed forest for collembolan and oribatid species abundance by means of the Wilcoxon’s signed rank test. Detrended correspondence analysis (DCA) and canonical correspondence analysis (CCA) were performed to look for similarity in species composition and in the oribatid and collembolan community structure. The community data matrix included 10 columns (one per plot) and 74 rows (the number of collembolan and oribatid species). The cells of the data matrix represent the mean population density of each species. In the variables data matrix, the nominal variable ‘‘treatment’’ (with five classes) was included by defining five dummy environmental variables: ‘‘sowing’’, ‘‘planting’’, ‘‘mixed’’, ‘‘reference’’ and ‘‘spreading’’. Thus, the variables data matrix included 10 columns and five rows. Additionally, simple regression analyses were conducted to test the utility of the soil physical and chemical parameters in predicting the soil mesofaunal community characteristics. 3. Results the most abundant taxa in all plots. No significant differences in their density or relative abundance were found between treatments. The density and relative abundance with respect to the total number of Acari of the mesostigmatid and astigmatid mites were also independent of the treatments. The density of the cryptostigmatid (oribatid) mites was independent of the treatments, but their relative abundance was affected by the treatments (Table 6) with the lowest value measured in the spreading treatment plots (Table 3). The relative abundance of the oribatids was also dependent on the plots (Table 6). The dominance of the prostigmatid mites was dependent on the treatments and on the plots (Table 6), with the lowest value in the reference plots and the highest in the spreading treatment plots. The Cryptostigmata/Prostigmata ratio was independent of the treatments. 3.3. Mesofaunal vertical distribution Significant differences in vertical distribution were found between treatments and between plots (Table 6). More than 50% of the soil mesofauna occupied the upper organic layers of the reference forest soils, whereas most of them (approximately 89%) occupied the deeper mineral layers in the sowing and in the soil spreading treatments. Planted plots (with or without sowing) showed an intermediate state, with about 33% of the mesofauna inhabiting the upper organic layer. 3.4. Relative abundance of predators Differences in percentage predators were not significant but this trophic class was especially abundant in the soil spreading treatment (Table 3) due to the high abundance of predatory mites belonging to the families Rhodacaridae (Mesostigmata), Bdellidae, Caligonellidae, Pseudochelidae and Rhagidiidae (Prostigmata). 3.1. Total mesofaunal density and composition 3.5. Oribatid and collembolan community structure Differences in total mesofaunal density (Table 3) between treatments and reference plots were not significant. The number of taxa per sample was significantly dependent on the treatments and significantly higher in the reference plots than in the restoration treatments (Table 6). 3.2. Relative abundance of the taxa Mites (67–79% of the total mesofauna) and collembolans (4–21% of the total mesofauna) were Twenty-six collembolan and at least 47 oribatid species (the Oppia species are grouped, except for O. concolor) were identified overall (Tables 4 and 5). For collembolans and oribatids, the species diversity measured by Shannon index was significantly affected by the treatments (Table 6). For oribatids, the diversity in the reference and in the sowing treatment was significantly greater than in the other treatments. For collembolans, it was greater in the reference than in any restoration treatment. For both taxa, the lowest species 71 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 3 Effect of restoration treatments on the density of various taxa (individuals/m2) and on some soil community indicators (mean  S.E.; n = 12) Treatments Reference Spreading Pseudoscorpionida Phalangida Araneida Acari Isopoda Pauropoda Penicillata Julida Glomerida Geophilomorpha Symphyla Protura Diplura Collembola Microcoryphia Zygentoma Orthoptera Isoptera Embioptera Psocoptera Homoptera Thysanoptera Coleoptera Hymenoptera Diptera Immature insects 135  55 88 17  11 17909  2549 17  11 390  192 177  95 127  91 17  17 175  175 11170  2011 135  63 102  56 562  172 42  42 1644  1480 17  11 261  85 126  74 320  185 100  33 Total density Taxa/sample % Of predators % Fauna in organic layer % Mesostigmata a % Cryptostigmataa % Astigmataa % Prostigmata a Cryptos/Prostig 26363  3321 7.9  0.6 18.0  1.5 56.6  6.1 13.4  2.1 71.2  2.3 2.4  1 13  1.3 6.6  1 a 88 88 424  175 772  422 3719  585 88 88 85  85 42  42 Sowing Planting Mixed 12104  2414 88 21339  9039 42  42 42  42 25315  6698 42  42 42  42 42  42 136  91 88 42  42 42  42 42  42 42  42 1197  616 85  57 42  42 3444  1069 42  42 2403  574 42  42 85  85 2507  1006 88 88 85  85 88 127  88 127  91 344  163 88 135  98 1464  900 102  57 25  18 288  180 144  64 346  159 209  143 1281  686 254  109 42  42 227  86 92  75 609  253 709  289 228  98 814  368 25  18 516  225 51  42 220  91 14126  2420 4.7  0.5 28.1  6.0 11.9  4.13 8.4  2.5 46.4  7.4 8.8  4.3 36.5  9.3 5.2  2.2 31567  6753 5.3  0.5 20.4  2.5 10.9  3.3 10.7  2.9 52.2  4.6 8.8  2.7 28.3  3.6 2.4  0.4 17364  3295 5.1  0.3 16.9  3.8 33.0  7.4 11.6  3.5 67.5  7.6 3.5  1 17.4  4.7 10.6  3.5 26739  9784 5.8  0.4 20.5  3.7 30.9  8.7 14.5  4.9 53.1  6.6 6.2  2.9 26.2  6.2 4.5  1.9 % of Mesostigmata, Cryptostigmata, Astigmata and Prostigmata with respect to the total number of Acari. diversity was found in the soil spreading treatment. As shown in Table 7, differences in the diversity were coupled with differences in the dominant species. When considering the oribatid and collembolan species composition and abundance, the highest similarity between the treatments and the reference plots was obtained with the planting and with the mixed treatments for oribatids and with the sowing and mixed treatments for collembolans (Tables 4 and 5). 3.6. Multivariate analysis The first two axes of the DCA explained 36% of the variance. The ordination diagram displayed different species composition by plots (Fig. 1). Except for the spreading treatment, the plots are not associated by treatments and do not display any clarifying spatial distribution pattern. The first two axes of the CCA (Fig. 2) explained 32.1% of the variance within species. The speciesenvironment correlation coefficient was 0.965 for the first axis and 0.952 for the second. The correlation coefficients (inter-set correlations) of the five dummy environmental variables with the first canonical axis were: 0.96 for reference, 0.21 for sowing, 0.19 for planting, 0.40 for mixed and 0.23 for spreading. With the second axis, the correlation coefficients were: 0.10 for reference, 0.89 for sowing, 0.07 for planting, 0.62 for mixed and 0.13 for spreading. By Monte Carlo permutation tests on the effect of the 72 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 4 Collembolan density in individuals/m2 (mean  S.E.; n = 12), species composition, species diversity (Shannon index) and similarity between treatments and reference plots with regard to species abundance by means of the Wilcoxon’s test Reference Xenylla brevisimilis (Gama, 1964) Ceratophysella tergilobata (Cassagnau, 1954) Ceratophysella denticulata (Bagnall, 1941) Friesea mirabilis (Tullberg, 1871) Friesea truncata (Cassagnau, 1958) Pseudachorudina meridionalis (Bonet, 1929) Choreutinula inermis (Tullberg, 1871) Deutonura cf sinistra (Denis, 1935) Protaphorura cf nemorata (Gisin, 1952) Mesaphorura sp. Neotullbergia sp. Parisotoma notabilis (Schäffer, 1896) Cryptopygus albaredai (Selga, 1962) Isotomiella minor (Schäffer, 1896) Folsomia manolachei (Bagnall, 1939) Folsomides parvulus (Stach, 1922) Archisotoma sp. Isotomurus sp. Entomobrya sp. 1 Entomobrya sp. 2 Lepidocyrtus lanuginosus (Gmelin, 1788) Lepidocyrtus lignorum (Fabricius, 1793) Pseudosinella sp. Heteromurus major (Monniez, 1889) Sminthurinus aureus (Lubbock, 1862) Bourletiella sp. Total collembola Species diversity Wilcoxon’s test p-values 450  140 382  218 Spreading Sowing Planting Mixed 42  42 202  109 1124  740 42  42 42  42 475  475 170  96 88 17  11 85  85 202  160 42  42 25  18 321  153 85  85 340  130 17  17 401  365 170  72 17  11 176  60 740  249 236  107 976  665 42  42 42  42 85  57 42  42 382  236 127  66 212  171 263  254 144  91 687  326 712  430 159  55 88 17  11 42  42 3719  585 2.03  0.14 – 1197  616 0.36  0.19 0.048 170  170 42  42 202  162 177  76 42  42 59  42 42  42 127  91 85  85 679  471 85  57 340  220 340  253 42  42 42  42 496  211 178  76 42  42 42  42 59  59 152  90 51  51 88 88 3444  1069 1.06  0.21 0.338 2403  574 1.08  0.20 0.180 2507  1006 0.56  0.20 0.21 five dummy environmental variables, only ‘‘reference’’ was significant (99 permutations, F-ratio = 1.77, p = 0.003). Thus, the collembolan and oribatid communities were different in the reference undisturbed forests and in the restored areas. Additionally, even though the dummy environmental variable ‘‘sowing’’ was not significant, the ‘‘sowing’’ treatment plots were distant from the others in the second axis, and we concluded that faunal communities obtained by sowing were different from those obtained by applying any other treatment. 3.7. Soil environmental parameters versus soil bioindicators Fig. 1. Detrended correspondence analysis (DCA) ordination diagram for collembolan and oribatid species showing treatment plots (A1 and A2, sowing; B1 and B2, mixed; C1 and C2, planting; D1 and D2, spreading; E1 and E2, reference). Total inertia = 2.299, axis 1 eigenvalue = 0.493, axis 2 eigenvalue = 0.334. Significant correlations with the mesofaunal diversity and species richness were found for soil organic matter content, soil pH and clay content (Table 8). The highest coefficients of determination (R2) were found between organic matter content and collembolan diversity and richness and between organic matter 73 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 5 Oribatid density in individuals/m2 (mean  S.E.; n = 12), species composition and species diversity (Shannon’s index) and similarity between treatments and reference with regard to species abundance, by means of the Wilcoxon’s test Reference Allogalumna subaequale (Mihelčič) Aphelacarus acarinus (Berlese) Arthrodamaeus reticulatus (Berlese) Berlesezetes auxiliaris (Grandjean) Camisia spinifer (Koch) Carabodes aerolatus (Berlese) Carabodes willmanni (Bernini) Ceratozetes conjunctus (Mihelčič) Phyllozetes emmae (Berlese) Cosmochthonius lanatus (Michael) Chamobates cuspidatus (Michael) Damaeus flagellifer (Michael) Epilohmannia cylindrica (Berlese) Eueremaeus granulatus (Mihelčič) Eupelops acromios (Hermann) Eupelops sp. Galumna sp. Haplozetes sinuatus (Pérez-Íñigo Jr.) Hemileius initialis (Berlese) Hermanniella dolosa (Grandjean) Hypovertex sp. Incabates sp. Liochthonius sp. Lucoppia burrowsi (Michael) Metabelbella sp. Microzetes petrocoriensis (Grandjean) Neoliodes ionicus (Sellnick) Oppia Oppia concolor (Koch) Oribatella berlesei (Michael) Oribatula tibialis (Nicolet) Papillacarus aciculatus (Berlese) Passalozetes africanus (Grandjean) Peloribates europaeus (Willmann) Pilobates carpetanus (Pérez-Íñigo) Pilogalumna sp. Platyliodes sp. Rhysotritia sp. Scheloribates laevigatus (Koch) Scheloribates sp. Scutovertex sculptus (Michael) Sphaerochthonius splendidus (Berlese) Steganacarus magnus (Nicolet) Tectocepheus velatus (Michael) Trimalaconothrus sp. Xenillus tegeocranus (Hermann) Zetorchestes flabrarius (Grandjean) Immature Cryptostigmata Total oribatids Species diversity Wilcoxon’s test p-values 42  34 192  152 944  227 351  132 17  17 50  36 297  221 83  41 1122  419 33  26 271  145 17  17 117  50 88 67  40 Spreading 144  91 42  42 Sowing Planting Mixed 42  42 1756  888 85  57 42  42 88 101  69 42  42 127  91 552  193 483  261 42  42 88 42  42 85  57 85  57 85  57 1001  422 135  70 33  33 328  177 500  424 308  290 42  42 42  42 17  17 88 25  25 433  271 33  26 17  17 168  90 88 42  42 85  57 42  42 25  25 88 167  167 3473  1257 88 42  34 233  108 127  127 492  261 601  472 466  317 532  376 42  42 340  181 398  144 170  114 449  241 951  283 202  133 88 17  17 42  42 42  42 229  229 1061  466 485  143 150  101 42  42 652  232 88 133  50 2341  1716 93  85 88 1969  1696 218  102 42  42 177  103 1103  425 244  152 127  88 305  254 125  125 85  57 85  57 1001  463 297  98 25  25 3471  2086 1220  630 541  238 88 51  43 2503  574 3853  1562 3142  1123 2301  416 9782  6536 12719  1881 2.54  0.18 – 6039  1662 1.42  0.22 0.007 13693  4531 2.06  0.13 0.002 8862  2374 1.50  0.13 0.127 16055  8865 1.44  0.19 0.132 74 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 6 Results of the nested ANOVA for the effects of treatments and plots within treatments on various faunal parameters d.f. Number of taxa per sample Treatments 4 Plots (in treatments) 5 Error 50 MS F 18.8 4.3 3.0 6.13 1.41 P Treatments Oribatidsa Collembolans Sowing T. velatus A. acarinus P. notabilis L. lignorum L. lanuginosus Planting S. laevigatus T. velatus O. tibialis Folsomides sp. Entomobrya sp. 2 Isotomurus sp Mixed Galumna sp. S. splendidus X. brevisimilis C. denticulada Spreading Galumna sp. O. tibialis T. velatus Pilogalumna sp Isotomurus sp. L. lanuginosus Reference Oppia ssp. Ch. cuspidatus Platyliodes sp. L. lanuginosus X. brevisimilis I. minor C. tergilobata 0.0004 0.2368 % Cryptostigmata with respect to the total acari number Treatments 4 1371.7 4.66 Plots (in treatments) 5 1899.7 6.46 Error 50 294.0 0.0028 0.0001 % Prostigmata with respect to the total acari number Treatments 4 1028.7 3.72 Plots (in treatments) 5 1504.3 5.44 Error 50 276.5 0.0099 0.0004 % Fauna inhabiting the organic layer Treatments 4 4192.3 Plots (in treatments) 5 1842.3 Error 50 333.5 12.57 5.52 <0.0001 0.0003 <0.0001 0.3524 Oribatid species diversity Treatments Plots (in treatments) Error Table 7 Species contributing 50% of the total collembolans and oribatids for each treatment a 4 5 48 2.9 0.3 0.3 9.64 1.13 Collembolan species diversity Treatments 4 Plots (in treatments) 5 Error 39 3.4 0.2 0.3 8.65 0.50 <0.0001 0.7704 Immature oribatids were not included in counting. large population sizes and ability to occupy a wide range of ecosystems, microhabitats and niches make the soil arthropods useful to monitor year-to-year ecosystem environmental changes (Kremen et al., 1993; content and oribatid specific diversity. Soil pH explained more than 50% of the variability in the collembolan species richness and specific diversity and the percentage of clay explained approximately 60% of the oribatid specific diversity. 4. Discussion The aim of the post-mining restoration treatments was to accelerate secondary succession towards the recovery of the former forest in the newly exposed soils. Given that soil matrix, soil micro-, meso- and macrobiota and plant interactions evolve throughout the successional process, different restoration methods should be compared for effectiveness, at a given moment, by evaluating the maturity of the restored areas by means of physical, chemical and biological indicators sensitive to the restoration age. Together with plants, many soil arthropods sensitive to ecological changes occurring over time after restoration have been recognised as efficient indicators of ecosystem maturity for conservation planning as well as land reclamation monitoring (Greenslade and Majer, 1993; Kremen et al., 1993). Short generation times, Fig. 2. Canonical correspondence analysis (CCA) ordination diagram for collembolan and oribatid species showing treatment plots (A1 and A2, sowing; B1 and B2, mixed; C1 and C2, planting; D1 and D2, spreading; E1 and E2, reference). Arrows pointed towards dummy environmental variables (1, sowing; 2, mixed; 3, planting; 4, spreading; 5, reference). Total inertia = 2.299, axis 1 eigenvalue = 0.418, axis 2 eigenvalue = 0.320. Theo environmental variables scales are multiplied by two in relation to plots and species scale. P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Table 8 Best regressions found between mesofaunal parameters and soil abiotic factors (n = 10) pH (H2O) Organic matter % Clay Collembolan diversity (H) R2 = 0.919 R2 = 0.56 p = 0.012 p < 0.0001 Simple ( ) Logarithmic Oribatid diversity (H) R2 = 0.946 p < 0.0001 Logarithmic R2 = 0.62 p = 0.034 Logarithmic Number of collembolan species R2 = 0.684 R2 = 0.53 p = 0.009 p < 0,0001 Simple ( ) Simple (+) Longcore, 2003). However, soil invertebrate communities are subject to natural stochastic variation, and their diversity, distribution and abundance may be determined by a combination of several factors, such as vegetative cover (Al-Assiuty et al., 1999; Kallimanis et al., 2002), soil nutrient status, organic mater content (Ponge et al., 2003) and soil texture or soil acidity (David et al., 1999) which may differ in importance from one invertebrate group to another (Berch et al., 2001; David et al., 1999). In this sense, the use of soil invertebrates as bioindicators requires first to establish how the different species respond to a variety of environmental factors (Hodkinson and Jackson, 2005). In our plots, an additional source of variability comes from the fact than they belonged to different mining companies and were restored with similar but not identical soil materials. This last factor may explain why, in some cases, plot effects are as strong as treatment effects. As for vegetation, post-restoration recovery of the invertebrate community is slow and not shorter than 15 years (Neumann, 1991; Webb, 1994), with 80–102 years estimated for the forest collembolan community to recover (Addison et al., 2003). The colonization of newly restored soils by the soil fauna coming from the adjacent undisturbed forests depends mainly on active locomotion for the macrofauna and on aerial immigration or on phoresy for mesoarthropods (St. John et al., 2002). The immigration strategies are taxon-specific, with collembola and trombidiidae establishing first within 15 months (Wanner and Dunger, 2002). The maximal mesofaunal density is usually achieved during the 2–3 year ‘‘pioneer’’ phase, which is followed by a drastic reduction in density to levels of less than 20% within the following 75 10 years (Koehler, 1998). Subsequent changes in taxonomic composition and relative abundance may be related to successional changes in factors, such as plant cover, soil pH and organic matter content, etc. (Black et al., 2003). In our work, we found no differences in total mesofaunal density between the undisturbed forests and the 12-year-old restored plots, but these simple density data, obtained from a unique sampling campaign, are not enough to evaluate the success of the soil community recovery and may reflect a transitory state in the still young restoration successional process. On the other hand, similar densities may correspond to different soil community structures indicative of different soil maturity stages. In this sense, the richness of taxa was higher in the undisturbed forests than in any treatment, due to the presence of some large, mobile, typical forest taxa, such as the jumping Bristletails (Microcoryphia), woodlice (Isopoda), centipedes (Geophilomorpha) or millipedes (Glomerida) in the reference plots. As a progressive increase of these macroarthropods may be expected from grassland to scrubland and forests (Ferguson, 2001), we may conclude that the richness of taxa is indicative of the maturity of the vegetal community of the restored areas. From this perspective, the soil faunal diversity in our plots was still not that expected for a mature forest, but rather showed the characteristics of a successional intermediate stage. The dominance of the mite suborder Prostigmata was an interesting indicator. The prostigmatid mites are abundant in temperate meadows (Kethley, 1990) and lawns (Koehler, 1998). Prostigmata are also favored by anaerobiosis and are indicative of human soil disturbance (Clapperton et al., 2002; Philips, 1990). In our work, Prostigmata (belonging in most cases to the families Bdellidae, Caligonellidae and Rhagidiidae) were most abundant in the spreading treatment, indicating the low degree of effectiveness of this treatment in facilitating the recovery of the undisturbed forest soil conditions. The Cryptostigmata/Prostigmata ratio has been discussed by many authors as an indicator of the organic matter content in soils (Cepeda-Pizarro and Withford, 1989; Kethley, 1990). Given that Cryptostigmata are representative of undisturbed sites whereas Prostigmata abundance increases with disturbance (Clapperton et al., 2002), high values of the ratio correspond with undisturbed forest soils, since low values are characteristic of the forest transformation to more immature stages (Sgardelis and Usher, 1994). In our work, this ratio was maximal in the undisturbed 76 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 forests and in the planting treatment and minimal in the spreading treatment. In mature Mediterranean soils, mesofauna inhabit mainly the upper organic layers, from which they migrate to the deeper layers or to the litter depending on how favorable or adverse climatic and trophic conditions are (Andrés, 1998; Andrés et al., 1999). In forests, the soil mesofauna are particularly concentrated in the aboveground/underground interface area, where litter is abundant and where niche diversity is the highest. During the first post-restoration stages, the newly spread soil layer is vertically homogeneous, and the later progression through stratification is promoted by the development of the vegetation cover and by its surface and sub-surface supply of organic matter. Therefore, the mesofaunal concentration in the upper soil layers may be taken as an indicator characteristic of mature forest soil stratification. In our restored plots, mesofaunal stratification was mainly favored by the treatments including planting (with or without sowing), whereas sowing or soil spreading resulted in most animals inhabiting the deeper soil layers. After soil restoration, the secondary succession process implies an increase in the structural diversity and energy availability of the ecosystem, which facilitate the development of higher trophic levels (Ferguson, 2001). The predators/decomposers ratio is then expected to be higher in forests than in grasslands or scrublands. On the other hand, since polyfagous taxa, with opportunistic feeding habits, are more tolerant to disturbance than predators, a decline in the predators/ decomposers ratio may be related to soil degradation (Kimberling et al., 2001). Contrary to what we expected, the index did not discriminate between restored and undisturbed plots, probably because this index may be useful in detecting soil degradation processes but not in evaluating the soil maturity state in a particular ecosystem. Diversity indices are considered insensitive when calculated for the soil mesofauna as a whole (Ekschmitt et al., 2003; Siepel and van de Bund, 1988; van Straalen, 1998) but their resolution may be improved when calculated for more homogeneous taxa. In our treatments, the maximal specific diversity for collembolans and oribatids was found in the reference forests, and sowing appeared as a promising treatment in restoring the oribatid diversity. The spreading treatment plots exhibited the lowest species diversity for both taxa. As biodiversity is dependent on the combination of multiple environmental and ecological factors, the interpretation of this index is difficult in terms of the success of restoration treatments. However, as in other studies, we found a significant correlation between biodiversity and some specific soil characteristics, such as soil organic matter content for oribatids and collembolans (Noble et al., 1996; Ponge et al., 2003), and soil pH for collembolans (Loranger et al., 2001; van Straalen and Verhoef, 1997). Comparison of the oribatid and collembolan community structure between treatments by means of rankspecies plots was not successful in our plots because of overlapping curves (Krüger and Scholtz, 1998). An alternative attempt by applying a Wilcoxon’s pairedranks test to the relative abundance of the oribatid and collembolan species suggested than the mixed treatment was the most adequate in order to promote the recovery of the community structure of both groups. Finally, as van Straalen (1998) and Addison et al. (2003) have suggested, the greatest data coherence was achieved by using multivariate analyses. As the solutions coming from the CCA and the DCA were different, and the species-environment correlation values in CCA were high, we inferred that the contribution of the treatments to the species variability was small (Ter Braak, 1986). But, as Gauch (1982) pointed out, an ordination diagram that explains only a low percentage of variance may be quite informative. In our case we may conclude that none of the treatments was successful in restoring a soil forest faunal community. Overall, two main conclusions may be extracted from this work, one concerning the most efficient mesofaunal indicators for post-mining restoration, and the other concerning the evaluation of the postmining restoration treatments. Even if they are relatively expensive in terms of classification effort, the multivariate statistical methods applied to the collembolan and oribatid community composition provide the most useful information in order to detect similarities or differences between the restored and the reference soils. Even so, additional indices, such as the collembolan or oribatid species diversity, directly related to the progressive organic matter accumulation in the upper soil layers, may be used as bioindicators of the forest soil recovery (Ponge et al., 2003). In our work, the soil mesofaunal trends suggested that 12 years after restoration, the areas treated by soil spreading and herbaceous species sowing are as herbaceous ecosystems, characterized by high densities of prostigmatid and astigmatid mites living mainly in the mineral soil layers. Where the new soil has been planted with shrubs and trees, the bioindicators suggest the beginnings of a soil forest structure, with animals P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 living mainly in the upper soil layers, and a high Cryptostigmata/Prostigmata ratio. When the treatment was sowing and planting trees over the new soil, the restored areas are in an intermediate stage between the forest and the herbaceous systems structure. The greatest gap between treated and undisturbed areas was found where the restored soil was not later vegetated. However, differences between the restored and the undisturbed areas are not easy to interpret in terms of ‘‘restoration success’’. As the direction and speed of the secondary succession depends on a multiplicity of factors (such as the pre-disturbance characteristics of the ecosystem, the colonization rate from the adjacent ecosystems, the physical disturbances of the biotope or the community endogenous processes) which are continuously evolving in time (McCook, 1994; Terradas, 2001), the description of the ecosystem variables at a given moment do not allow one to predict its future evolution. Some authors (Parmenter and MacMahon, 1990) have suggested that the pre-disturbance ecosystem recovery is not possible after mining. Our restored plots may become stabilized in the situation described, may progress through forest characteristics or may even regress to more simple stages. Monitoring in time is necessary to clarify such matters. Acknowledgment This work was supported for the CICyT-FEDER 2FD97-1644-C03-03 project. References Addison, J.A., Trofymow, J.A., Marshall, V.G., 2003. Abundances, species diversity and community structure of collembola in successional coastal temperate forests on Vancouver island. Can. Appl. Soil Ecol. 24, 233–246. Al-Assiuty, A.I.M., Khalil, M.A., Abdel-Lateif, H.M., 1999. Changes in oribatid faunal structure associated with land conversion from annual crop into orchard. Pedobiologia 43 (1), 85–96. Alcañiz, J.M., Comellas, L., Pujolà, M., 1996. Manual de restauració d’activitats extractives amb fangs de depuradora. In: de Sanejament, J. (Ed.), Recuperació de Terrenys Marginals. Generalitat de Catalunya, Barcelona, p. 69. Andrés, P., 1998. Décomposition de la matière organique dans le sol de deux écosystèmes forestiers (Parc Naturel du Montseny, Espagne). I. Abondance, biomasse et distribution des microarthropodes. Vie Milieu 48, 215–225. Andrés, P., 1999. Ecological risks of the use of sewage sludge as fertilizer in soil restoration: effects on the soil arthropods populations. Land Degrad. Dev. 10, 67–77. Andrés, P., Mateos, E., Ascaso, C., 1999. Soil arthropods. In: Ferran Rodà, et al. (Eds.), Ecology of Mediterranean Evergreen Oak Forests. Springer-Verlag, Berlin, pp. 341–354. 77 Ash, H.J., Gemmell, R.P., Bradshaw, A.D., 1994. The introduction of native plant species on industrial waste heaps: a test of immigration and other factors affecting primary succession. J. Appl. Ecol. 31, 74–84. Berch, S.M., Baumbrough, B., Battigelli, J., Kroeger, P., Strub, N, de Montigny, L., 2001. Preliminary Assessment of Selected Communities of Soil Organisms under Different Conifer Species. British Columbia Ministry of Forests Research Program. Research Report 20, 27 pp. Bezdicek, D.F., 1996. Development and evaluation of indicators for agroecosystems health. Agriculture in Concert with the Environment ACE Research Projects Western Region, 1991–1995, 6 pp. Black, H.I.J., Parekh, N.R., Chaplow, J.S., Monson, F., Watkins, J., Creamer, R., Potter, E.D., Poskitt, J.M., Rowland, P., Ainsworth, G., Hornung, M., 2003. Assessing soil biodiversity across Great Britain: national trends in the occurrence of heterotrophic bacteria and invertebrates in soil. J. Environ. Manage. 67, 255–266. Bonet, A., Pausas, J.G., 2004. Species richness and cover along a 60year chronosequence in old-fields of southeastern Spain. Plant Ecol. 174, 257–270. Bradshaw, A., 1997. Restoration of mined lands—using natural processes. Ecol. Eng. 8, 255–269. Cepeda-Pizarro, J.G., Withford, W.G., 1989. The relationship between abiotic factors and the abundance patterns of soil microarthropods on a desert watershed. Pedobiologia 33, 79–86. Clapperton, M.J., Kanashiro, D.A., Behan-Pelletier, V.M., 2002. Changes in abundance and diversity of microarthropods associated with Fescue Prairie grazing regimes. Pedobiologia 46 (5), 496– 511. David, J.F., Devernay, S., Loucougaray, G., Le Folc’h, E., 1999. Belowground diversity in a Mediterranean landscape: relationships between saprophagous macroarthropod communities and vegetation structure. Biodivers. Conserv. 8, 753–767. Doran, J.W., Zeiss, M.R., 2000. Soil health and sustainability: managing the biotic component of soil quality. Appl. Soil Ecol. 15, 3– 11. Ekschmitt, K., Stierhof, T., Dauber, J., Kreimes, K., Wolters, V., 2003. On the quality of soil diversity indicators: abiotic and biotic parameters as predictors of soil faunal richness at different spatial scales. Agric. Ecosyst. Environ. 98, 183–273. Ferguson, S.H., 2001. Changes in trophic abundance of soil arthropods along a grass-shrub-forest gradient. Can. J. Zool. 79, 457–464. Gauch, H.G., 1982. Multivariate Analysis in Community Ecology. Cambridge University Press, Cambridge, p. 298. Green, J.J., Vallejo, R., Serrasolses, I., Martins-Loucao, M.A., Hatzistathis, A., Barea, J.M., Cortina, J., Watson, C.A., 1999. Restoration of degraded ecosystems in Mediterranean regions (REDMED). In: Papanastasis, V.P., et al. (Eds.), Grasslands and Woody Plants in Europe, vol. 4. Grassland Science in Europe, Herpas, Thessaloniki. Greenslade, P., Majer, J.D., 1993. Recolonization by collembola of rehabilitated bauxite mines in western-Australia. Aust. J. Ecol. 18 (4), 385–394. Hodkinson, I.D., Jackson, J.K., 2005. Terrestrial and aquatic invertebrates as bioindicators for environmental monitoring, with particular reference to mountain ecosystems. Environ. Manage. 35 (5), 649–666. Jorba, M., Andrés, P., 2000. Effects of sewage sludge on the establishment of the herbaceous ground cover alter soil restoration. J. Soil Water Con. 55 (3), 322–327. Kallimanis, A.S., Argyropoulou, M.D., Sgardelis, S.P., 2002. Two scale patterns of spatial distribution of oribatid mites (Acari, 78 P. Andrés, E. Mateos / Applied Soil Ecology 33 (2006) 67–78 Cryptostigmata) in a Greek mountain. Pedobiologia 46 (6), 513– 525. Kethley, J., 1990. Acarina: Prostigmata (Actinedida). In: Dindal, D.L. (Ed.), Soil Biology Guide. Willey, New York, pp. 667–756. Kimberling, D.N., Karr, J.R., Fore, L.S., 2001. Measuring human disturbance using terrestrial invertebrates in the shrub-steppe of eastern Washington (USA). Ecol. Indic. 1, 63–81. Koehler, H., 1998. Secondary succession of soil mesofauna: A 13 year study. Appl. Soil Ecol. 9, 81–86. Kremen, C., Collwell, R.K., Erwin, T.L., Murphy, D.D., Noss, R.F., Sanjayan, M.A., 1993. Terrestrial arthropods assemblages: their use in conservation planning. Conserv. Biol. 7, 796–808. Krüger, K., Scholtz, C.H., 1998. Changes in the structure of dung insect communities after ivermectin usage in a grassland ecosystem. I. Impact of ivermectin under drought conditions. Acta Oecol. 19 (5), 425–438. Longcore, T., 2003. Terrestrial arthropods as indicators of ecological restoration success in coastal sage scrub (California, USA). Restor. Ecol. 11 (4), 397–409. Loranger, G., Bandyopadhyaya, I., Razaka, B., Ponge, J.F., 2001. Does soil acidity explain altitudinal sequences in collembolan communities? Soil Biol. Biochem. 33, 381–393. Ludwig, J.A., Hindley, N., Barnett, G., 2003. Indicators for monitoring minesite rehabilitation: trends on waste-rock dumps, northern Australia. Ecol. Indic. 3, 143–153. McCook, I.J., 1994. Understanding ecological community succession: causal models and theories, a review. Vegetatio 110, 115–147. Moreno, R., 2000. Aplicación de lodos de depuradora en la restauración de canteras: efectos sobre la comunidad vegetal. Trabajo de investigación de Doctorado. Universitat Autònoma de Barcelona, CREAF, Bellaterra, unpublished. Neumann, F.G., 1991. Responses of litter arthropods to major natural or artificial ecological disturbances in mountain ash forests. Aust. J. Ecol. 1, 19–32. Noble, J.C., Withford, W.G., Kaliszweski, M., 1996. Soil and litter microarthropods populations from two contrasting ecosystems in semi-arid Western Australia. J. Arid Envir. 32, 329–346. Ojeda, G., Alcañiz, J.M., Ortiz, O., Tarrasón, D., 2003. Escorrentı́a y erosión en dos suelos fertilizados con lodos de depuradora. In: Actas del I Simposio Nacional Sobre Control de la Erosión y Degradación del Suelo, IMIA, Madrid, July 9–11, pp. 271–274. Ortiz, O., 1998. Activitat biològica i revegetació en sòls de pedreres de calcària restaurats amb fangs de depuradora. PhD Thesis. Universitat Autònoma de Barcelona, Bellaterra. Barcelona, unpublished. Parmenter, R.R., MacMahon, J.A., 1990. Faunal community development on disturbed lands: an indicator of reclamation success. In: Chambers, J.C., Wade, G.L., (Eds). Evaluating reclamation success: the ecological consideration. USDA, Forest Service, Northeastern Forest Station, Radnor, Pennsylvania, pp. 73–89. Philips, J.R., 1990. Acarina: astigmata (Acaridida). In: Dindal, D.L. (Ed.), Soil Biology Guide. Wiley, New York, pp. 757–778. Ponge, J.F., Gillet, S., Dubs, F., Federoff, E., Haese, L., Sousa, J.P., Lavelle, P., 2003. Collembolan communities as bioindicators of land use intensification. Soil Biol. Biochem. 35, 813–826. Ruf, A., 1998. A maturity index for predatory soil mites (Mesostigmata: Gamasina) as an indicator of environmental impacts of pollution on forest soils. Appl. Soil Ecol. 9, 447–452. Schloter, M., Dilly, O., Munich, J.C., 2003. Indicators for evaluating soil quality. Agric. Ecosyst. Environ. 98, 255–262. Sgardelis, S.P., Usher, M.B., 1994. Responses of soil Cryptostigmata across a boundary between a farm woodland and an arable field. Pedobiologia 38, 36–49. Shopper, W.E., 1993. Municipal Sludge Use in Land Reclamation. Lewis Publishers, Florida, p. 163. Siepel, H., van de Bund, C.F., 1988. The influence of management practices on the microarthropods community of grasslands. Pedobiologia 31, 339–354. Sort, X., Alcañiz, J.M., 1996. Contribution of sewage sludge to erosion control in the rehabilitation of limestone quarries. Land Degrad. Dev. 7, 69–76. Sort, X., Alcañiz, J.M., 1999a. Modification of soil porosity alters application of sewage sludge. Soil Till. Res. 49, 337–345. Sort, X., Alcañiz, J.M., 1999b. Effects of sewage sludge amendment on soil aggregation. Land Degrad. Dev. 10, 3–12. St. John, M.G., Bagatto, G., Behan-Pelletier, V., Lindquist, E.E., Shorthouse, J.D., Smith, I.M., 2002. Mite (Acari) colonization of vegetated mine tailings near Sudbury, Ontario, Canada. Plant Soil 245, 295–305. Ter Braak, C.J.F., 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67 (5), 1167–1179. Terradas, J., 2001. Ecologı́a de la vegetación. De la ecofisiologı́a de las plantas a la dinámica de las comunidades y paisajes. Omega, Barcelona, 703 pp. UB/CEIB-ESAB/CREAF, 2001. Restauración de áreas afectadas por minerı́a en Cataluña. CICyT-FEDER 2FD97-1644-C03-03, unpublished report, 103 pp. Vallejo, V.R., Bautista, S., Cortina, J., 1999. Restoration for soil protection after disturbances. In: Trabaud, L., (Ed.), Life and Environment in the Mediterranean. Adv. Ecol. Sci. 3. WIT Press, pp. 301–343. van Bruggen, A.H.C., Semenov, A.M., 2000. In search of biological indicators for soil health and disease suspension. Appl. Soil Ecol. 15, 13–24. van Straalen, N.M., 1998. Evaluation of bioindicator systems derived from soil arthropod communities. Appl. Soil Ecol. 9, 429–437. van Straalen, N.M., Verhoef, H.A., 1997. The development of a bioindicator system for soil acidity based on arthropod pH preferences. J. Appl. Ecol. 34, 217–232. Wanner, M., Dunger, W., 2002. Primary immigration and succession of soil organisms on reclaimed opencast coal mining areas in eastern Germany. Eur. J. Soil Biol. 38, 137–143. Webb, N.R., 1994. Postfire succession of Cryptostigmatic mites (Acari, Cryptostigmata) in a calluna-heathland soil. Pedobiologia 38 (2), 138–145. Zas, R., Alonso, M., 2002. Understory vegetation as indicator of soil characteristics in Northwest Spain. Forest Ecol. Manag. 171, 101– 111.