1 Introduction

Areas of the past ecological damage, or brownfields, represent one of the most relevant ecological problems; remediation of such lands is growing more and more important. Brownfields are decommissioned lands, previously used as industrial waste sites, landfills, and dumps, as well as other contaminated areas that were previously in use. The disturbance loads and scopes of the brownfields vary from rare global- and nation-scale ecological disaster areas, to common regional and local long-term waste deposit sites. Conservation agencies of many countries fight the effects of these areas on the surrounding ecosystems, and such efforts often gain state-level priority. Supporting measures include compiling of the inventories and data bases of the brownfields, flexible remediation rules determine the sites suitable for various functions, remediation measures themselves, as well as the estimation instruments, are constantly developed. In the Russian Federation, the effects of brownfields are very pronounced, and liquidation of these sites is an acute problem (Saraev et al. 2015). During the last two decades, brownfields became the veritable source of ecological damage, representing danger to the health of the people living nearby. In Russia, the problem is now being addressed by a specialized state programme aimed at mitigation of the past environmental damage.

As the primary pollution by heavy metals is decreasing globally, the brownfield sites are becoming the major source of secondary pollution (Vodyanitskii 2013). It is well-known that the upper soils are particularly susceptible to the toxic contamination and accumulation of contaminants (Wolterbeek and Verburg 2001; Baran et al. 2014), and heavy metals are the main pollutants of the brownfield soils.

There are legally defined maximum permissible concentrations (MPCs) that specify the levels of pollutants in soils. At brownfield, as well as other sites, chemical tests are traditionally used to measure such concentrations (Remon et al. 2005; Hu et al. 2013). However, the list of pollutants and methods for their detection, specified in normative regulations, is not comprehensive, which undermines the accuracy of the environmental assessment (Maxam et al. 2000; Ahtiainen et al. 2002). MPCs are country-specific (e.g., SANPIN 2.1.7.1287-03; GN 2.1.7.2041-06 (2006); GN 2.1.7.2511-09 (2009). are only used in the Russian Federation); a unified, internationally accepted levels are not, as yet, developed. Additionally, existing soil quality standards and MPCs are controversial in Russia (Kapelkina 2013): they consider neither the soil type and resilience nor intended land use. For instance, MPCs are the same for agricultural and industrial soils, and thus may differ from levels defined in other countries by one or two orders of magnitude. Public health validation of the permissible concentrations of the chemicals derived from four experimentally defined parameters, i.e., their ability to enter (i) the plant tissue, (ii) ground water, (iii) the atmosphere, and (iv) a generalized sanitary hazard index (the effects of the chemical on the self-purification and biological activity of the soil). The smallest of these four levels is set as a minimum permissible concentration. For some chemicals, only provisional, approximate permissible concentrations (APCs) exist, derived from pH and particle-size composition of the soil. Both measures are unsuitable for industrial lands. Useful regulatory standards have to be based on the influence of the pollutants on the public health. Additionally, present MPCs and APCs do not consider the age of contamination, although it is known that, e.g., mobility of the heavy metals falls with time, thus reducing their hazard level (Vodyanitskii 2013).

Another measure of the soil quality is an integrated contamination index Zc (Saet et al. 1990), which aggregates pollution levels of several individual heavy metals. This index provides an easy-to-interprete, clear-cut instrument for classification of soil contamination levels compared to the background. Value ranges of this index delineate five such categories: uncontaminated soils, permissible contamination level, moderately hazardous, hazardous, and extremely hazardous contamination (SANPIN 2.1.7.1287–03). This index, although widely used even in the state-level ecological regulations, is far from satisfactory. Its main problem lies in sensitivity to the number of the chemical elements used in the analysis: as this number increases, so does the resulting value of Zc (Smagin 2013).

A modern approach to ecological evaluation of the soil quality should take into account biotic parameters. Hence, many researchers call for interdisciplinary approach to brownfield assessment that involves chemical testing and toxicological and bioassay methods (Linkov et al. 2006; Alvarenga et al. 2012; Ribe et al. 2012; Feng et al. 2016; Voronich et al. 2016; Filenko 2007).

Bioassay is a laboratory method for determining the quality of environmental entities with the use of test organisms. The unique advantage of bioassay is its ability to register the total toxic impact of multiple pollutants on the living organism (Terekhova 2011; Olkova 2014). For instance, several chemical pollutants may be present in concentrations below permitted levels, but still have negative effect on the ecosystem, and bioassay methods can reveal this integral influence much better than chemical testing (van Gestel et al. 2001; Lors et al. 2011). Besides, bioassay may function as a part of the TRIAD diagnistic complex (Chapman Dagnino et al. 2008; Terekhova et al. 2014). Another benefit of the bioassay method lies in its ability to predict upcoming negative changes in the ecosystem before they take effect.

In their existing form, bioassays already form a reliable source of the integral assessment of brownfields, and deserve further development as a part of environmental monitoring for the contaminated sites (Bardina et al. 2014a, b, c, 2016). Still, as they gain popularity in modern ecological studies, biotesting methods encounter diverse demands. They need to be accurate, quick, cost-effective, and produce repeatable, reproducible results (Filenko and Terekhova 2016). To ensure the latter, one must employ certified test cultures and procedures.

Existing bioassays predominantly analyze soil eluate (i.e., aqueous extract), while contact methods are in great demand. Additionally, reliable transition is needed from the results of acute toxicological test to predicting the outcome of more prolonged, and costly (Solomon et al. 2008), chronic experiment.

The success rate of any particular variant of a bioassay closely depends on local conditions, including the landscape features, mesoclimate, soil, and geochemistry, as well as specific traits of the test objects. On the other hand, no single species can be a universal indicator of the environmental health of a soil (Broos et al. 2005; Foucault et al. 2013). Several species from various trophic levels: producers, consumers, and decomposers, should be used for comprehensive environmental assessment. Thus, important and highly relevant areas of research include the following: (1) identifying the applicability of various bioassay systems for specific objects, (2) regional geo-environmental adaptation of bioassay methods, and (3) developing comprehensive panels of the biotests to be used together (Manzo et al. 2008; Matejczyk et al. 2011).

The aim of the present study is to test the usefulness of several dissimilar bioassay methods on four brownfield sites, contaminated by various waste types. The investigation objectives are to determine the physical-chemical and chemical parameters of the soil samples from various brownfield sites; to apply various acute eluate and contact soil bioassays, in order to compare the resulting estimates; and to evaluate the integral sensitivity of applied set of test systems for the soils from the areas suffering the long-term storage of the solid wastes.

2 Material аnd Methods

2.1 Study Sites and Sampling

The study was carried out at four brownfield sites in North-West Russia. These sites are located in four different geomorphological subareas, each with its own characteristic soil type. All four sites suffered ecological damage in the near past, each site being contaminated by different combinations of toxic substances.

Site 1 is an area about 6 ha, located in the Neva lowland near the coastal zone of Lake Ladoga (59° 43′ 44″ N 31° 36′ 59″ E, Fig. 1), with Podzolics surfacely-gleyic (Luvic Stagnosols Dystic of WRB, 2006) soils typical for the locality. Hydrological conditions of the terrain include a weak surface runoff and significant groundwater infiltration. The site itself was used for storage of industrial waste, municipal solid waste, wood processing waste, agricultural waste and construction debris between 1980 and 2005. The layer of topsoil at the site has been partially destroyed. The area was covered with intermittent shrubbery and scattered small-volume dumps of the mixed waste. We took five soil subsamples from one 10m2 plot from the depth of 0–20 cm and joined them to form a single mixed sample.

Fig. 1
figure 1

Map of the study sites (S1, S2, S3, S4). Scale 1:1,500,000

Full size image

Site 2 is an 8-ha area, located near Lake Lublinskoye (60° 20′ 11″ N 29° 53′ 59″ E). Surrounding terrain is terraced, gently rolling, with sandy Podzolics illuvial-ferrugenous (Carbic Podzols) soils. Hydrological conditions of the terrain include an underground runoff to the lake. The site has been used as mixed waste deposit in 1990s. The area was covered with scattered small-volume dumps of the mixed waste. We took five subsamples of soil from one 10-m2 plot from the depth of 0–20 cm, and joined them to obtain one mixed sample.

Site 3 is a brownfield of 150 ha, lying on the bank of Voronka River (59° 43′ 12″ N 29° 18′ 23″ E). The surrounding landscape is an Ordovic plato, with developed limestone karst processes and Sod-calcareouses (Redzic leptosols eutric) loam soil. Hydrological conditions are characterized by a weak surface runoff in the direction of the nearby swamp and river, as well as a significant groundwater infiltration. The study site is contaminated by construction debris and industrial and mixed wastes, which were stored for 30 years, between 1970 and 2000. We took a total of five samples from each of two 10-m2 plots with different contaminants: (1) industrial waste and (2) mixed waste. Samples were taken separately from 0 to 5 cm and 5 to 20 cm (SANPIN 2003).

Finally, site 4 is a 6.7-ha industrial waste dump and surrounding territory, located near lake Ladoga on the bank of the Neva River (59° 40′ 22″ N 31° 01′ 04″ E). Typical landscapes of the locality are gently rolling hills, with Sod-podzolics illuvial-ferrugenous (Umbric Albeluvisols Abruptic) soils on higher grounds and Podzols illuvial-ferrugenous (Carbic Podzol) soils in depressions. Hydrological conditions of the terrain include a noticeable lateral and vertical migration of water. The site of the dump was used for long-term unregulated storage of industrial wastes of sulfuric acid production. They have a complex composition, with prevalence of iron and other metals typical for the sulfide iron ores, and also include burned pyrite, industrial slag and secondary metabolites. The depositing started in 1965 and stopped in 1978, but the existing waste was never removed. Surrounding land was also contaminated over the years because of both surface runoff and groundwater infiltration of dissolved pollutants (Saraev et al. 2015). For sampling, we selected five 10-m2 plots, located about 10 m away from the edge of the existing dump, in the south (1), south-east (2), east (3), north (4) and south-west (5) of the contaminated site 4. Again, we delineated two separate sampling depths: 0–5 cm and 5–20 cm. From each depth, five subsamples were taken within one plot and joined in a single mixed sample.

2.2 Physical-Chemical and Chemical Methods

We employed a range of physical-chemical and chemical tests, including measurements of pH and conductivity of the soil extracts. Heavy metals, arcenic, and phosphorous were determined by inductively coupled plasma mass spectrometry. Concentration of polychlorinated biphenyls, oil products, and chlorinated pesticides (HCH, DDT) was measured using gas chromatographic method with electron-capture detector (GC-2010, Shinadzu, Japan). Concentration of 3,4-benzo(a)pyrene was measured using high-performance liquid chromatography with fluorescence detection method (Fluorat-02, Russia). To estimate the integrative level of chemical contamination, we calculated the total index of pollution Zc:

$$ \mathrm{Zc}=\sum {C}_{\mathrm{s}}/{\mathrm{C}}_{\mathrm{b}}-\left(\mathrm{n}-1\right), $$

where C s is the concentration of a chemical substance in the sample (mg/kg), C b is the regional background concentration (mg/kg), and n is the number of substances. Zc values below 16 indicate permissible contamination level, values in the range between 16 and 32 – moderately hazardous, values from 32 to 128 – hazardous and and Zc values above 128 indicate extremely hazardous contamination (SANPIN 2003).

2.3 Eluate Bioassay Methods

There are two major approaches to biotesting of the soil toxicity. The first uses the aqueous extract, or eluate, of the soil sample, with various aquatic organisms (algae, ciliates, daphnia, etc.) as test organisms. The second approach is contact bioassay, where the test organism contacts directly with the untreated test sample. This type of analysis helps to establish the level of impact of solid pollutants (Voronich et al. 2016; Bardina et al. 2016; Selivanovskaya and Galitskaya 2006). Here, both types of biotests were used, and final soil toxicity was estimated based on the most sensitive result.

We prepared eluates by mixing 10 parts of distilled water with 1 part of the sampled soil for 24 h to allow complete extraction. These were used in three types of bioassays: with ciliates Paramecium caudatum Ehrenberg (consumers), green algae Scenedesmus quadricauda (Turp.) Brebisson and seeds of common oats Avena sativa L. (producers). Daphnia were not employed, since they are intolerant to the acidic reaction, pronounced in all sample extracts from site 4. We did not attempt to neutralize the eluate, since this process alters the chemical composition of the sample and influence the measurements of toxicity (Filenko 2007).

We chose P. caudatum as a test species because these protists are sensitive to heavy metals (Eriksen 1990) and respond to the presence of dangerous substances by directional movement across the concentration gradient of these substances (chemotactic response). At the same time, Paramecium demonstrates negative geotaxis. When placed in a vertical test tube, these ciliates aggregate in the upper portions of the medium. This tendency is used in a biotest method, where a thickened medium with ciliate culture is placed at the bottom of the test tube, and examined eluate is poured on top (PND 2010). A stable interface boundary, which forms between the liquids, does not block ciliate movements between upper and lower zones. After 30 min, the ciliates redistribute themselves between two zones. The higher the toxicity of the sample, the smaller the proportion of ciliates in the upper zone of the test tube. We measured the concentration of the ciliates (cells/ml) and determined the sample toxicity index (T) with the following equation:

$$ T=\left({I}_{\mathrm{c}}-{I}_{\mathrm{s}}\right)/{I}_{\mathrm{c}} $$

where I c is the mean of the ciliate concentration in the control and I s—in the analyzed sample.

Samples were classified into three groups of toxicity: (1) low (at 0.0 < Т ≤ 0.4), (2) moderate (0.4 < Т ≤ 0.7), and (3) high degree of toxicity (Т > 0.7). Sometimes, samples may contain harmless substances that are attractive for ciliates. In such cases, I s may even exceed I c, thus giving negative values of toxicity index. These results indicated the absence of toxicity and were treated as zeros.

To increase the sensitivity of detection, we prepared two subsamples: the first was used as is; the second was diluted with distilled water 100 times. The solutions were tested, and dilution rate, at which the toxic effects of the extract disappeared, was used to judge the sample quality.

The use of S. quadricauda in biotesting is based on the sensitivity of its growth rate. We used a standard protocol (Grigoriev and Tyutkova 2011), where two populations of algae were grown in a culture medium based on the eluate and on distilled water. After an interval of 45 h, their optical densities were measured and compared using the formula:

$$ I=\left[\left({X}_{\mathrm{c}}-{X}_{\mathrm{s}}\right)/{X}_{\mathrm{c}}\right]\cdotp 100\%, $$

where X s and X c are the average values of optical density in the sample and control, respectively. A growth rate decrease of 20% or more (growth suppression) and increase by 30% or more (growth stimulation), were used as threshold values of acute toxicity.

In all tests we used pure cultures of the Scenedesmus that were in the exponential phase of growth. To distinguish between toxicity levels of the soil extracts, we prepared three subsamples: the undiluted original eluate, and extracts diluted 10 and 100 times. All these were tested, and dilution rate, at which the toxic effects of the extract disappeared, was used to judge the sample quality. Samples that only exhibited toxicity in undiluted state were designated “weakly toxic.” Samples that remained toxic when diluted 10 times but not 100 times were identified simply as “toxic.” Finally, samples that remained toxic even if diluted 100 times were marked “highly toxic” (Grigoriev and Tyutkova 2011).

Finally, we used the seeds of common oat (A. sativa) as the test objects since phytotesting also proved useful in assessment of ecological conditions of soils (Terekhova et al. 2016). We planted 25 oat seeds in Petri dishes, added either distilled water (control), or undiluted soil extract, and placed the dishes into climatic chamber (20–23 °C). After 5 days, the root length of the seedlings was measured and compared; each sample was tested in test replications. We determined the sample toxicity index (E) with the following equation:

$$ E=\left[\left({X}_{\mathrm{c}}-{X}_{\mathrm{s}}\right)/{X}_{\mathrm{c}}\right]\cdotp 100\%, $$

where X s and X c are the average values of root length of the seedlings in the sample and control, respectively. Eluate was judged toxic if the root length of the seedlings in the sample exceeded that of oats grown in the control by more than 20% (MR 2007).

2.4 Contact Bioassay Methods

The main advantage of contact bioassay lies in revealing the effect of solid pollutants as opposed to the dissolved ones (Terekhova et al. 2016). Methods employing seeds of the higher plants, as well as natural microbial communities, were suggested and developed in the recent years. Examples include international standards for plants ISO 11269-1, ISO 11269-2 (Fomin and Fomin 2001) as well as national standards (Kapelkina et al. 2009). The latter standard recommends barley (Hordeum vulgare L.) and common wheat (Triticum aestivum) as test cultures.

In the present study, we used the seeds of wheat, which is better adapted to the regional climate. Twenty seeds were placed into Petri dishes on the surface of soil sample (1-cm thick), moistened to 60% of moisture-holding capacity and allowed to germinate and grow for 5 days at the same humidity. For control, we planted the same number of seeds into clean, uncontaminated soils, similar in humus content and granulometric composition to each test sample. For each sample, we performed four replications of the test. In each case, we measured two response variables: the rate of germination and the length of the roots, which are the most sensitive parameters in phytotesting (Wang et al. 2001). We compared sample and control measurements of germination rate (N 1) and root length (N 2) separately, using the formula:

$$ N=\left({M}_{\mathrm{c}}-{M}_{\mathrm{s}}\right)\cdotp 100/{M}_{\mathrm{c}}, $$

where М c is the average value under control conditions and M s is the average value of test sample. The degree of sample toxicity was judged by suppression of germination and reduction of the root length: V—practically non-toxic (0 < N 1 ≤ 20 and 0 < N 2 ≤ 20); IV—slightly toxic (0 < N 1 ≤ 20 and 20 < N 2 ≤ 50); III—moderately toxic (20 < N 1 ≤ 70 and 50 < N 2 ≤ 70); II—highly toxic (70 < N 1 < 100 and 70 < N 2 < 100); I—extremely toxic (N 1 = 100 and N 2 = 100).

The choice of controls is extremely significant in contact bioassays (Terekhova et al. 2016). To establish the background levels for comparison, one has to select the soils, which were not affected by anthropogenic activities, and which closely resemble the samples from the brownfields in physical and physical-chemical characteristics. To ensure this, we used control soil quantities, individually matched to samples from each of the tested sites.

Microorganisms are also sensitive indicators of the biological status of soils (Vodyanitskii 2013; Voronich et al. 2016). We used the natural community of microorganisms, present directly in the studied soils, for the second contact bioassay. To determine toxicity of each sample, we estimated biological activity of microbial community as manifested by soil respiration, which is one of the most important indicators of destruction processes caused by microorganisms (Anderson et al. 2011). Soil respiration was deduced from the intensity of carbon dioxide release, measured by modified adsorption method (Alef 1995). For this test, we used samples taken to the lab immediately upon collection (after no more than 48 h). Moist soil sample, equivalent to 1 g of soil dehuminified at 104 °C, was placed into an airtight glass jar for 24 h. CO2, released during the incubation, was adsorbed in NaOH (0.02 mol/L), which was titrated with H2SO4 (0.01 mol/L). The jars with NaOH, but without the soil, were used to measure CO2 levels in the air itself. The rate of the respiration was calculated in milligrams CO2/ 100 g of dry soil/ 24 h. Uncontaminated soils (control), similar in humus content and granulometric composition to each test sample, were treated in the same way to obtain CO2 release readings. Each test was repeated in four replications.

2.5 Statistics

To estimate central tendencies and range of measured biological characteristics, we calculated arithmetic means and standard errors (sample size n always equaled 4). To compare samples and controls, we used one-way analysis of variance (ANOVA) and Tukey’s post hoc test. Differences were considered significant at p < 0.05. Data analysis was performed with Statistica 10.0.

3 Results and Discussion

3.1 Physico-Chemical and Chemical Parameters of Soils

Soils at sites 1, 2, and 3 had close to neutral and neutral reaction (pH ranged from 5.9 to 6.9). Alkalization of these soils was caused by anthropogenic pollution. At site 4, soil had a more pronounced acidic reaction (pH 3.9–5.4) because of the acidic surface runoff from the waste dump and migration of the products of sulfur oxidation. The conductivity of all samples ranged from 0.05 to 0.12 mS/cm, indicating the absence of salination.

Some pollutants were common to all sites, e.g., all examined soils a had higher content of total phosphorus (428–699 mg/kg) compared to that of the background (302 mg/kg), due to anthropogenic pollution. Heavy metals were also ubiquitous toxicants (Table 1).

Table 1 Concentrations of organic and inorganic contaminats in the soils
Full size table

Heavy metals typically settle in the most fertile humic soil layer (0–20 cm), which also defines the crop ranges (Minkina et al. 2010). We expected that soils from different sites would be unequal in the levels of contamination with heavy metals. This is why measurements of Zc index were included into our study.

As expected, soils from various study sites differed in the level of contamination with heavy metals. Soil at the site 1 had the highest total index of pollution Zc of 65.5, and thus contamination of the area was classified as hazardous. Substances of hazard categories 1 and 2 (Pb, Cd, Zn, and Cu, respectively) were present at levels, exceeding maximum permissible concentration or tentative allowable concentration. Among organic pollutants, concentration of polychlorinated biphenyls was above MPC (0.16 mg/kg against 0.06 mg/kg).

At site 2, we found a low value of Zc (7.4), which placed the total heavy metal pollution of this soil into permissible category. Still, three elements: Pb, Cd and Zn, all from hazard category 1, slightly exceeded maximum or tentative permissible concentration. Polychlorinated biphenyls, at a level somewhat above threshold, were also found at site 2 (0.072 mg/kg).

Total heavy metal concentration at site 3 was judged permissible (Zc = 7.3). The main pollutants, however, were organic substances: 3,4 benzo(a)pyrene (0.085 mg/kg against MPC 0.02 mg/kg) and polychlorinated biphenyls (0.20 mg/kg against MPC 0.06 mg/kg). These substances are extremely toxic even in small concentrations and cause cell death.

Soil samples from site 4 had a Zc value of 29.9, or moderately hazardous total chemical pollution level. Among individual heavy metals, Pb (hazard category 1), Zn (hazard category 1), Cu (hazard category 2), and As (hazard category 1) were especially pronounced. The levels of organic pollutants did not exceed the established Russian standards. The concentration of oil products was below the background level of 180 mg/kg in all four study sites. Site 4 is located on the outskirts of an existing landfill, storing industrial waste. The landfill itself is a source of extreme environmental hazard, as revealed by a previous chemical study of the area: the landfill soil had high level of total heavy metal pollution (Zc = 607) (Bardina et al. 2014a, b, c). Aside from this, complex biochemical processes of waste decomposition take place over long periods of time, accompanied by formation of toxic organic and inorganic compounds. As a result, specific anthropogenic soils, sometimes extremely hazardous and toxic, are formed in the landfill body. Strong contamination of the surrounding territories (exemplified by site 4) happens due to sliding of the substrate from the landfill and, perhaps, other migration mechanisms.

Thus, chemical testing revealed strong heavy metal pollution and mild organic pollution at site 1, no total contamination by heavy metals at sites 2 and 3 (although individual element might have exceeded the limits) and significant presence of organic toxicants at site 3, and notable contamination with heavy metals but no organic pollution at site 4. Pb, Zn, Cr, Cu, Mn, and V demonstrated the most significant excess over their regional geochemical baseline at all sites.

3.2 Ecotoxicological Evaluation of Soils with Eluate Tests

The choice of an appropriate test organism for a bioassay, and the resulting usefulness of the procedure, largely depend on the chemical composition of the studied samples (Filenko and Terekhova 2016). Thus, we kept in mind the varying sensitivity of hydrobionts while choosing the organisms used in the tests. One of the most common species, used to evaluate aqueous solutions is a crustacean Daphnia magna straus 1820. They are easily cultured and demonstrate high sensitivity to various pollutants (Terekhova 2011). Unfortunately, these animals are best used for testing the solutions with рН about 7.0–8.3 (ISO 6341: 2012). Since soil samples from site 4 were overly acidic, we could not employ Daphnia directly. Using additional substances to bring the pH to the recommended range was also deemed undesirable, as that would necessarily affect the chemical composition of the sample (Terekhova 2011). Therefore, we chose other test species: P. caudatum, S. quadricauda and A. sativa L.

Results of bioassay with P. caudatum as test organism are given in Figs. 2 and 3. Soil extracts from site 1 were moderately toxic (effect disappeared at ×100 dilution). Samples from site 2 (Fig. 2a) and the upper soil horizon (0–5 cm) of site 3 (Fig. 2b) were not toxic to ciliates. Samples from 5 to 20 cm were found toxic, with negative effect on ciliates disappearing only at ×100 dilution (Fig. 2c). Finally, the soils at site 4 also had pronounced toxic effect, which disappeared only at ×100 dilution (Fig. 3). Biotests using P. caudatum were among the quickest and least expensive, although very sensitive to heavy metal contamination (Terekhova 2011).

Fig. 2
figure 2

Toxicity index of soil extracts, measured with Paramecium caudatum biotest. a Sites 1 and 2 (depth 0–20 cm). b Site 3 (depth 0–5 cm). c Site 3 (depth 5–20 cm)

Full size image
Fig. 3
figure 3

Toxicity index of soil extracts from site 4, measured with Paramecium caudatum biotest. a Samples from depth 0–5 cm. b Samples from 5 to 20 cm

Full size image

The results of biotesting of soil extracts using green algae S. quadricauda are presented on Figs. 4 and 5. The sample from site 1 had low toxicity, which disappeared at ×10 dilution (Fig. 4a). Soil from site 2 was non-toxic, while site 3 had mixed status. Samples from plot 1 were not toxic to algae, while plot 2 demonstrated no effect at depths 0–5 cm (Fig. 4b) and a pronounced effect at greater depths (5–20 cm), disappearing only with ×100 dilution (Fig. 4c).

Fig. 4
figure 4

Growth inhibition (I) of green algae Scenedesmus quadricauda. a Samples from sites 1 and 2 (0–20 cm). b Samples from site 3 (0–5 cm). c Samples from site 3 (5–20 cm)

Full size image
Fig. 5
figure 5

Growth inhibition (I) of green algae Scenedesmus quadricauda in soil extracts from site 4. a Samples from depth 0–5 cm. b Samples from 5 to 20 cm

Full size image

At site 4, all of the examined soil extracts had an acute toxic effect on Scenedesmus; there were no differences between the toxicity of the horizons 0–5 cm (Fig. 5a) and 5–20 cm (Fig. 5b) (p values > 0.1). Both undiluted and diluted 10 times eluates from plots 1, 2 and 3 completely suppressed algal growth (I = 100%). With a further dilution of extracts (×100) toxic effect disappeared. At site 4, soil extracts from plot 4 had the least inhibitory effect on the algae: undiluted eluate suppressed population growth by 27%, and the toxic effect disappeared at ×10 dilution. Finally, extracts from plot 5 had notable toxic effect: undiluted eluate completely inhibited the growth of algae, and the toxic effect was still present in ×10 dilution (but not in ×100).

Biotests with seeds of common oat A. sativa L revealed no toxic effect at sites 1 and 2 (Table 2), while sites 3 and 4 had reduced the root growth in seedlings above the critical level. However, no dramatic effects, such as complete suppression of germination and growth, were observed in any of the samples. These results indicate that, at the studied sites at least, oat is more resistant to contamination than ciliates and algae.

Table 2 Results of eluate biotest with the seeds of Avena sativa L.
Full size table

The highest toxicity was seen at site 4 and the lowest at site 2 (Table 5). Results of the ciliate and algae bioassays correspond with the results of chemical testing for contamination with heavy metals: both methods revealed sites 1 and 4 as polluted, although the reported severity of danger varied. Such partial agreement supports the established view on ciliate sensitivity to heavy metals (Voronich et al. 2016; Bardina et al. 2014a, b, c). However, chemical test results not always got in line with results of bioassays: mass spectrometry showed the site 1 as the most polluted by heavy metals, while bioassays consistently indicated the site 4 as the most hazardous. This may be due to an acidic environment at site 4 and high lability of heavy metals in acidic conditions.

Higher toxicity of soil samples from greater depth (5–20 cm vs 0–5 cm) at site 3 may be explained by downward migration of soluble forms of toxicants happening alongside water infiltration. Generalized relative sensitivity of the three eluate biotests may be summarized as: P. caudatum = S. quadricauda > A. sativa.

3.3 Ecotoxicological Evaluation of Soils with Contact Bioassay

We performed two types of contact biotests, using grains of Trinicum aestivum and natural microbial community.

Results of the bioassays with common wheat as test organism are given in Table 3. At site 1, we found moderate degree of toxicity; at site 2, the measured toxicity was low. The soils from site 3 were not toxic to wheat seeds. Indeed, our previous results indicate that at least 1 mg/kg of benzo(a)pyrene is needed to elicit a response (Bardina et al. 2016).

Table 3 Results of contact bioassay with seeds of Triticum aestivum L.
Full size table

Finally, most plots from site 4 had a dangerous level of toxicity. The single exception was plot 4, where the degree of phytotoxicity was low. Presumably, it is a result of uneven secondary contamination of the lands around the industrial waste dump at the heart of site 4. Acidic reaction of the surroundings facilitates the transition of the total forms of heavy metals into active forms, thus increasing phytotoxicity of the soil. Our earlier studies also found direct correlation between the response of T. aestivum and concentration of heavy metals (Bardina et al. 2014a, b, c). The most sensitive parameter, measured here as well as in other studies, was the root length, while the overall rate of germination was less affected (Isak et al. 2013; Terekhova et al. 2016).

Table 4 presents the results of microbial contact bioassay. The respiratory activity of microorganisms from most samples was significantly lower than the baseline. At sites 2, 3 (plot 1), and 4, the respiration rate was 30–70% less than a threshold value. It is known, that a critical level of soil system stability is the loss of not more than 30% of microbial biological activity (Yakovlev and Evdokimova 2011). Thus, we found a widespread and significant degradation of microbial communities in the studied soils, which, in turn, indicates low resistance of the soils as a whole to the toxic effects of pollutants. At site 3 (plot 2), the rate of reduction in biological activity did not exceed 30%; thus, degradation of microbial communities at this location may yet be reversible. However, site 1 is the most notable exception. At this site, we found no differences in biological activity of microbial communities of the control and test sample. We believe that the microorganisms at this locality are resistant to the existing toxicity (although chemical testing found this site as the most polluted with heavy metals). This apparent contradiction is resolved by noting that in a polluted locality which retains its vegetation cover, heavy metals lose their mobility and even enhance soil respiration and CО2 production (Vodyanitskii 2013).

Table 4 Respiration rate of soil microbial communities
Full size table

Results of two contact bioassays mostly agree with each other (Table 5). The two exceptions were sites 1 and 3. Soil sample from site 1 was moderately toxic to wheat but not toxic for microbial community. The latter may be caused by adaptation of the microorganisms to long-term contamination. Soils from site 3 demonstrated toxicity for microbial community but were not toxic for wheat seeds. This is consistent with our previous experiments, where response of wheat to pollution with benzo(a)pyrene was observed only when the content of this pollutant in loamy soils exceeded 50 MPC (Bardina et al. 2016). On the other hand, wheat is strongly sensitive to heavy metals (Bardina et al. 2014a, b, c), which is seen in complete agreement between estimates given by this test organism (Table 5) and chemical testing (Table 1).

Table 5 Integrated estimates of the soil toxicity (across all employed biotests)
Full size table

At some sites, contact bioassay methods were found to be more sensitive to the presence of toxicants than eluate bioassays. For instance, soil toxicity at site 2 was found by contact methods, while eluate bioassays did not detect the toxicants (probably organic in nature, as indicated by chemical tests). Notwithstanding some differences in sensitivity, however, all the bioassays revealed a dangerous level of toxicity at site 4.

4 Conclusions

The present study is focused on ecological evaluation of the brownfields using not only chemical, but also biologial testing methods. Chemical examination revealed that sited differed in pH reaction and identified heavy metals as primary pollutants. Various bioassay methods were more sensitive to the presence of toxicants than the methods of chemical analysis. In accord with previous publications, we found that reaction of different biotest systems depended on the chemical composition of the pollutants (Ram et al. 2004; Canna/Michaelidou et al. 2000). Various test species have unequal sensitivity to the wide range of toxicants. Aquatic organisms (protozoa and algae), used in eluate bioassay, were more sensitive than the seeds of higher plant. Both of the contact bioassays were sensitive to pollutants, since toxic effect of several samples was revealed more clearly with these methods than that with either chemical testing or eluate bioassays. In this study, we did not reveal any test method as superfluous: the wider the list of used methods, the larger the scope and reliability of pollution detection. Finally, we found that both eluate and contact bioassays are well-suited and should be used together for ecotoxicological assessment of brownfield soils. The latter finding agrees with the results for other types of objects (Terekhova 2011; Filenko and Terekhova 2016). We believe that compilation of a database of bioassay systems, suitable for testing brownfield soils, is an important step in remediation of these territories.