The Influence of Freeze–Thaw Processes and the Organic Supplementation on the Structural Parameters of Luvisol

Abstract

From an agronomic perspective, the most important property of soil aggregates is their stability. The stability of the soil structure can vary greatly, even over a short time. In this context many authors draw attention to the influence that climatic factors have on the stability of aggregates. The process of freeze–thaw cycles (FTC) could result in changes in the aggregates’ stability. The research objective was to determine the effect of alternating freeze–thaw cycles on selected parameters of soil structure in Luvisol depending on the type and duration of action of organic supplementation. Cyclic freeze–thaw (FT) processes significantly modified most of the analysed properties of soil aggregates. This effect was usually not modified by organic supplementation, nor by its duration of action. FT cycles reduced the density of soil aggregates. FT cycles reduced the resistance of soil aggregates to dynamic and static water action. Despite the lower resistance to the destructive action of water, the decomposition of primary aggregates resulted in a more favourable size distribution of secondary aggregates in aggregates subjected to FT cycles. The research objective was to determine the effect of alternating freeze–thaw cycles on selected parameters of soil structure in Luvisol depending on the type and duration of action of organic supplementation.

1. Introduction

In agronomic practice, soil structure is a key determinant of plant production properties and—de facto—a range of physical soil properties. In field conditions, stable soil structure ensures a favourable content of capillary pores, which serve to retain water that can be used by plants, and an appropriate content of macropores, which determine water conductivity, air capacity, and air permeability [1,2]. From an agronomic perspective, the most important property of soil aggregates is their stability. The stability of the soil structure can vary greatly, even over a short time. This has been reported by, inter alia, Algayer et al. [3], who drew attention to the influence that climatic factors (moisture, temperature) have on the stability of aggregates. Leuther and Schlüter [4] agreed and emphasised the importance of the movement of water from the unfrozen zone to the frozen zone. According to them, this causes the uneven distribution of ice crystals and the wetting and drying of the soil, which in turn affects soil structure. Soil freeze–thaw cycles (FTCs) may mitigate the effects of soil compaction, as pointed out by Wang et al. [5], who called these processes “ecologically friendly”. They believe that the impact of FTCs may vary depending on, among other things, soil texture, moisture, and bulk density. As a result, they stated that there are still doubts as to the beneficial effects of FTCs on the physical properties of soil. Some authors have considered FTC processes in the context of climate change [6]. According to the authors variations in seasonal soil freeze/thaw state are important indicators of climate change and influence ground temperature, hydrological processes, surface energy, and the moisture balance.

Given the value of any research that contributes to expanding knowledge about the influence that FTCs have on the properties of soil structure, the present study was undertaken. At the same time, due to the complex nature of these properties, the research objective was to determine the effect of alternating FTCs on selected parameters of soil structure in Luvisol depending on the type and duration of action of organic supplementation. Previous studies have confirmed the beneficial effect of adding cattle manure on soil structural properties [7]. Three research hypotheses were adopted: (i) freeze–thaw cycles significantly change soil structure parameters; (ii) this effect varies depending on the type of organic supplementation; and (iii) this effect varies depending on the duration of action of the organic supplement.

2. Materials and Methods

2.1. Experimental Design

The study adopted the concept of soil structure models developed by the Poznań University of Life Sciences (Poland) [8]. The intention of its authors was to develop a methodology that would allow for extensive studies of soil structure. Because most research methods could not be carried out on natural aggregates, they developed a research methodology that uses not natural aggregates, but aggregate models cut from the soil using a 1 cm³ cylindrical sampler (diameter 11.3 mm; height 10 mm). The cylinder sample was made by Technical Department of the Agricultural University in Poznań (Poznań, Poland). In the model research concept, the authors understood water resistance to be the time it takes for aggregates to dissolve when soaked (i.e., resistance to the static action of water) and the energy (or, in cases of weakly resistant models, number of water drops) needed to break the aggregates (i.e., resistance to the dynamic action of water). The authors use the term “secondary aggregation” to refer to the phenomenon of soil forming stable secondary aggregates with a diameter of >0.25 mm under the influence of water—which is called “aggregate stability” in the literature (e.g., [3]).

The results described in the paper complement research conducted in a three-factor laboratory experiment set up in 2015 to assess the effect that applying organic additives, microbiological inoculation, and time had on selected structural parameters [7]. It was decided to conduct a laboratory experiment rather than a field experiment due to the need to strictly control conditions, in particular the homogeneity of the texture, the temperature, and the moisture of the soil. All these variables were factors that, in a field experiment, would be uncontrolled and could therefore not be taken into account during statistical calculations. A commentary on results of this experiment is provided in a monograph by Gajewski [7]. In parallel, the impact of FTCs on selected combinations was analysed, and the results are presented herein. To establish the above-mentioned experiment, the arable horizons of two Luvisols were collected: (i) one developed from loamy sand (ii) one developed from loess (silty loam); the texture was determined according to the USDA standard [9]. Luvisols occupy large areas in Central and Eastern Europe [10]. They are the most important group of soils in Poland, covering ~45% of its area. They are mostly used for agricultural purposes, as they are characterised by average-to-good agricultural value and are suitable for growing wheat, maize, and rapeseed [11]. The soil material collected in the field was derooted and then divided into parts of equal weight. The tested additives and water were applied to individual samples in sufficient quantities to bring the samples to a moisture content corresponding to the field capacity (FC). The samples were sealed in plastic bags and left for 24 h for the moisture content to equalise throughout the mass. After this time, the samples were placed in plastic containers with a length of 24 cm, width of 14 cm, and depth of 2 cm. The entire experiment included 100 combinations [7]. To evaluate the impact of FT processes, “control” combinations “C” (without organic additive) were selected, as well as combinations to which composted cattle manure “M” had been added. Manure was added at a rate equivalent to 30 tons of fresh matter per hectare. At five different times (after 30, 60, 90, 270, and 365 days), 70 models were sampled. Throughout the experiment, the soil in the containers was maintained at a moisture content corresponding to FC. For this purpose, holes were drilled in the containers to allow the free gravitational drainage of water. The effect of freezing and thawing of aggregates was tested on both soils. This paper presents the results obtained for the soil developed from loess (i.e., Luvisols). According to the WRB taxonomy [12], this soil was classified as Albic Luvisol and according to Polish Soil Classification [13], as typical clay-illuvial soils. The soil material was collected in Brochocin, SW Poland (51°16′42″ N 17°05′19″ E). Its basic properties are presented in

Table 1. Properties of examined soil material.

Content (%) of Fraction of Diamater in (mm)
2–1	1–0.5	0.5–0.25	0.25–0.1	0.1–0.05	0.05–0.02	0.02–0.005	0.005–0.002	<0.002	TC (g kg−1)	TN (g kg−1)	C:N	pH in 1MKCl	CaCO3 (%)
0.50	1.50	2.5	10.5	3.0	51.0	24.0	3.0	4.0	14.1	1.43	9.9	7.40	1.26

and the experimental scheme in

Table 2. The experimental scheme.

Term 1 (30 Days)	Term 2 (90 Days)	Term 3 (180 Days)	Term 4 (270 Days)	Term 5 (365 Days)
NFC	NFC	NFC	NFC	NFC
FC	FC	FC	FC	FC
NFM	NFM	NFM	NFM	NFM
FM	FM	FM	FM	FM

Note: NFC—control: aggregates not frozen; FC—control: frozen aggregates; NFM—manure: aggregates not frozen; FM—manure: frozen aggregates. Aggregates that were subjected to freeze–thaw cycles were taken from the same containers as aggregates that were not subjected to these cycles.

. The chemical properties of the cattle manure were as follows: TC: 407.2 g kg⁻¹; TN: 20.9 g kg⁻¹; and C:N 19.5. Throughout the entire 365-day experimental period, the soil moisture in the containers was kept at a constant level corresponding to FC moisture content; the temperature in the laboratory ranged from 20 to 24 °C.

2.2. Analytical Methods

2.2.1. Freeze–Thaw Cycles

A random selection of half of the moist (FC moisture) aggregate models from the tested combinations was placed in Petri dishes and subjected to cyclic (five cycles) freeze–thaw processes. In line with the well-known concept of Henry [14], realistic temperature fluctuations were used. Immediately after collection, the models were transferred to a refrigerator–freezer where they were maintained at +5 °C for 12 h. The temperature was then lowered to −5 °C, and the models were kept there for another 12 h. The final stage of freezing was to maintain the models (12 h) at −15 °C. Thawing was performed in reverse order, maintaining the 12 h time. The freezing temperature was selected based on long-term climate measurements for the Lower Silesia region for the years 1981–2000 [15], i.e., from the region that the soil material came from. This choice is in agreement with the latest climate data for this region [16]. The number of cycles used by FT cycle researchers is also problematic. The literature contains articles that use three cycles [17], six cycles [18] or even 19 [4]. The number of cycles used herein is a “compromise” between these studies Another methodological difference is the soil moisture at which researchers conduct FT cycles. There are studies [17] in which samples were frozen at sample-time moisture, whereas others [19] used FC moisture.

2.2.2. Structural Properties

All analyses described below were carried out on aggregates either subjected to cyclic freeze–thaw processes or not (F and NF, respectively) and were performed on aggregates dried at room temperature.

The following analyses were performed according to the Rząsa and Owczarzak model [8]:

-

Analysis of the aggregate resistance to dynamic water action (RDWA): This consists of breaking down the aggregate with drops of 0.05 g mass falling from a height of 1 m, i.e., with kinetic energy = 4.905 10⁻⁴ (J). The results presented in the tables are the average of 15 repetitions (Figure 1).

-

Analysis of aggregates’ resistance to soaking (static water action; RSWA): This is measured by the disintegration time of the aggregate immersed in water in a plexiglass container on nylon threads spaced 6 mm apart. The results presented in the tables are the average of 15 repetitions (Figure 2).

-

Secondary aggregation state: The analysis involves separation on a set of sieves with a 7, 5, 3, 1, 0.5, and 0.25 mm mesh diameter of the primary aggregate fraction that has been disintegrated as a result of the dynamic and static action of water. According to the methodology, the optimal number of aggregates used to analyse this property is five. The analysis of the secondary aggregation state was therefore carried out in three repetitions—each consisting of five aggregates from the analyses of the aggregates’ resistance to static and dynamic water action.

The methodologies used in soil structure studies vary greatly, which reduces the comparability of results [20]. At the same time, the conclusions drawn from the last-cited article allow us to assume that trends in the variability of structural properties correlate, whatever methodology is applied. This difficulty requires that the researcher interpret not the absolute values of individual features but the trends in their variations.

2.2.3. Properties of Soil Material and Organic Supplements

The following properties were determined in the initial soil material from the individual combinations and in the manure:

-

Texture (as a general characteristic of the soil material used) by the sieve method (sand) and areometric method (silt and clay) after dispersion with sodium hexametaphosphate, in accordance with the Ryżak et al. [21];

-

TC (total carbon) and TN (total nitrogen) content using a Vario Max CNS (Elementar) analyser;

-

CaCO₃ content using a Scheibler apparatus;

-

particle density (PD) by the pycnometric method [22]. The soil material used for analysis of PD was not derived from excised aggregate models. This trait would be impossible to analyse in single-aggregate models because the mass of one sample (modelled aggregate) does not exceed 1.8 g. Therefore, it was decided to determine this trait in five samples of approximately 10 g each taken from each container (with each combination being represented by a single container); these five were then combined to create a single, mixed sample. The variability of PD within a single combination was assumed to be small. Accordingly, it is acceptable in this case to use one average value of PD and to use five different bulk densities to calculate the total porosity of the five aggregate models.

It was also assumed that PD would not change under the influence of the FT factor. Therefore, for combinations in which FT was the only discriminating factor, it was assumed that the PD value was the same.

-

Bulk density—calculated from the mass and volume of soil aggregate models. This analysis was performed in five replicates (five aggregate models)

-

Total porosity—calculated from the determined PD and bulk density of the soil [23].

2.2.4. Verification of Research Hypotheses

The research hypotheses herein complement the hypotheses formulated in the experiment conducted in 2015 (Section 2.1). The earlier studies investigated how organic additives, effective microorganisms, and time affected the properties of soil aggregate models [7]. In this paper, the influence of the organic additive (manure) factor (A) and time (C) is considered in the context of the variability of action of the FTC factor (B) and its interaction with other factors.

2.3. Statistical Analysis

The results were statistically analysed by ANOVA, using Duncan’s test to determine the significance of differences at the level of p ≤ 0.05. In order to compute the statistical analyses, STATISTICA version 12.0 (StatSoft Inc., Tulsa, OK, USA) was applied.

3. Results and Discussion

3.1. Chemical Properties of Soil Aggregate Models

The chemical properties of soil aggregates are presented in

Table 3. Influence of experimental factors on the content of total carbon and total nitrogen.

Organic Supplement A	Days of Incubation B					Mean for A
	Content of Total Carbon (g kg −1)
	30	60	90	270	365
NFC	14.2 abc	14.2 abc	14.2 abc	13.8 a	13.8 a	14.0 a
NFM	15.0 b–f	14.6 a–e	14.8 a–f	14.4 a–d	14.2 a–c	14.6 b
Mean for B	14.6 c	14.4 bc	14.5 ab	14.1 a	14 a	-
Total nitrogen (g kg −1)
NFC	1.42 b–g	1.39 a–d	1.39 a–d	1.43 c–h	1.34 a	1.39 a
NFM	1.51 i–l	1.42 b–g	1.52 j–m	1.44 d–h	1.43 c–h	1.46 b
Mean for B	1.46 bc	1.49 c	1.46 bc	1.44 b	1.39 a	-

Note: Explanation of abbreviations as in Table 2; the results come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

. The study did not include an analysis of their variability under the influence of FT cycles; they are included for illustrative purposes only.

3.2. Bulk Density and Total Porosity of Aggregates

In the tested combinations, the bulk density (BD) of aggregates ranged from 1.30 Mg m⁻³ (combination M/F/270) to 1.46 Mg m⁻³ (combination C/NF/90;

Table 4. Bulk density and total porosity of modelled aggregates.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		Bulk Density (Mg m−3)
		30	60	90	270	365
Control	NF	1.40 d–f	1.40 d–f	1.46 g	1.42 e–g	1.44 fg	1.42 b		1.38 b
	F	1.36 b–d	1.34 a–c	1.32 ab	1.32 ab	1.38 c–e	1.34 a
Mean for AxC		1.38 bc	1.37 bc	1.39 c	1.37 bc	1.41 c	-
Manure	NF	1.38 c–e	1.30 a	1.34 a–c	1.32 ab	1.40 d–f	1.35 a		1.34 a
	F	1.36 b–d	1.32 ab	1.34 a–c	1.30 a	1.38 c–e	1.34 a
Mean for AxC		1.37 bc	1.31 a	1.34 ab	1.31 a	1.39 c	-
BxC	NF	1.39 c–e	1.35 a–c	1.40 de	1.37 b–d	1.42 e	Mean for B	NF	1.39 b
	F	1.36 b–e	1.33 ab	1.33 ab	1.31 a	1.38 c–e		F	1.34 a
Mean for C		1.38 bc	1.34 a	1.37 ab	1.34 a	1.40 c	-
Total porosity (%)
Control	NF	46.8 a–e	46.6 a–e	44.5 a	45.8 a–c	45.2 ab	45.8 a		47.3 a
	F	48.3 d–h	48.9 e–h	49.8 gh	49.6 gh	47.5 b–g	48.8 b
Mean for AxC		47.6 ab	47.8 ab	47.2 a	47.7 ab	46.4 a	-
Manure	NF	47.1 b–f	50.2 h	48.9 e–h	49.4 f–h	46.4 a–d	48.4 b		48.6 a
	F	47.9 c–h	49.4 f–h	48.9 e–h	50.3 h	47.1 b–f	48.7 b
Mean for dla C		47.5 ab	49.8 c	48.9 bc	49.8 c	46.8 a	-
BxC	NF	47.0 a–c	48.4 c–f	46.7 ab	47.6 b–d	45.8 a	Mean for B	NF	47.1 a
	F	48.1 b–e	49.2 d–f	49.4 ef	49.9 f	47.3 a–c		F	48.8 b
Mean for dla C		47.5 ab	48.8 c	48.0 bc	48.8 c	46.6 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

). Factor B modified this feature significantly (at p < 0.01; F = 26.16 **), and FT cycles reduced the aggregate bulk density. The effect of FTCs (freeze–thaw cycles) was not time-dependent, but the interaction of AxB was significant (at p < 0.01; F = 17.51 **), and significant differences were found only for the control combination. Total porosity (TP) ranged from 44.5 to 50.3%, with these two percentage values having been obtained in both the combination with the highest soil density and the combination with the lowest soil density. Similarly to the variation in density, factor B was significant (at p < 0.01; F = 26.15 **). The BxC interaction was not significant, in contrast to the AxB interaction (at p < 0.01; F = 17.51 **). Similarly to BD, the effect of factor B was significant only in combinations not supplemented with manure. The beneficial effect of FTCs on bulk density has also been noted by other authors [24,25,26]. Konrad [26] explains the changes in the physical properties of soils subjected to FTCs by changes in the original soil structure, destruction of binding forces, and re-arrangement of soil particles. Yang et al. [27] argue that loess formations are particularly susceptible to such transformations.

Recognising the beneficial effect of FTCs on porosity, Xie et al. [25] observe that it degrades the stability of aggregates. Authors who have indicated the beneficial effect of FT processes on TP also include Fouli et al. [28]. Other results were presented by Sahin et al. [29]. They tested the effect that FT processes and the addition of three doses of sewage sludge on selected physical properties. Their extensive studies observed that FTCs increased bulk density, but only for combinations with the least amount of sludge added. However, they did not notice a negative effect for variants supplemented with higher doses of sludge. The authors concluded that the negative impact of FTCs on the analysed properties can be reduced by adding the tested organic supplement. The influence of FTCs on the properties of loess soils has also been investigated by Xiao et al. [30]. They noted that, in soils with low initial bulk density, porosity initially decreases under the influence of FTCs, and then—after ten cycles—it increases. Yet another finding was presented by Leuther and Schlüter [4], who, despite the use of 19 FTCs, observed no change in porosity in soil with a silty loam texture. In the context of the effect of FTC number on soil total porosity, very extensive studies were conducted by Lv et al. [31]. The authors analysed how soil properties change, including total porosity (TP) and differential porosity (DP). They applied a very wide range of cycles: 0, 10, 20, 30, 40, 50, 60, and 70. The validity of the analysis of such a wide range of FTCs has been proven by research results. Their results confirm the complex nature of the variability in the physical properties of soil under the influence of FTCs. They observed a TP fluctuation that initially increased (relative to the unfrozen control) and reached a maximum after ten FTCs. Then, by the 50th cycle, it had decreased, approaching the control level, which it maintained after subsequent FTCs. The variability of DP was also not clear. The authors divided pores into micropores (diameter < 1 µm), small pores (1–4 µm), medium pores (4–16 µm), and large pores (>16 µm). Initially, as the number of FTCs increased, the proportion of micropores and small pores first fell and then rose. Conversely, the share of medium and large pores first rose and then fell. Viklander [32] disagrees with the claim that FTC number has an important influence on soil density. The author reports a decrease in porosity regardless of the number of FTCs.

3.3. Resistance to Dynamic and Static Water Action

The decomposition of aggregates caused by atmospheric precipitation promotes soil crusting and thus intensifies and increases the risk of erosion [33]. Moreover, the integrity and strength of aggregates can be reduced by factors such as intensive land use and climatic stress, including the alternating freeze–thaw cycles that characterise the cold seasons in temperate climates [34]. Therefore, it is important to assess whether and how these processes can change aggregates’ resistance to raindrops. The resistance of the aggregates (calculated based on the number of drops; see: Section 2.2.2) to dynamic water action ranged from 0.021 J (C/F/270&365) to 0.038 J (M/NF/365;

Table 5. Soil aggregates’ resistance to the dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		Soil Aggregates Resistance to the Dynamic Water Action [J]
		30	60	90	270	365
Control	NF	0.024 de	0.025 e	0.024 de	0.023 cd	0.024 de	0.024 b		0.023 a
	F	0.022 ab	0.023 cd	0.022 ab	0.021 ab	0.021 ab	0.022 a
Mean for AxC		0.023 bc	0.024 d	0.023 bc	0.022 ab	0.022 ab	-
Manure	NF	0.034 gh	0.037 jk	0.036 ij	0.037 jk	0.038 k	0.037 d		0.035 b
	F	0.029 f	0.035 hi	0.033 g	0.034 gh	0.036 ij	0.034 c
Mean for AxC		0.032 e	0.036 g	0.035 f	0.036 g	0.037 h	-
Mean for BxC	NF	0.029 cd	0.031 e	0.03 de	0.03 de	0.031 e	Mean for B	NF	0.030 b
	F	0.026 a	0.029 cd	0.027 b	0.027 b	0.028 c		F	0.028 a
Mean for C		0.028 a	0.03 c	0.029 b	0.029 b	0.03 c	-
		Soil aggregates resistance to the static water action (s)
Control	NF	34.0 d–g	39.0 jk	37.0 g–j	35.0 e–h	34.0 d–g	36.0 c		34.0 a
	F	30.0 a–c	35.0 e–h	33.0 c–f	31.0 a–d	30.0 a–c	31.0 a
Mean for AxC		32.0 a	37.0 cd	35.0 bc	33.0 ab	32.0 a	-
Manure	NF	34.0 d–g	41.0 k	39.0 jk	36.0 f–j	34.0 d–g	37.0 d		35.0 b
	F	29.0 ab	35.0 e–h	37.0 g–j	33.0 c–f	28.0 ab	33.0 b
Mean for AxC		31.0 a	38.0 d	38.0 d	35.0 bc	31.0 a	-
Mean for BxC	NF	34.0 cd	40.0 f	38.0 ef	36.0 de	34.0 cd	Mean for B	NF	36.3 b
	F	29.0 a	35.0 d	35.0 d	32.0 bc	30.0 ab		F	32.0 a
Mean for C		31.0 a	37.0 c	37.0 c	34.0 b	32.0 bc	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

). Factor B had a significant effect (p < 0.01; F = 157.76 **), and the frozen aggregates were less resistant than the non-frozen ones. Despite its statistical significance, the effect of FTCs seems to be small in practical terms. The effect of alternating freezing and thawing was time-dependent (p < 0.05; F = 3.11 *); in NF and F combinations alike, resistance to dynamic water action was highest in aggregates collected after 60 days, while the interaction of AxB was insignificant.

The resistance of soil aggregate models to static water action ranged from 28 s (M/F/365) to 41 s (M/NF/60; Table 5). Factor B had a significant effect (p < 0.01; F = 76.90 **). The interactions of BxC and AxB were insignificant. FT cycles had a significant effect in combinations both with and without addition of manure. Similarly to the resistance to dynamic water action, FTCs reduced the resistance of aggregates. The strength of the unfavourable impact of frost processes is seen when comparing the average value of the C/NF combination (no freeze–thaw control combinations) vs. M/F (freeze–thaw combinations with manure added). The M/FT combinations were less resistant, which proves the importance of natural processes—the “power of nature” in shaping this property of aggregates. The variability in resistance to both dynamic and static water action under the influence of cyclic freeze–thaw processes has been researched by Owczarzak et al. [35]. They tested the effect of tannery sludge and FTCs on the resistance to dynamic and static water action of soil aggregates from two loamy sand soils. For resistance to dynamic water action, the authors found no significant differences between frozen and non-frozen aggregates. The variation in resistance to static water action was less clear. The authors found that FTCs usually increased the resistance, and only in the aggregates taken from the combinations supplemented with the highest dose of sludge did water resistance decrease. The authors do not explain this puzzling variability. Similar studies have also been reported by Rząsa and Owczarzak [36]. They described a pot experiment in which sewage sludge was added to individual test combinations in doses of 2, 4, 8, 12, 16, 40, and 80 tons of dry matter per 1 hectare. The experiment was therefore of a strictly fundamental scientific nature, as most of the doses used were too high to be implemented in agricultural practice. The results indicated a protective effect of organic matter; the authors noted that FTCs reduced resistance to dynamic water action, but not at the highest dose of the additive. The supporters of the claim that FTCs weaken aggregates includes Chang and Liu [37]. They claim that, when water freezes, the porosity of the aggregate increases, and when ice melts, the porosity decreases, so these changes weaken the aggregate. Ma et al. [38] share a similar opinion and explain that freezing water exerts pressure on soil particles, displacing them and weakening the aggregate. The influence of frost processes on the effect of raindrops in the context of hypothetical acceleration of surface runoff of loess soils was studied by Gharemahmudli et al. [39]. They observed a clear increase in runoff generation and soil loss as a result of FTCs. The results obtained by Rząsa and Owczarzak [40] indicate that FTCs disrupt intermolecular bonds, thereby disturbing the aggregate. Ferrick and Gatto [41] provide this as the reason for the increase in soil susceptibility to erosion. They found an almost-two-fold acceleration of surface washout compared to the unfrozen control. Resistance to dynamic and static water action did not correlate significantly with soil bulk density or total porosity. So too, other authors [42,43,44] have reported that the water resistance of soil aggregates does not always reach a maximum at maximum bulk density. A different opinion was expressed by Lv et al. [45], who studied the soil resistance to raindrops of the Loess Plateau (China). They found that it increased linearly with the compaction of the tested soils. Other authors [46] reported that the duration of aggregate soaking increases with organic matter content in soils, especially with the increase in the hydrophobic part.

3.4. Secondary Aggregation State

The results of past studies have shown that not only is the degree of decomposition (sum of aggregates with a diameter > 0.25 mm) important, but so too is its nature (i.e., the percentage content of individual fractions of secondary aggregates) [7,8,47]. The state of secondary aggregation after the dynamic and static action of water is presented in

Table A1. Influence of experimental factors on the content of secondary aggregate fraction (%) with 5–3 mm diameter formed after dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean For A
		Dynamic Water Action (%)
		30	60	90	270	365
Control	NF	0 a	0 a	0 a	0 a	0 a	0 a		0 a
	F	0.2 ab	0 a	0.4 b	0 a	0 a	0.12 a
Mean for AxC		0.1 a	0 a	0.2 a	0 a	0 a	0 a
Manure	NF	0 a	0 a	0 a	0 a	0 a	0 a		0.3 b
	F	0.7 c	0 a	0.6 c	0 a	1.2 d	0.5 b
Mean for AxC		0.4 ab	0 a	0.3 ab	0 a	0 a	-
Mean for BxC	NF	0 a	0 a	0 a	0 a	0 a	Mean for B	NF	0 a
	F	0.5 b	0 a	0.5 b	0 a	0.6 b		F	0.3 b
Mean for C		0.2 a	0 a	0.3 a	0 a	0.3 a	-
Static water action (%)
Control	NF	0 a	0 a	0 a	0 a	0 a	0 a		0.2 a
	F	1.9 d	0 a	0 a	0 a	0 a	0.38 ab
Mean for AxC		1 b	0 a	0 a	0 a	0 a	0 a
Manure	NF	0 a	0 a	0 a	0 a	0 a	0 a		0.4 a
	F	0 a	1.3 bc	0 a	0.9 b	1.4 c	0.72 b
Mean for AxC		0 a	0.7 b	0 a	0 a	0.7 b	-
Mean for BxC	NF	0 a	0 a	0 a	0 a	0 a	Mean for B	NF	0 a
	F	1.0 b	0 a	0 a	0.5 ab	0.7 b		F	0.6 b
Mean for C		0.5 a	0.4 a	0 a	0.3 a	0.4 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

,

Table A2. Influence of experimental factors on the content of secondary aggregate fraction (%) with 3–1 mm diameter formed after dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		Dynamic Water Action (%)
		30	60	90	270	365
Control	NF	0.3 a	0 a	0.2 a	0 a	0 a	0.1 a		0.2 a
	F	0.5 a	0.6 a	0.5 a	0 a	0 a	0.3 a
Mean for AxC		0.4 a	0.3 a	0.4 a	0 a	0 a	-
Manure	NF	1.5 b	1.8 bc	2 b–d	2.1 b–d	1.8 bc	1.9 b		2 b
	F	1.8 bc	2.1 b–d	2.7 d	2.4 cd	2.1 bd	2.2 b
Mean for AxC		1.7 b	2 b	2.4 c	2.3 bc	2 b	-
Mean for BxC	NF	0.9 a	0.9 a	1.1 a	1.1 a	0.9 a	Mean for B	NF	1 a
	F	1.1 a	1.3 ab	1.6 b	1.2 ab	1 a		F	1.9 b
Mean for C		1 a	1.1 a	1.4 b	1.1 a	1 a	-
Static water action (%)
Control	NF	0.8 ab	0.7 a	0.8 ab	1 a–c	0.9 ab	0.8 a		1.0 a
	F	1.3 a–d	1.2 a–d	1.3 a–d	1.5 a–e	1.1 a–d	1.3 b
Mean for AxC		1.1 a	1.0 a	1.1 a	1.3 ab	1.0 a	-
Manure	NF	0.9 ab	0.8 ab	0.9 ab	1.2 a–d	1.7 b–e	1.1 b		1.6 b
	F	1.6 a–e	2.1 d–f	2.7 f	1.9 c–f	2.4 ef	2.1 c
Mean for AxC		1.3 ab	1.5 ab	1.8 bc	1.6 b	2.1 c	-
Mean for BxC	NF	0.9 a	0.8 a	0.9 a	1.1 a	1.3 ab	Mean for B	NF	1 a
	F	1.5 b	1.7 bc	2.0 c	1.7 bc	1.8 bc		F	1.7 b
Mean for C		1.8 a	1.8 a	2.0 a	2.1 a	2.1 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

,

Table A3. Influence of experimental factors on the content of secondary aggregate fraction (%) with 1–0.5 mm diameter formed after dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		After Dynamic Water Action (%)
		30	60	90	270	365
Control	NF	0.7 a	0.8 ab	1.2 a–d	1.1 a–d	1 a–c	1.0 a		1.1 a
	F	0.9 a–c	1.2 a–d	1.8 d–f	1.3 a–d	0.8 ab	1.2 a
Mean for AxC		0.8 a	1.0 a	1.5 b	1.2 ab	0.9 a	-
Manure	NF	1.4 a–e	1.3 a–d	1.5 b–f	2.2 fg	1.2 a–d	1.5 b		1.7 b
	F	2.1 e–g	1.6 c–f	2 e–g	2.6 g	1.3 a–d	1.9 c
Mean for AxC		1.8 bc	1.5 b	1.8 bc	2.4 c	1.3 ab	-
Mean for BxC	NF	1.1 a	1.1 a	1.4 ab	1.7 b	1.1 a	Mean for B	NF	1.3 a
	F	1.5 b	1.4 ab	1.9 bc	2.0 c	1.1 a		F	1.6 b
Mean for C		1.3 a	1.3 a	1.7 ab	1.9 ab	1.1 a	-
After static water action (%)
Control	NF	1.1 a	1.5 a–c	1.2 ab	1.2 ab	1.2 ab	1.2 a		1.3 a
	F	1.3 ab	1.5 a–c	1.3 ab	1.5 a–c	1.6 a–d	1.4 a
Mean for AxC		1.2 a	1.5 a	1.3 a	1.4 a	1.4 a	-
Manure	NF	2.2 a–e	2.4 b–e	2.2 a–e	2.3 a–e	2 a–e	2.2 b		2.5 b
	F	2.8 de	2.9 e	2.8 de	2.7 c–e	2.6 c–e	2.8 c
Mean for AxC		2.5 b	2.7 b	2.5 b	2.5 b	2.3 b	-
Mean for BxC	NF	1.7 a	2 ab	1.7 a	1.8 ab	1.6 a	Mean for B	NF	1.7 a
	F	2.1 ab	2 b	2.1 ab	2.1 ab	2.1 ab		F	2.1 b
Mean for C		1.9 a	2.1 a	1.9 a	1.9 a	1.9 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

and

Table A4. Influence of experimental factors on the content of secondary aggregate fraction (%) with 0.5–0.25 mm diameter formed after dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		Dynamic Water Action (%)
		30	60	90	270	365
Control	NF	0.7 a	1.4 a	1.3 a	1.5 a	1.5 a	1.28 a		1.4 a
	F	1.1 a	1.8 a	1.4 a	1.9 a	1.6 a	1.56 a
Mean for AxC		0.9 a	1.6 b	1.4 b	1.7 b	1.6 b	-
Manure	NF	3.5 b	3.6 b	3.2 b	3.5 b	3.8 b	3.52 b		3.8 b
	F	4.1 b	3.9 b	3.3 b	4.3 b	4.3 b	3.98 b
Mean for AxC		3.8 cd	3.8 cd	3.3 c	3.9 cd	4.1 d	-
Mean for BxC	NF	2.1 a	2.5 a	2.3 a	2.5 a	2.7 ab	Mean for B	NF	2.4 a
	F	2.6 a	2.9 b	2.4 a	3.1 b	3 b		F	2.8 b
Mean for C		2.4 a	2.7 a	2.3 a	2.8 a	2.8 a	-
Static water action (%)
Control	NF	3.3 ab	3.2 ab	3.1 a	3.1 a	3.2 ab	3.2 a		3.6 a
	F	4.2 a–d	4.1 a–c	3.6 a–c	3.8 a–c	3.9 a–c	3.9 b
Mean for AxC		3.8 ab	3.7 ab	3.4 a	3.5 a	3.6 a	-
Manure	NF	4.1 a–c	4.5 a–d	4.8 b–d	4.5 a–d	3.8 a–c	4.3 bc		4.6 b
	F	4.6 a–d	5.7 d	5.1 cd	4.7 a–d	4.1 a–c	4.8 d
Mean for AxC		4.4 b	5.1 c	5.0 c	4.6 bc	4.0 ab	-
Mean for BxC	NF	3.7 a	3.9 ab	4 ab	3.8 ab	3.5 a	Mean for B	NF	3.8 a
	F	4.4 bc	4.9 c	4.4 bc	4.3 b	4.0 ab		F	4.4 b
Mean for C		4.1 ab	4.4 b	4.2 ab	4.0 a	3.8 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

. They show the percentage share of the secondary aggregate fractions (formed after the decomposition of primary aggregates) of dimensions 5–3 mm, 3–1 mm, 1–0.5 mm, and 0.5–0.25 mm. The sum of these fractions is presented in

Table 6. The content of secondary aggregate fraction (%) with > 0.25 mm diameter formed after dynamic and static water action.

Organic Supplement A	NF/F B	Days of Incubation C					Mean for AxB		Mean for A
		After Dynamic Water Action (%)
		30	60	90	270	365
Control	NF	1.7 a	2.2 a	2.7 a	2.6 a	2.5 a	2.3 a		2.8 a
	F	2.7 a	3.6 a	4.1 a	3.2 a	2.4 a	3.2 a
Mean for AxC		2.2 a	2.9 a	3.4 a	2.9 a	2.5 a
Manure	NF	6.4 b	6.7 b	6.8 b	7.8 bc	6.8 b	6.9 b		7.8 b
	F	8.7 bc	7.6 bc	8.7 bc	9.4 c	8.8 bc	8.6 c
Mean for AxC		7.6 b	7.2 b	7.7 b	8.6 b	7.8 b
Mean for BxC	NF	4.1 a	4.5 ab	4.7 a–c	5.2 a–c	4.6 a–c	Mean for B	NF	4.6 a
	F	5.7 a–c	5.6 a–c	6.4 c	6.3 bc	5.7 a–c		F	5.9 b
Mean for C		4.9 a	5.0 a	5.6 a	5.7 a	5.1 a	-
		After static water action (%)
Control	NF	5.2 a	5.1 a	5.5 a	5.3 a	5.2 a	5.2 a		6.1 a
	F	8.7 a–d	6.7 a	6.9 a–c	6.4 a	6.5 a	7.0 b
Mean for AxC		6.9 ab	5.9 a	6.2 a	5.9 a	5.8 a
Manure	NF	6.79 ab	7.1 a–c	7.8 a–d	8.2 a–d	8.5 a–d	7.7 b		9.0 b
	F	8.3 a–d	10.8 cd	11.3 d	10.7 b–d	11.1 d	10.4 c
Mean for AxC		7.5 a–c	8.9 bc	9.5 bc	9.5 bc	9.8 c
Mean for BxC	NF	6.0 a	6.1 ab	6.6 a–c	6.8 a–c	6.8 a–c	Mean for B	NF	6.5 a
	F	8.5 a–c	8.8 bc	9.1 c	8.6 a–c	8.8 bc		F	8.7 b
Mean for C		7.2 a	7.4 a	7.9 a	7.7 a	7.8 a	-

Note: Explanation of abbreviations as in Table 2; the results given for non-frozen aggregates come from the monography of Gajewski [7]; α = 0.05/values marked with the same letters do not differ significantly.

. After both dynamic and static water action, secondary aggregates of 5–3 mm were not formed in most combinations (Table A1). The exceptions were the combinations C/F/30, C/F/90, M/F/30, M/F/90, and M/F/365 after dynamic action of water, and C/F/30, M/F/60, M/F/270, and M/F/365 after static action of water. For this fraction the effect of FTCs was significant (p < 0.01; F = 65.36 ** for DW and p < 0.01; F = 60.62 ** for SW). FT cycles caused an increase in the content of the 5–3 mm fraction. Meanwhile, the 3–1 mm fraction was present in most combinations (Table A2). After dynamic water action, in the combinations in which their presence was detected, it ranged from 0.2% to 2.7% (C/NF/90; M/F/90). However, after static action of water, it was within the limits of 0.7% to 2.7% (C/NF/60; M/F/90). For both types of water action, the effect of factor B was significant (p < 0.01; F = 65.36 ** for DW and p < 0.01; F = 31.81 ** for SW), with a higher share of this fraction having been found in the combinations subjected to FTCs. The share of secondary aggregates of 1–0.5 mm ranged from 0.7% (C/NF/30) to 2.6% (M/F/270) after dynamic water action, and from 1.1% (C/NF/30) to 2.9% (M/F/60) after static water action (Table A3). FTCs significantly increased the share of this fraction (p < 0.01; F = 11.41 ** for DW and p < 0.05 F = 5.26 * for SW). The share in secondary aggregation was largest in the 0.5–0.25 mm fraction (Table A4). After dynamic water action, it ranged from 0.7% (C/NF/30) to 4.3% (M/F/270 and M/F/365), and after static action it ranged from 3.1% (C/NF/90; C/NF/270) to 5.7% (M/F/60). The variability in the share of this fraction formed after both dynamic and static water action depended on factor B (p < 0.05; F = 5.32 * for DW and p < 0.01; F = 8.73 ** for SW) and was higher for aggregates subjected to FTCs. The interaction of AxB and BxC in shaping the variability of the percentage share of the above-described secondary aggregate fractions formed after dynamic and static water action was non-significant in most cases. The aggregate fractions described above constitute the sum of aggregates with a diameter >0.25 mm, which Rząsa and Owczarzak [8] define as the secondary aggregation state. They identify this as a key determinant of the agronomic quality and structural stability of soil. The variability of this property is shown in Table 6. After dynamic water action, it was lowest (1.7%) in the C/NF/30 variant and highest (9.4%) in the M/F/270 combination. The disintegration of primary aggregates after static water action resulted in secondary aggregation, in which the share of the >0.25 mm diameter fraction was smallest in the C/NF/60 variant and largest in the M/F/90 variant. These shares were, respectively, 5.1% and 11.3%. After both types of water action, the effect of factor B was significant (p < 0.01; F = 14.42 ** for DW and p < 0.01; F = 19.23 ** for SW), and subjecting the aggregates to FTCs increased the share of the >0.25 mm fraction, thus positively affecting this characteristic. The interactions of AxB and BxC were negligible. These results are partially confirmed in the article of Li and Fan [48], who analysed the effect of FTCs on the size distribution of water-resistant soil aggregates. They found increases in the share of the 1–0.5 mm and 0.5–0.25 mm fractions, but decreases in the >5 mm, 5–3 mm, 3–2 mm, and 2–1 mm fractions. An increase in the share of water-resistant aggregates of diameter <0.5 mm and a decrease in the share of aggregates of >4.75 mm under the influence of FTC was noted by Edwards [49]. The variation in the fraction of aggregates with a diameter >0.25 mm as a response to FTCs was also studied by Zhang et al. [50], who made similar observations to those in this paper (i.e., an increase). They, unlike Li and Fan [48] and Edwards [49], found an increase in the share of all fraction classes. Interestingly, they state that the number of FTCs correlated positively with aggregate stability. In the context of the relationship between the number of FTCs and aggregate stabilities, Lersch [51] reached different conclusions, finding it to be at its highest level after 2–3 cycles. The beneficial effect of FTCs on aggregate stability has been noted by Øygarden [52], Park et al. [53], and by Žabenska and Dumbrovský [54]. Rząsa and Owczarzak [36] noted a very clear beneficial effect of FTCs on the structural stability of soils of various textures, explaining it as resulting from the complicated nature of their movement during alternating FTCs. The literature is dominated by papers whose authors report a negative impact of FTCs on structural stability. For example, Wang et al. [55] analysed 96 soils in this respect and found a negative effect of FTCs on aggregate stabilities. The variability in structural stability due to FTCs has also been assessed by Kværnø and Øygarden [56]. They tested the stability in a rain simulator and in a wet-sieving apparatus in an analysis of soils with silt and clay textures. They found that the stability of both soils fell as a result of FTCs, and this effect was particularly pronounced in the combinations exposed to the rain simulator and greater in the silty soil. They also found a negative correlation between the number of FTCs and the stability of aggregates. A similar opinion was expressed by Bajracharya et al. [57]. Supporters of the claim that FTCs negatively affect structural stability include Staricka and Benoit [58], who, of the 96 soils they examined, observed such an effect in 85 cases. The effect of freezing temperatures on aggregate stabilities has been tested by Oztas and Fayetorbay [59], who found that, as temperature fell, so too did the stability of aggregates. The influence of FTCs on soil properties, including structural stability, is also reported by Sun et al. [60] to be a complex mechanism. They state that the variability in structural stability, and thus whether and how FTCs will affect it, is due to, inter alia, the number of cycles and soil texture. As they note, soil responses may vary greatly depending on these factors. This requires that researchers in the field constantly verify the conclusions that have already been drawn, while also explaining why conclusions sometimes diametrically contradict one another.

3.5. Climate Change and Soil Freezing Dynamics

Soil freezing is influenced strongly by both air temperature and insulation by the snowpack [61]. Some authors suggest that a trend towards decreasing snowfall has already commenced and will likely further intensify in a future climate [62]. The forecasted decrease in snowfall in Poland has also been reported by Somorowska [63]. This should lead to thinner and more intermittent snow cover, greater fluctuations in soil temperature, and thus more frequent FT cycles [64]. Ghazi et al. [65], in turn, predicted a significant decrease in the number of days with frost in Poland over the coming decades. Based on research conducted in China, Peng et al. [6] found that the mean annual area extent of seasonal soil freeze/thaw state decreased significantly for completely frozen ground, while the area extents of partially frozen and unfrozen ground both increased. According to the authors, those results will be useful for advancing the understanding of seasonally frozen ground dynamics, such as the impacts on ecosystems and hydrological processes. These contradictions between authors require future research in this area.

4. Conclusions

Cyclic freeze–thaw (FT) processes significantly modified most of the analysed properties of soil aggregates. This effect was usually not modified by organic supplementation, nor by its duration of action. FT cycles reduced the bulk density of soil aggregates.

FT cycles reduced the resistance of soil aggregates to dynamic and static water action. Despite the lower resistance to the destructive action of water, the decomposition of primary aggregates resulted in a more favourable size distribution of secondary aggregates in aggregates subjected to FT cycles.

The initial hypotheses were partially confirmed. FTCs significantly influenced the tested structural properties, but their influence was not modified by time or organic additives.

The unfavourable influence of FTC processes on the resistance of soil aggregates to the dynamic and static action of water indicates the validity of applying measures to limit this effect.

Future studies analysing the effect of FTCs on structural parameters should be extended to soils with different (than tested) soil properties, such as texture, organic matter content, soil genesis, and soil morphology.