INTRODUCTION

Intensive urbanization requires comprehensive study of urban pedogenesis and of the anthropogenic transformation of soils [15, 28]. The aggregate composition of soil is a result of a complicated system of diverse interactions. It determines the water-air and temperature soil regimes, affects the microbiological activity of soils, and ensures the availability of nutrients for plants [3, 21, 31, 32]. Urban pedogenesis is accompanied by a decrease in soil ability to perform its ecological functions, which primarily concerns the absorption and neutralization of pollutants of various kinds and the prevention of their migration to adjacent media and contaminating soils of farmlands in the suburbs [1, 12, 30]. Thus, the study of soil structure under transformation occurring in the course of pedogenesis is an urgent task of modern soil science.

The aim of this study is to compare the ratio between structural fractions in different horizons of Calcic Chernozems and Urbic Technosol on Chernozems in the Rostov agglomeration and to determine the role of carbon in structure-forming processes.

OBJECTS AND METHODS

The study was performed in Rostov agglomeration, which is the largest one in the south of Russia. The city of Rostov-on-Don is its administrative center. A larger part of Rostov-on-Don, Aksai, and adjacent urban settlements are located on the interfluves of the Azov Plain and on gentle slopes of terraces on the right bank of the Don River. We studied soils with different degrees of anthropogenic transformation: untransformed or slightly transformed chernozems (Calcic Chernozems (Loamic, Pachic) of urban recreational zones and urbostratozems on chernozems (Urban Technosols) [19, 26, 34].

The studied soil types were characterized by soil pits located on watersheds with a thick (to 20 m) layer of Quaternary loess-like loam on marine Neogene sediments [6, 9, 10]. In total, 35 soil pits were studied, and the statistical analysis included data on 196 horizons.

According to the methodological approach, the sampling included dark-humus soil horizon (AU), carbonate-accumulative horizon (ВСА), soil-forming rock (C), and their buried analogues ([AU], [BCA], and [C]), which were then compared. The urbic horizons were grouped into two clusters according to the content of particles <0.01 mm: heavy-textured URh (40–60%) and coarse-textured URl (10–40%). This division is based on the genesis of urbic horizons: heavy horizons are formed of the material of natural chernozems, and coarse horizons are mainly composed of filled material (the content of particles <0.01 mm in native chernozems is 60–70%) [6, 7].

The structural status was determined by the Savvinov method: successive sieving of dry soil sample (dry sieving) and its sieving in water (wet sieving) on sieves with different mesh diameters [8].

Total (TC), inorganic (IC), and organic carbon (TOC) have been determined in the soil sample, using the TOC-L CPN Shimadzu carbon analyzer in the SSM-5000A auxiliary device for dry samples. The analysis is based on high-temperature catalytic combustion of the sample and subsequent detection of the released carbon dioxide. The analysis comprises two stages of sample combustion: at a temperature of 900°C (determination of the total carbon) and at 200°C with the addition of orthophosphoric acid (for inorganic carbon). Organic carbon was determined by subtraction of inorganic carbon from the total carbon [35].

The statistical analysis was based on the Student criterion to identify significant differences between the horizons in the content of structural fractions [14]. Since the regression analysis did not reveal the relationships between the studied parameters, we used the graphical analysis of the correlation between the content of particular structural fractions and of organic and inorganic carbon interpolated by the least square method in the STATISTICA program. This analysis enabled us to reveal changes in the relationships between soil structure and the content of organic and inorganic carbon.

RESULTS AND DISCUSSION

Table 1 presents the results of statistical processing of data on the aggregate composition (dry sieving) of particular genetic soil horizons in Rostov agglomeration. The urbic horizons are characterized by specific coarse angular blocky structure (>10 mm [18]), the portion of which is higher both in heavy (the Student criterion for compared pair URh–AU is 5.1 at n = 69; P = 0.05; tcr = 1.99) and coarse-textured horizons (the Student criterion for compared pair URl–AU is 4.3 at n = 64; P = 0.05; tcr = 1.99). The content of granular fractions (3–2 and 1–0.5 mm) in these horizons becomes lower, and that of fractions 0.5–0.25 and <0.25 mm increases in the URl and decreases in the URh horizons. The difference between the urbic and the dark-humus horizons of chernozem in the content of these structural units is also statistically significant.

Table 1.   The content of structural fractions in genetic horizons of soils of the Rostov agglomeration (dry sieving) (M ± m are given here and in the following tables, where M is the arithmetic mean, and m is the error of the mean, n = 33–52)
Full size table

The analysis of the structural composition of urban soils shows that direct correlations between the aggregate composition and the carbon content are not always recorded. For example, the organic carbon content in the urbic horizons is second only to that in the AU horizons, while the content of fraction >10 mm in the urbic horizons is significantly higher than in the AU, BCA, and C horizons. This is related to higher density of urbic horizons [6] as a result of an increased anthropogenic load on them. The data on the fraction <0.25 mm, the content of which is 2.6 times higher in the coarse-textured urbic horizons, explain the greater amount of organic carbon in these soils (Table 2).

Table 2.   The mean content of organic and inorganic carbon in genetic horizons of soils of the Rostov agglomeration and Student’s t-test (tcr = 2–2.04) for the compared pairs of horizons
Full size table

Many authors point to the relationship between the organic carbon content and the structural status of soils [2, 4, 23, 27, 29]. An increase in the percentage of agronomically valuable water-stable fractions parallel to the organic carbon content has been revealed in some soils of Great Britain [33]. However, the data of a three-year-long experiment performed under alfalfa (Мedicago) on the ordinary chernozem in Rostov oblast show that these correlations are seasonal and depend on weather conditions. The general correlation between the content of water-stable aggregates (WA) and the total amount of organic carbon is very slight or absent, while in autumn with cool and rainy weather, there is a direct dependence of WA on the humus content (r = +0.512), which is related to changes in its composition [4].

Wet sieving (Table 3) shows better water stability of aggregates in urbic horizons as compared to those in the horizons of native and buried profiles and does not confirm the expected quality of structural units in urban soils. This is probably explained by the contamination of the surface horizons by aromatic hydrocarbons of petroleum products, which is very typical for urban soils, near streets with heavy traffic in particular [16, 17, 22, 24]. Organic substances of aromatic pollutants are mainly hydrophobic [25], and according to the newest concepts, hydrophobic–hydrophilic relations in soil play an important role in structure formation [20]. If amphiphilic organic substances attach their hydrophilic part to a mineral particle, and their hydrophobic part interacts with hydrophobic parts of another amphiphilic organic particle, the aggregate is resistant to the destructive effect of water [32].

Table 3.   Mean content of structural fractions in genetic horizons of soils of the Rostov agglomeration (wet sieving)
Full size table

The burial of chernozems under urbic layers also affects water stability of aggregates. The content of water-stable aggregates >3 and 2–1 mm in the group of AU horizons is significantly higher as compared to [AU] horizons (the Student criterion is 2.4 and 2, respectively, at n = 67; P = 0.05; and tcr = 1.99), and the content of the fraction 0.5–0.25 mm is lower. In the middle and deep BCA–[BCA] and C–[C] horizons, significant differences in the status of water-stable fractions are not revealed.

Various types of regression analysis (linear, logarithmic, quadratic, and multiple) have not revealed any reliable dependences, and we cannot correlate the relationship of organic and inorganic carbon content with the aggregate composition of soils. The determination coefficient R2 does not exceed 0.2, which points to a slight effect of carbon on the content of structural fractions.

It should be pointed out that under the conditions of urban pedogenesis, the regression model cannot describe the complicated relationships between the content of the structural fractions and carbon, which are well pronounced in natural soils [2, 4, 23, 27, 29, 33].

Fig. 1.
figure 1

Changes in the content of air-dry fractions of aggregates: (a) >10 mm, (b) 7–5 mm, (c) 5–3 mm, (d) 2–1 mm, and (e) 1–0.5 mm in the phase coordinates of the contents of organic and inorganic carbon in the native (I) and buried (II) soil profiles.

Full size image

These relationships were visually assessed by a graphical analysis of the isopleths of the content of aggregate fractions on the phase plane with the coordinates of organic and inorganic carbon content in soil interpolated by the least square method (Fig. 1). This approach revealed a significant transformation of the correlations as a result of soil burial. The share of coarse angular blocky aggregates (>10 mm) increases parallel to the carbon content in the profile of native chernozem (Fig. 1a (I)) and contrary to it in the buried profile (Fig. 1a (II)). The content of this fraction in the latter profile is close to zero in a wide range of organic (1.5–3%) and inorganic (1.4–2%) carbon contents and more sharply increases after the plateau in the region of 23–30%.

Fig. 1b shows the relationship between the studied parameters and the amount of aggregates 7–5 mm. In the native and buried profiles, the maximal content of this fraction is allocated to the area of high organic and inorganic carbon content. However, in the region of low TOC and IC, the plots differ: the fraction content gradually increases in the native profile and sharply decreases in the buried profile.

Figure 1c describes the matrix of correlations for the aggregate fraction 5–3 mm. An axis of the high fraction content is pronounced in the buried soil profile at the level of 1.5–2% of TOC and throughout all IC values. The decrease in the fraction content in both directions from this axis is sharper in the direction of increasing TOC. Such axis is also seen in the native soil profile, but the pattern is smoother, and there is an increase in the fraction content at the small content of inorganic carbon.

The diagrams in Figs. 1c (5–3 mm) and 1d (2–1 mm) reflect similar regularities of the shift of the maximal and minimal content of fractions during burial.

Figure 1e (1–0.5 mm) also reflects sharp deformations of the correlation between the structural-aggregate fraction and organic and inorganic carbon in buried horizons. The native profiles are characterized by a gradual rise to a plateau and by a decrease in the content of fractions in the area of combined minimal and maximal values.

Thus, it may be concluded that the burial of the soil profile changes correlations between the structural-aggregate fraction and the content of organic and inorganic carbon. Numerous peaks usually appear, the diagram becomes more complicated, and there are areas of a sharp decrease in the content of fractions in comparison with the similar fractions in native chernozem, but similar features still remain. In few cases, the diagrams for coarse fractions in the buried and native soils are opposite. The difference is mainly related to a decrease in the total amount of organic carbon in buried profiles, in its labile components (carbohydrates, proteins, and water-soluble humus) in particular [5, 10, 12]. There are studies, showing that these substances play a significant role at the initial stages of structure formation [2, 26]. Thus, it may be assumed that the structure formation is slowed down in the areas of a sharp decrease in the content of fractions on the phase planes, reflecting the ratio between the content of organic and inorganic carbon in buried soils. These areas are unique for different structural fractions.

CONCLUSIONS

An increase in the portion of the air-dry fraction >10 mm both in the urbic and buried horizons is a clear evidence of urban pedogenesis at the macrostructure level. According to the data of dry sieving, coarse-textured (URl) urbic horizons differ from heavy-textured (URh) ones by higher content of fractions 0.5–0.25 and <0.25 mm. Standard relationships between the soil structure and the content of organic and inorganic carbon are not recorded in urbic horizons. The processes of structure formation in them are probably more strongly affected by the soil texture, compaction, and the presence of hydrophobic pollutants.

In comparison with native soils, the buried part of the urban soil profile is characterized by a tendency to an increase in the content of coarse angular blocky aggregates as a result of higher portion of agronomically valuable aggregates 7–5, 5–3, 2–1, and 1–0.5 mm. The water stability of the structure is a more constant parameter. The comparison of native and buried horizons shows a significant increase in the content of fraction 0.5–0.25 mm in the buried dark-humus horizons due to lower content of aggregates >3 mm and 2–1 mm.

Regression models cannot be used to describe the effect of different carbon forms on the content of structural fractions.

The graphical analysis shows that the relationships between the content of structural fractions and the amount of organic and inorganic carbon interpolated by the least square method are transformed as a result of soil burial. It is assumed that in native soils, each structural fraction is characterized by its own range of values on the phase plane with the coordinates of organic and inorganic carbon content, where the structure formation is intensive. The buried profiles are characterized by areas of a sharp decrease in the content of structural fractions in comparison with similar horizons of native soils.