Issues related to classification of garden soils from the urban area of Toruń, Poland

ABSTRACT

Nowadays, the city area of Toruń is dominated by anthropogenic and technogenic soils – developed by humans or significantly altered, mainly disturbed in terms of morphology and chemical and physical properties. This study is a continuation of research on the soil cover of the city. The aim of the presented study is to assess extent of garden soils in the city and characterize its properties on the base of five soil profiles in four exemplary gardens in Toruń and evaluate if they meet the classification criteria for Hortic Anthrosols according to World Reference Base (WRB) for Soil Resources. Within the administrative boundaries of the city, 66 allotment gardens are located which totally cover an area of more than 300 ha. They occupy 3% of the city area. None of the studied mineral surface horizons meets the criteria for hortic horizon according to WRB 2015, due to too low phosphorus content. Other hortic criteria were fulfilled. The research on classification issues of garden soils should be continued on larger scale to evaluate if WRB criteria are not too strict taking into account the features of most typical, few decade-old garden soils.

KEY WORDS:

• Urban soils
• garden soils
• WRB soil classification
• hortic
• heavy metals
• SUITMAs

1. Introduction

Soils of Toruń urban area were intensively researched over the last decade (Capra et al. 2015; Charzyński and Hulisz 2017). Most of the investigations were focused on moderately or heavily disturbed soils (SUITMAs – soils of urban, industrial, traffic, mining and military areas), especially due to housing construction activities and traffic (e.g., Charzyński et al. 2017; Hulisz et al. 2016; Mendyk and Charzyński 2016). Less attention has been paid to urban greenery and horticultural soils (Charzyński et al. 2013).

Nowadays, the city area of Toruń is dominated by anthropogenic and technogenic soils – developed or significantly altered by humans, mainly disturbed in terms of morphology and chemical and physical properties. Only 25% of the soil cover represent relatively natural soils. They remained on the outskirts of the city, within 20 forest complexes and areas located on the Vistula River floodplain (Bednarek and Jankowski 2006).

Garden soils (hortisols) are one of the purposely modified soil types. This soil type is distinguished in many national systems, e.g., German (Arbeitskreis für Bodensystematik der Deutschen Bodenkundlichen Gesellschaft 1998), Polish (Polish Soil Classification 2011) and Czech (Nemeček et al. 2001). They mainly cover areas of allotments. Allotment gardens can be found in almost all Polish cities and towns. Most of them were created to meet the shortcomings of the centrally planned economy of socialist Poland with a deficit of fruit and vegetables. Initially, they occupied areas on the outskirts of municipalities or even outside the city limits. They were different types of wastelands and wetlands, which for various reasons were not suitable for housing and industrial development. The expansion of cities in the following decades has meant that many gardens are currently surrounded by new housing estates and are often located relatively close to the city centers. This phenomenon is clearly visible in the area of Toruń.

Studies of garden soils, both in Poland and in other parts of the world, are primarily associated with determining the degree of contamination with heavy metals or other pollutants, especially in the context of health risk assessment. Such publications are quite numerous. For example, Bielińska (2006) studied the content of heavy metals in garden soils in Lublin, Biała Podlaska, Zabrze and Kraków, while Kabała et al. (2009) performed similar studies in Wrocław. The quality of soils in selected allotment gardens, located in the vicinity of major transport routes in Łódź, was investigated by Niewiadomski and Szubert  (2014). The influence of intensive heavy metal emission by heavy traffic on the content of these elements in the soils of gardens in Warsaw was researched by Dmochowski et al. (2011). The impact of human activities on the accumulation of metals in garden soils in the Hungarian city of Szeged was investigated by Szolnoki et al. (2013). Kim et al. (2015) considered metal concentrations in rooftop gardens located in Seoul, the Republic of Korea – in the growing media and crop plants cultivated there. Izquierdo et al. (2015) determined the pseudototal and bioaccessible trace element contents in soils of six urban gardens in Madrid, Spain. Jean-Soro et al. (2015) investigated arsenic and lead concentrations in one of the urban community gardens from Nantes, France.

Issue of anthropogenic influences in garden soils in context of diagnostic criteria used in World Reference Base (WRB) soil classification was discussed in Mester et al. (2017).

The aim of the presented study is different from the aforementioned papers apart from the last one. It is to evaluate whether the mineral surface horizons of allotment garden soils meet the classification criteria for hortic according to the World Reference Base for Soil Resources (2015) and to assess the extent of garden soils in the city and to characterize the physical and chemical properties of soils of selected allotment gardens in Toruń.

2. The study area

Toruń is located in Northern Poland – 53°01′ N, 18°36′ E (Fig. 1). It is a medium-size city (116 km²) with the population of nearly 192,000 (Census of the City Council of Toruń 2017). There are 66 allotment gardens within the administrative boundaries of the city, which together cover an area of 348.8 ha. They occupy 3% of the urban area. The largest complex of such gardens is located on the left bank of the Vistula River, in the district of Rudak. Most of the oldest gardens, created in the 1950s, is presently in the central part of the city, surrounded by built-up areas constructed in the last 50 years. However, many youngest gardens, created in the 1980s, despite the constant urban development of Toruń, are away from the city center. The garden complexes were named after the factory they belonged to (e.g., Metalchem), landscape features (e.g., By the brook – Nad Strumykiem), location (e.g., Polder), famous historical figures (e.g., A. Mickiewicz – Polish romantic poet) or historical event/anniversary (10th anniversary of the Polish People’s Republic). Particular garden complexes are divided into small individual parcels (e.g., 3 ares each parcel) belonging to separate possessors. This fragmentation enabled conducting labor-intensive agrarian treatments. Distribution of allotments within the administrative boundaries of the city, including the time they were established and partition into individual plots, is shown in Fig. 1.

Figure 1. Toruń, allotment garden complexes and study site location.

Study sites considered in this paper are located in the Toruń Basin. Three researched gardens with one soil profile each are located in the floodplain of the Vistula River, originally covered with alluvial soils (mostly Fluvic Phaeozems and Fluvic Gleysols). The floodplain terrace lies directly at the riverbed and was periodically flooded before the river flow regulation between 1835 and 1892 (Glazik and Kubiak-Wójcicka 2006) and the construction of levees. Silty and sandy sediments are most common here (Niewiarowski and Weckwerth 2006). The study profiles are located in allotment gardens established about 35 years ago. When creating the gardens, the land surface was raised by adding mineral materials to improve the physical properties of the soil, i.e., to decrease excessive humidity by reducing the impact of the high groundwater table.

It is not possible however to determine precisely from which natural soil pedon number 2 was formed. Texture, organic carbon and nitrogen content indicate alluvial soil. The absence of carbonates and low pH values indicate that this could also be a Fluvic Umbric Gleysol (Kabała et al. 2016). Properties of the garden soils are mainly determined by agrotechnical treatments, but some features are still inherited from original soils.

The last two research sites are located on the fourth terrace developed during Older Dryas (from 14–13 ka BP to 12.4 ka BP) and Bolling (12.4–12.1 ka BP) originally built from Pleistocene fluvial deposits covered in many places by sand dunes (Niewiarowski and Tomczak 1973; Niewiarowski and Weckwerth 2006). These areas were primarily dominated by Brunic Arenosols. Profiles 4 and 5 are located in the former botanical garden of Nicolaus Copernicus University in Toruń.

3. Material and methods

The field work was carried out in October and November 2012 after determining in which year the gardens were established. This information was not available in municipal archives, and meetings with authorities of each allotment garden were necessary to obtain information about the date of foundation (Fig. 1). Five soil profiles were excavated. Soil samples were collected from all designated horizons in each soil profile (Fig. 2). After air-drying at room temperature and sieving (2 mm), the soil samples were analyzed for chemical and physical properties. Standard methods of analysis were used to determine physicochemical properties (Van Reeuwijk 2002). The texture was measured by the aerometric method combined with the sieve method. The pH in 1 mol l⁻¹ KCl was measured using the potentiometric method in 1:2.5 soil:solution suspensions; the total organic carbon (OC_TOT) and total nitrogen (N_TOT) content was determined using a dry combustion CN analyzer (Vario Max CN); P₂O₅ content was measured using Olsen et al. method (1954). Moreover, hydrolytic acidity and sorption properties were identified. The position of soils in the urban area results in the predicted contamination with heavy metals; therefore, also analysis of heavy metal (Pb, Zn, Cu and Cd) content was performed. The total content of these elements (denoted by _TOT) in the soil samples was determined using the ICP MS (Inductively Coupled Plasma Mass Spectroscopy) technique after dissolving with aqua regia.

Figure 2. Morphology of the study profiles (photos by P. Charzyński).

To achieve the objective of the study, it was examined whether the surface horizon of the soils meets the criteria of the hortic horizon according to WRB (IUSS Working Group WRB 2015). There are six criteria:

1. Munsell color value and chroma of ≤3, moist;

2. weighted average of ≥1% soil OC;

3. 0.5 M NaHCO₃ extractable P₂O₅ content of ≥100 mg kg⁻¹ fine earth in the upper 25 cm;

4. base saturation (by 1 M NH₄OAc, pH 7) of ≥50%;

5. ≥25% (by volume) of animal pores, coprolites, or other traces of soil animal activity; and

6. thickness of ≥20 cm.

Finally, interviews with owners of gardens were performed at the sites of sampling to learn about agronomic treatments, fertilization, and vegetable growing.

4. Results and discussion

Morphology of soils located on the floodplain (sites 1–3) has many similarities. The studied soils represented poorly diversified texture distribution, and according to the USDA (United States Department of Agriculture), they represented mainly sand and loamy sand. Profile 2 is the most interesting profile in terms of morphology. The mineral surface horizon is underlain by an older tilled AC horizon with peat remains. Artifacts and other objects introduced into the soil by humans – egg shells and charcoal – are present in both horizons. In this profile, also gleyic properties in the underlying material (Cl) can be observed.

Profile 5 is distinguished by the presence of a technogenic layer (30–95 cm), consisting of 80% of artifacts.

The investigated profiles have a different texture, which is related to the characteristics of soils that formed the analyzed garden soils. Soils located in the floodplain of the Vistula River, and developed from alluvial soils, are characterized by the finer texture (loamy sand mostly). Garden soils of the fourth terrace are characterized by the most coarse texture. The lack of impact of groundwater indicates that these places were originally covered with Brunic Arenosols (Table 1).

Table 1. Selected physical and physicochemical properties of the soils.

Horizon	Depth [cm]	pH H2O	pH KCl	CaCO3 [%]	TOC [%]	Total N [%]	C/N	P2O5 [mg kg^−1]	CEC [cmol(+) kg^−1]	Base saturation [%]	Textural class	Munsell color
												Dry	Moist
Profile 1 – Zagroda Garden
Ap	0–25	7.5	7.3	1.7	2.,68	0.203	13	86	18.0	96	LS	2.5Y 4/1	2.5Y 3/1
AC	25–50	7.9	7.5	1.1	0.76	0.074	10	45	13.7	98	LS	10YR 5/2	10YR 4/2
C1	50–80	8.3	7.6	0.5	0.25	0.020	12	23	7.5	98	LS	10YR 5/3	10YR 4/3
C2	>80	8.3	7.5	^a	0.22	0.024	9	22	7.8	98	LS	10YR 5/3	10YR 4/3
Profile 2 – St. Christopher’s Garden
Ap	0–30	6.7	6.3	n	2.59	0.197	13	51	17.1	90	LS	2.5Y 5/2	2.5Y 3/1
AC	30–50	4.8	4.0	n	0.36	0.037	10	12	4.2	38	LS	10YR 5/3	10YR 3/3
C	50–80	4.9	3.9	n	0.11	0.011	10	7	3.6	46	LS	10YR 6/3	10YR 4/4
Cl	>80	5.4	4.4	n	0.14	0.015	10	3	8.6	82	SL	10YR 5/4	10YR 4/4
Profile 3 – Relax Garden
Ap	0–20	7.5	6.9	0.3	1.66	0.145	12	61	11.6	92	LS	2.5Y 4/2	2.5Y 3/2
AB	20–70	7.4	6.7	^a	0.83	0.072	12	42	2.6	85	S	10YR 5/2	10YR 4/2
C	>70	7.4	6.3	n	0.26	–	–	24	1.3	83	S	10YR 6/3	10YR 5/4
Profile 4 – Botanical Garden of NCU
Ap	0–30	7.4	6.9	0.4	1.49	0.079	19	92	8.2	89	S	2.5Y 4/1	2.5Y 3/1
AC	30–45	7.7	7.1	n	0.39	–	–	68	2.9	86	S	10YR 5/3	10YR 4/3
C	>45	7.6	6.9	n	0.20	–	–	31	0.9	71	S	10YR 6/3	10YR 4/3
Profile 5 – Botanical Garden of NCU
Ap	0–30	7.6	7.2	0.4	1.49	0.103	14	77	10.4	94	S	2.5Y 4/2	2.5Y 3/1
2Au	30–95	7.7	7.1	0.8	2.64	0.128	21	117	10.8	93	S	2.5Y 4/2	2.5Y 3/1
3C	>95	7.6	6.9	n	0.20	–	–	31	0.9	71	S	10YR 6/3	10YR 4/3

^aTrace amount.

n: lack of carbonates; –: not determined.

Texture: S: sand; LS: loamy sand; SL: sandy loam.

The color of all investigated mineral surface horizons fulfills criteria of the hortic horizon (Table 1). Also the weighted average organic carbon content in the mineral surface horizon of every profile was higher than 1%, ranging from 1.49% to 2.68%. These criteria for the hortic horizon are thus fulfilled in every case. The highest amount of organic carbon occurs in the mineral surface horizons (Table 1). The OC content decreased below this horizon.

The nitrogen content of soil mineral surface horizons is usually 0.06–0.3% (Bednarek et al. 2004). The nitrogen content in the analyzed soils is within the above range, while lower-lying horizons show a much smaller content of this element or in some cases amounts below detection limit of nitrogen (Table 1).

The high content of OC and N is due to organic fertilizers used by every garden’s owner, mainly in the form of farmyard manure and residues from garden plants.

The data presented in Table 1 show that the content of organic carbon corresponded to the nitrogen content. The C/N ratio in the analyzed soils ranged from 9 to 13 in profiles 1–3 and from 14 to 21 in profiles 4 and 5. Profiles 1–3 are characterized by a higher biological activity and faster organic matter transformations.

The content of carbonates in the studied soils is low, and in some cases, they are even absent. The pH of the analyzed soils is high (except profile 2 with acidic soil reaction in the subsoil, pH_KCl ranging from 3.9 to 6.3; Table 1). Profiles 3, 4, and 5 have neutral pH and profile 1 has alkaline pH. The pH of the soils is closely related to the use of fertilizers containing calcium carbonate and magnesium carbonate. Lime applied in the studied gardens causes a significant increase in pH compared to natural soils of the investigated areas.

Another criterion for the mineral surface horizon to be classified as hortic is the base saturation (by 1 M CH₃COONH₄) at pH 7 of 50% or more. This condition is fulfilled for all the analyzed soils (Table 1). Base saturation is very high in each profile. The highest values of cation exchange capacity (CEC) were determined in mineral surface horizons and decrease with the depth. CEC is clearly dependent on the OC and colloidal clay content (Bednarek et al. 2004). The investigated soils from the higher Pleistocene terraces with a more sandy texture (profiles 4 and 5) are characterized by lower values of CEC compared to garden soils with a higher clay content (1–3).

Sources of phosphorus in the soil are native mineral and organic phosphorus compounds as well as chemical precipitation. However, the type and amount of fertilizers used significantly affect the phosphorus content (Sądej 2000). The greatest amount of phosphorus is applied to the soil with farmyard manure and phosphatic and potassium fertilizers. Both types were periodically used in the studied gardens. The content of this element clearly decreases with the increasing depth because phosphorus complexing with humic substances. Phosphorus content in analyzed profiles also corresponded to organic carbon amount – the higher the content of OC, the higher the amount of phosphorus.

Profile 5 is an exception, especially its technogenic layer. Phosphorus content is much higher than in the upper mineral surface horizon (Table 1). Such high values compared to all other horizons (both in this and other profiles) are due to the presence of large amounts of artifacts (especially animal bones). Nevertheless, none of the studied soils’ mineral surface horizons meets the criteria for the hortic horizon, due to low phosphorus content.

Another criterion of the hortic according to WRB is the presence of high (≥25%) amount of animal pores, coprolites, and other traces of soil animal activity. During the field studies, it was estimated that the researched profiles meet this criterion.

According to the WRB, studied soil No. 1 can be classified as Fluvic Terric Phaeozem (Arenic, Aric), No. 2 as Terric Umbrisol (Arenic, Aric), soil No. 3 as Terric Phaeozem (Arenic, Aric, Humic), while profile No. 4 as Haplic Phaeozem (Arenic, Aric) and soil No. 5 with the horizon containing 80% of artifacts becomes an Urbic Technosol (Arenic, Profundihumic, Mollic).

In addition, the total content was measured for selected, environmentally hazardous elements: zinc, copper, lead and cadmium. Some of the Toruń gardens are located in the vicinity of industrial areas or traffic routes, but this does not apply to gardens selected for the presented study. The content of zinc in soil varies, and its highest concentration of 578 mg kg⁻¹ was determined in the technogenic layer of profile 5 (Table 2), where many metallic artifacts (Fig. 3) were present. The larger amounts of lead were found in profiles 4 and 5 (Table 2). In other soils, this element was below the limit of detection. Background values of lead in Polish sedimentary rocks were investigated by Czarnowska (1996). The natural level of Pb in parent material is 9.8 mg kg⁻¹, varying from 0.5 to 21.0 mg kg⁻¹ depending of the type of rock and its texture.

Table 2. Content of selected trace elements in the studied soils.

Horizon	Depth [cm]	Zn	Pb	Cu
Profile 1 – Zagroda Garden
Ap	0–25	53	^a	^a
AC	25–50	37	^a	^a
C1	50–80	23	^a	^a
C2	>80	23	^a	^a
Profile 2 – St. Christopher’s Garden
Ap	0–30	33	^a	^a
AC	30–50	16	^a	^a
C	50–80	14	^a	^a
Cl	>80	23	^a	^a
Profile 3 – Relax Garden
Ap	0–20	80	^a	21
AB	20–70	51	^a	11
C	>70	6	^a	^a
Profile 4 – Botanical Garden of NCU
Ap	0–30	107	40	30
AC	30–45	9	^a	^a
C	>45	6	^a	^a
		30.0	98	7.1
Profile 5 – Botanical Garden of NCU
Ap	0–30	142	^a	^a
2Au	30–95	577	148	22
3C	>95	6	^a	^a

^aBelow the limit of detection.

Limits of detection: Zn = 3, Pb = 16, Cu = 7, Cd = 5 mg kg⁻¹.

Maximum permissible quantities of trace elements in soils according to Kabata-Pendias and Pendias (1999) in mg kg⁻¹: Zn = 300, Pb = 100, and Cu = 100 mg kg⁻¹.

Figure 3. Artifacts from profile No. 5 (photo by P. Charzyński).

The slightly elevated values of copper were detected in profiles 3, 4, and 5 (Table 2), while the average natural level was estimated at 7.1 mg kg⁻¹ (Czarnowska 1996).

Cadmium was not detected in any of the analyzed profiles because the limit of detection was = 5 mg kg⁻¹, while the amount of 3 mg kg⁻¹ is the maximum permissible amount of this element in soils according to Kabata-Pendias and Pendias (1999). What is more, its natural level in sedimentary rocks in Poland is only 0.18 mg kg⁻¹ (Czarnowska 1996). The total content of the studied heavy metals did not exceed values admissible for arable soils in Poland.

Monitoring results for the mid-1990s in selected 13 allotment garden complexes pointed to zinc and lead pollution. The extreme contamination with these elements was found in gardens located near the northern exit road (Jankowski 1995). The relationships between the content of the analyzed trace elements in soils and the distance from the busy thoroughfares were significant. Average quantities of the studied heavy metals in soils located near the traffic routes were almost two times higher than in soils away from the road. It should be noted, however, that these results are for the period when unleaded fuel was just introduced, so sources of such pollution are no longer present.

5. Conclusions

Physical, physicochemical, and chemical properties of garden soils are to some extent conditioned by features of soils from which they developed. Study soils No. 1–3 located within the floodplain of the Vistula River were formed from alluvial soils. Basic physical and chemical properties (texture, pH, and carbonate contents) also indicate that origin. Compared to the adjacent undisturbed alluvial soils, garden soils are characterized by a higher content of organic carbon and nitrogen in the mineral surface horizons. These changes in the properties are caused by purposeful human activity.

Locations of profile No. 4 and 5 were primarily covered by Brunic Arenosols. Such origin is mainly indicated by sandy texture of all horizons and lack of ground water strong influence. The content of organic carbon, nitrogen, and pH values are elevated as a result of agrotechnical treatments.

The study has shown that none of the mineral surface horizons meets the criteria of the hortic horizon, due to low phosphorus content – values below 100 mg∙kg⁻¹. It can be assumed that 35–50 years of horticultural practices were not enough to develop the diagnostic hortic horizon despite intensive fertilization and other agricultural practices (addition of peat and mineral materials). The authors have undertaken a challenge to preliminary evaluate the hortic horizon criteria. Those criteria was based on the research of convent garden soils cultivated for centuries. The vast majority of garden soils are much younger. Allotment gardens started to be established at the end of the nineteenth century, but most of them were founded after World War II. The research on the issues related to the classification of garden soils should be continued on a larger scale to evaluate whether the current WRB criteria are not too strict when taking into account the features of most typical, a few decade-old, garden soils.

Unfortunately, phosphorus analytical procedure listed in hortic criteria was not commonly used, so it was not possible to discuss results with literature analytical data. The problem of lack of unified phosphorus tests in WRB criteria of various diagnostics (three different ones used) is discussed by Kabała et al. (2018).

Nevertheless, the current WRB offers a well-designed set of terms related to anthropogenic and technogenic soils and is serving as a model solution for national soil classifications. Terminology related to urban soils will further evolve in this international system, under the influence of numerous new proposals (like it was in the past) to be elaborated within IUSS SUITMA Working Group.