Effect of Paulownia and Buckwheat Intercropping on Soil Microbial Biodiversity, Dehydrogenase Activity, and Glomalin-Related Soil Protein

Abstract

Intercropping of trees and classical crops has been proposed as a practice to help adapt to climate change and protect soil against erosion. However, the effects of intercropping on soil biology are not sufficiently quantified. Therefore, the aim of this study was to evaluate microbiological changes in the soil resulting from the intercropping of Paulownia and buckwheat. A field experiment, involving an intercropping and control no-tree variant, was conducted from 2019 to 2022 with a plot size of 30 m². Buckwheat rhizosphere soil samples were collected twice in both 2021 and 2022 in order to evaluate the effects of intercropping on a range of parameters describing soil microbiome status: abundance of microorganisms, bacterial and fungal community structure (using Illumina MiSeq sequencing), dehydrogenases (DHA) activity, and total glomalin-related soil proteins (T-GRSP). In addition, the colonisation of buckwheat roots by fungi, yield, and biometric traits of the plant were determined. Next-generation sequencing showed that Actinobacteria, Proteobacteria, and Acidobacteria were dominant in the microbiome of every variant of the experiment, regardless of the crop. In contrast, the mycobiome was dominated by fungi classified as Ascomycota and Mortierellomycota. This observation corresponded to an increase in buckwheat yield in intercropped plots. Biometric traits, namely buckwheat yield and total kernel weight per plant, showed higher values when buckwheat was intercropped with Paulownia compared to the control. DHA activity was stimulated by intercropping at the first sampling date, whereas glomalin concentration and abundance of microorganisms were not dependent on the cropping systems tested. This study shows that tree-based intercropping (TBI) systems promote a more diverse soil microbial community and function than conventional agriculture. Our results also suggest that TBI positively impacts buckwheat biometric traits, supporting its implementation in rural landscapes. The yield under intercropping cultivation amounted to 0.65 t ha⁻¹, while in control sites it was 0.53 t ha⁻¹. The total abundance of bacteria under intercropping cultivation was higher compared to monoculture in 2021 at the first term of sampling (4.3 × 10⁴) and in 2022 in the second term of soil sampling (4.6 × 10⁴).

1. Introduction

Intercropping of trees and cultivated annuals, i.e., the integration of trees into the agricultural landscape in the form of tree-based intercropping systems (TBI), is part of a system called agroforestry, which involves the cultivation of forest trees and shrubs alongside agro- and zootechnical activities within the same land area. Agroforestry constitutes a system that ensures a continuous supply of organic material to the soil [1].

The development of agroforestry systems seems particularly relevant in the context of the following: counteracting the negative effects of climate change, water and wind erosion of the soil, leaching of nutrients and plant protection products from the soil, loss of biodiversity, large areas of marginal agricultural land, and the increasing demand for bioenergy and organic products [2].

Buckwheat (Fagopyrum esculentum Moench) is a crop in high demand around the world for its nutritional, economic, and pharmaceutical qualities. To ensure food and nutrition security in the face of global climate change, this pseudo-cereal is a competent alternative to staple crops (rice, wheat, rice, and maize). Buckwheat also has favourable agroecological traits, such as enabling weed control due to the inhibitory effect of root secretions on the development of couch grass (allelopathy), effective protection of the soil against water erosion, high capacity to assimilate nitrogen and phosphorus from the soil, phytosanitary effects on the soil through effective control of nematodes (reduction of potato cyst nematode populations), and benefits to insects during flowering [3].

Paulownia spp. are fast-growing deciduous trees with a high tolerance for variable soil and climatic conditions and low water requirements [4]. Paulownia Clon in Vitro 112 is a tree developed under laboratory conditions by crossing and cloning Paulownia elongata and P. fortunei. The hybrid is considered suitable for biomass production and revegetation [5]. Furthermore, Valdivia et al. (2012) [6] found that trees of the genus Paulownia have an advantage over most other tree species in that they have a sparse canopy with late foliage and leaf drop, thus not blocking sun rays reaching the fields during periods where many commonly grown crops, including cereals, require sunlight the most. Intercropping trees and annual crops expand the agroecosystem biodiversity, promoting biological life in the soil [2]. The diversity of soil microorganisms is higher within agroforestry systems [7]. Knowledge of both the activity and diversity of soil microorganisms enables researchers to assess the state of the soil, which is beneficial for agriculture and ecology [8,9]. A diverse soil microbial community is crucial to agroecosystem productivity and serves as a high-sensitivity indicator of changes in soil organic matter. Even modest decreases in microbial abundance can be detrimental to soil ecosystems, as soil microorganisms are responsible for various biogeochemical processes, organic matter mineralisation, and nutrient cycling in the environment [10]. Microorganisms are a principal source of soil enzymes, which act as mediators and biological catalysts for various soil functions, such as organic matter breakdown, the release of inorganic nutrients (C, N, P, S) for plant growth, nitrogen fixation, nitrification, denitrification, detoxification of xenobiotics, and stabilisation of soil structure [11].

Enzyme-related activity in the soil is one of the biological properties used as a soil quality indicator due to its interrelationship with soil biology. In addition, it is a highly sensitive and easily measured indicator of soil biological management, microbial activity, soil fertilisation, degree of pollution, and anthropopressure. Dehydrogenases (DHGs) are key components of soil enzyme testing, as they determine the correct sequence of all biochemical pathways in soil [11]. The literature on the effects of intercropping on soil enzymes is rather limited. However, the available studies provide ambiguous results. Some researchers report an increase in soil enzyme activity, including dehydrogenases, as a result of chestnut tree-tea intercropping [12], whereas other authors have observed a negative impact of maize-peanut intercropping on dehydrogenases [13]. Therefore, any new data on the effects of intercropping on dehydrogenases are valuable.

Glomalin-related soil protein (GRSP) is a glycoprotein produced by arbuscular mycorrhizal fungi (AMF) that plays a vital role in regulating soil organic carbon (SOC) dynamics and maintaining soil aggregate stability Wang et al. (2016) [14]. Acting as a key agent in soil aggregate formation, GRSP contributes to improving soil structure and facilitates the accumulation of carbon (C) and nitrogen (N) [15]. Research indicates that GRSP can constitute up to 27% of the organic C content in soil and as much as 53% in peat, highlighting its significance as a major source of carbon in the active SOC pool. Additionally, glomalin may play a role in soil fertility by forming complexes with iron [16]. Moreover, its ability to bind with potentially toxic elements suggests a potential function in mitigating soil contamination [17].

Determining the relationship between the soil humus layer and the microbiome is possible through advances in sequencing technology [18]. Sequencing of DNA with high throughput is a particularly useful tool for understanding microbe–soil–plant relationships and visualising patterns of microbial co-occurrence [19]. Furthermore, it can simultaneously examine the entire community of microorganisms present in a soil sample. The diversity of bacteria and fungi is a major driver of fundamental metabolic processes in a dynamic environment, such as soil. It is thus important to explore the biodiversity of the soil environment as one of the elements that can influence the development of management strategies for terrestrial ecosystems to maintain their integrity, function, and long-term stability. Metagenomics can also be used to identify key microbial taxa that have a significant impact on the soil community [18,19]. The aim of this study was to capture microbiological changes in the soil environment during the intercropping of Paulownia and buckwheat. This study aimed to assess the soil microbiome structure, including the abundance of microorganisms, and to analyze the diversity and abundance of bacterial and fungal microbiomes. Additionally, this study aimed to evaluate the colonisation of buckwheat roots by fungi and to analyze the yield and biometric traits of the plants.

It was hypothesized that intercropping of Paulownia and buckwheat would lead to an increase in soil microbial diversity, which may enhance dehydrogenase (DHA) activity and positively impact the yield and biometric traits of buckwheat compared to the control.

2. Materials and Methods

2.1. Study Sites and Soil Sampling

In 2019, a rigorous field experiment with buckwheat (Fagopyrum esculentum Moench) and Paulownia (Clon in Vitro 112; P. elongata × P. fortunei) was set up in a randomised block (Figure 1) design at the Research and Teaching Station of Wrocław University of Environmental and Life Sciences (Wroclaw, Poland, voivodship Lower Silesia; 51°07′00″ N, 17°10′ E, 121 m above sea level). The intercropping of buckwheat between Paulownia (AP) trees was tested. The experiment also included control sites (AK), i.e., buckwheat plots grown without Paulownia, and consisted of 5 replications. Paulownia seedlings were planted on 3 May 2019. In spring, on 19 May 2020, we carried out technical pruning, where we established the main shoot (future trunk). Buckwheat (Kora cultivar) was sown on 11 May 2021 and 4 May 2022 at a density of 250 seeds m². The area of the buckwheat plot was 30 m². The trees were planted in rows of 5. In the plot, the row spacing was 5 m, and the tree-to-tree spacing was 4 m. The tree density was typical for growing Paulownia in round wood. For microbiological and enzyme analyses, we sampled rhizosphere soil from buckwheat grown without Paulownia (AK) and from buckwheat grown in an inter-row with Paulownia (AP). The soil was sampled twice before buckwheat sowing (T1) and at the buckwheat flowering stage (T2) in the first (2021) and second (2022) years of the study. Soil samples for analysis of each cultivation variant were collected from 5 replicates.

2.2. Chemical-Physical Soil Analysis

The rigorous field experiment was conducted on soil classified as humic ordinary alluvial soils according to the Polish Soil Classification (SGP6) or Eutric Fluvisols (Humic) according to the IUSS Working Group WRB (2022) [20]. According to agricultural evaluation, the soil was identified as F–IV b-a class (moderately poor) [20,21]. Soil samples for bioavailable forms of macronutrients (P, K, Mg) and mineral nitrogen content analysis were sourced before sowing buckwheat and after the end of vegetation from several randomly selected locations in the field for each site to obtain an average sample. Soil samples were sourced from the 0–30 cm layer using a soil sampler tool, air-dried, ground using a porcelain pestle and mortar, and sieved to <2 mm. A portion of each sample was then finely ground for analysis. Soil testing included the determination of basic soil characteristics and properties, such as soil morphology, soil classification, land use value, pH, and macronutrient content, including mineral nitrogen.

We measured the following soil properties:

• Content of bioavailable forms of P, K—assessed using the Egner–Riehm (DL) method;
• Soil Ca content—determined using the Sheibler method;
• Available magnesium content—measured using the Schachtschabel method;
• pH in distilled water and in 1 M KCl—determined using a potentiometric pH metre;
• Nitrogen content—determined using the modified Kjehdal method (total nitrogen determination).

The soil pH and microelement content are presented in

Table 1. Soil pH and macroelement content of the soil profile (0–30 cm) in 2019 and 2022.

Detailed Determination	2019		2022
	AK	AP	AK	AP
pH in KCl (soil pH)	6.2 (slightly acidic)	5.8 (slightly acidic)	5.7 (slightly acidic)	5.7 (slightly acidic)
Phosphorus P2O5 (mg 100 g−1 of soil)	48.5 (very high)	27.6 (very high)	22.6 (very high)	18.8 (high)
Potassium K2O (mg 100 g−1 of soil)	33.5 (very high)	33.5 (very high)	28.9 (very high)	24.1 (medium)
Magnesium Mg (mg 100 g−1 of soil)	7.6 (very high)	7.6 (very high)	6.7 (medium)	5.4 (medium)
N min. (kg ha−1)	43.8 (very low)	48.0 (very low)	56.8 (low)	21.4 (very low)

. The soil pH at both the intercropped site and the control site was slightly acidic in both years of the study. Phosphorus and potassium contents at both sites were very high in the year the experiment started. At the end of the 2022 season, in the intercropping variant, the potassium content was reduced to ‘medium’, while phosphorus was reduced to ‘high’. In terms of magnesium content, it was very high in the soil at both sites in the year the experiment started, but by the end of 2022, it had decreased to medium at both sites (Table 1). Over time, the mineral nitrogen content in the soil decreased under the intercropping of Paulownia and buckwheat, while the monoculture cultivation of buckwheat contributed to an increase in mineral nitrogen content in the 30-cm soil layer.

2.3. Agrotechnology of the Experiment

Prior to starting the experiment, following the 2018 winter triticale harvest, the field was ploughed with a subsoiler plough, and the soil was kept in a weed-free state using a cultivating unit (rotary harrow with crushing roller). In autumn, deep pre-winter ploughing was carried out to a depth of 25 cm. In the year we started our experiment, i.e., in spring 2019, the soil was ploughed again using a subsoiler plough and further amended using a cultivating unit. After planting the Paulownia seedlings, the soil in the inter-rows was loosened with a rotary harrow. In the following year, spring of 2020, the soil was ploughed shallow and then amended using a cultivator before sowing buckwheat. Following technical tree pruning in May 2020, the paths in the experiment were tended mechanically. After harvesting the buckwheat with a plot harvester, the soil was kept weed-free by harrowing (rotary harrow). In 2021, a cultivating unit equipped with a string roller was used to cultivate soil after winter. By May 2021, harrowing had been carried out twice as a weed control measure. After sowing the buckwheat, the paths were mechanically cultivated several times until harvesting using a combined harvester. Before winter, the soil was kept weed-free by harrowing with a rotary harrow. Tillage in 2022 included harrowing the field after winter with a cultivator, followed by ploughing with a subsoiler plough to a depth of 12 cm. Before sowing buckwheat, the soil was treated with a cultivating unit consisting of a rotary harrow and a crushing roller. The sowing of buckwheat was done with a plot seeder. Paths between the plots were tended during the growing season. No mineral or organic fertilisers or pesticides were used in the experiment.

2.4. Weather Conditions

As part of the experimental design, we considered meteorological weather data during the experimental period (

Table 2. Weather conditions and HTC in 2019–2022, provided by the meteorological station (AsterMet) in Swojczyce (Wrocław).

Month	Temperature (°C)					Rainfall (mm)					HTC (K)
	2019	2020	2021	2022	Mean 1990–2020	2019	2020	2021	2022	Mean 1990–2020	2019	2020	2021	2022
IV	10.8	9.5	7.5	6.4	9.6	24.2	6.4	32.5	32.7	32.8	0.7	0.2	1.4	1.5
V	12.1	11.6	12.4	12.7	14.3	76.8	77.2	59.0	20.9	58.9	2.1	2.1	1.5	0.4
VI	22.1	18.5	19.8	20.1	17.8	27.0	94.5	38.4	39.7	74.6	0.4	1.2	0.6	0.6
VII	19.3	19.5	20.5	20.7	19.7	44.5	53.2	41.1	132.5	86.6	0.7	1.3	0.6	2.1
VIII	20.3	22.5	17.3	17.7	19.2	59.8	6.2	11.0	92.5	63.6	0.9	0.1	0.2	1.4
IX	14.4	14.7	15.0	14.8	14.2	42.0	92.9	14.5	82.0	50.6	1.0	2.1	0.3	2.1
X	10.4	10.7	9.5	9.1	9.3	32.1	100.5	10.2	8.6	40.8	3.2	3.0	0.3	0.2
Mean/sum (IV–X)	15.6	15.3	14.6	14.5	14.5	306.4	430.9	206.7	408.9	377.0	-	-	-	-

). Meteorological conditions were analysed using data provided by the meteorological station (AsterMet) in Swojczyce (Wrocław, Poland). We used Selyaninov’s hydrothermal coefficient (HTC) to describe the impact of weather conditions on plant development, using the following formula:

K = P/(0.1 × T)

where K is Selyaninov’s hydrothermal coefficient, P is the total rainfall per month, and T is the sum of daily average temperatures per month [22].

2.5. Characteristics of Soil Microbiological Properties

2.5.1. DNA Extraction, PCR, and Illumina Amplicon Sequencing

Total DNA for metagenomic analysis was extracted from the soil samples using the PowerSoil^® DNA Isolation Kit (Qiagen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Meta-barcoding analysis of the microbial community was performed based on the following variable regions: V3–V4 of the 16S rRNA gene for bacteria and the hypervariable region of ITS1 for fungi. Specific primer sequences 341F and 785R for 16S rRNA analysis and ITS1FI2 and 5.8S for ITS1 analysis were used to amplify the selected regions and prepare the library. All steps, including amplification, indexing, and library quantification, were performed according to the protocol “Metagenomic Sequencing Library Preparation” (Illumina, San Diego, CA, USA). Sequencing was performed on the MiSeq, 2 × 300 PE (paired-end) in order to obtain at least 50,000 read pairs per sample.

2.5.2. Bioinformatic Analysis

Bioinformatics analysis ensuring read classification was carried out with the QIIME 2 software package based on the reference sequence databases Silva 138 (for bacteria) and UNITE v8.2 (for fungi). The DADA2 package was also used, which allowed for the specification of sequences of biological origin from those newly created in the sequencing process. This package was also used to isolate unique sequences of biological origin, i.e., amplicon sequence variant (ASV).

Metagenomic analysis of fungal populations was performed using MiSeq Reporter (MSR) v2.6 software. The analysis consisted of the following stages: reading quality control, initial data processing using the Cutadapt 4.0 tool, selecting unique ASV and operational taxonomic unit (OTU) sequences, and assigning taxonomies to the generated sequences based on reference databases. In addition, α-diversity indices (Shannon diver-sity H′, Simpson index D, and observed ASV/OTU) were assessed from the Illumina MiSeq results. Raw data files are available at the Sequence Read Archive (SRA), NCBI and data information can be found at the BioProject PRJNA1073284.

2.5.3. Quantification of Cultivable Bacteria and Fungi

To determine the colony-forming units of bacteria and fungi in the test soils (AK, AP, T1, and T2), we performed surface culture with successive dilutions on medium developed by Bunt and Rovira (1955) and Martin (1950), respectively [23,24]. Analyses were performed in 3 average samples taken from each test plot. The results are presented in colony-forming units in 1 g of soil (CFU × g⁻¹). Actual results were subjected to a two-way analysis of variance (ANOVA) with a post-hoc Tukey HSD test.

2.5.4. Soil Dehydrogenase Activity and T-GRSP Concentration

Dehydrogenase activity (DHA) was analysed according to PN-ISO 23753-1 (2008) using 2,3,5-triphenyltetrazolium chloride (TTC), which is reduced via enzymatic activity to triphenylformazan (TPF). We mixed 5 mL of a 3% solution of TTC (Merck, Darmstadt, Germany) with soil in a 1:1 ratio (g:mL) and incubated the mixture at 25 °C for 16 h. The released triphenylformazane (TPF, Merck) was extracted using methanol and spectrophotometrically analysed (RayLeight, VIS 723G, Beijing Rayleigh Analitical Instrument Co.Ltd., China) at 485 nm. A control sample consisted of soil without added TTC. The amount of TPF released was expressed as micrograms per gram of dry soil per hour.

Total glomalin-related soil proteins (T-GRSP) were also determined and extracted from soil samples using the method developed by Wright and Upadhyaya (1999) [25] with alterations. Soil samples (10 g) were flooded with 0.05 M citrate buffer (pH 8.0) and autoclaved at 121 °C for 60 min. Extraction was repeated several times until the organic fraction was completely eluted from the soil. Following each autoclaving, the supernatant containing T-GRSP was poured off and centrifuged at 10,000 rpm for 10 min. Soil samples were covered with sterile buffer and autoclaved again. The extracts, collected after each heating and centrifugation, were combined and stored at 4 °C until analysis. The T-GRSP content in the supernatants was quantified via the Bradford method using bovine serum albumin (Sigma-Aldrich, Inc., Saint Louis, MO, USA) as a standard.

2.5.5. Colonisation of Buckwheat Roots by Filamentous Fungi

We assessed the degree of colonisation of buckwheat roots by potential pathogenic fungi. Filamentous fungi were isolated from buckwheat roots collected at flowering (T2) and their genus and/or species were determined based on their morphological characteristics [26,27]. For this purpose, root sections that had been previously disinfected with 0.5% NaOCl (for 15 min, rinsed 4 times with sterile distilled water) were plated on potato dextrose agar (PDA) medium (BTL, Department of Enzymes and Peptones, Łódź, Poland) supplemented with streptomycin (30 µg mL⁻¹). The plates were incubated at 28 °C for 72 h. The fungi were then reisolated onto the PDA medium. Ten replicates (60 inoculations each) were performed for each sample. The results are presented as the percentage of root colonisation by fungi.

2.5.6. Analysis of Buckwheat Biometric Traits and Yield

Before harvesting buckwheat, we randomly sampled 10 plants from each plot. The samples were sourced from the middle rows to mitigate border effects. Biometric analysis included plant height, number of branches, number of twigs, number of inflorescences, number of kernels, and weight of full kernels per plant. The yield per plot was adjusted to a 15% moisture content and converted to tonnes per hectare. To assess the effect of the cultivation method and time by year of study, we performed a two-factor ANOVA using Statistica 13.1. When differences between means were significant, they were compared using the Tukey HSD test, with a significance level of α = 0.05. Actual DHA and T-GRSP results were subject to two-way ANOVA with a post-hoc Tukey HSD test. Homogenous groups were established from largest to smallest. Additionally, a Pearson correlation was performed to assess the relationships between biological parameters (number of bacteria, fungi, DHA, and GRSP).

3. Results

3.1. Diversity and Structure of the Microbiome and Mycobiome

The relative abundance of bacteria at the cluster level is shown in Figure 1. The following nine bacterial phyla were most prevalent in the rhizosphere soil samples: Actinobacteriota, Proteobacteria, Chloroflexi, Gemmatimonadota, Bacteroidota, Firmicutes, Myxococcota, Verrucomicrobiota, and Planctomycetota. Other phyla usually did not exceed 5% relative abundance (Figure 2). In all combinations examined, the Actinomycetes were the largest group with up to 45% (2022/T2/AK). Proteobacteria were also a significant group in all combinations (about 20–30%, depending on the combination). Samples collected in 2021 had a lower relative abundance of Actinobacteria based on 16S rRNA gene fragment sequencing analysis for bacteria compared to samples from the 2022 collection in all combinations. 16S rRNA sequencing also showed that the intercropping variant in the 2022 collection samples had no effect on Proteobacteria, regardless of the sampling date. In contrast, for the 2021 samples, there was an increase in this population for the intercropping (AP) samples collected at the first sampling time compared to the AK combination. We observed no such difference at the second sampling date (T2).

There were four dominant fungal phyla: Ascomycota, Basidiomycota, Mortierellomycota, and Mucoromycota. Unclassified fungi also formed a significant part of the population (Figure 3). In all combinations studied, Ascomycota was the largest group, accounting for up to about 65% (2022/T2/AK). The relative abundance of Ascomycota was lowest in 2021 samples collected from AP at T2. In 2022, there were slightly fewer Ascomycota in AP than in AK on both sampling dates. Analysing different variants in the plot experiment, the Basidiomycota phylum was found to be the least abundant in AP during the first sampling (T1) compared to the other variants and sampling dates in 2021. In contrast, the Mortierellomycota phylum was most abundant in the 2022 samples collected on the first sampling date (T1) for the AK combination. The Mucoromycota phylum was much less abundant than the other three main fungal phyla, but its proportion increased significantly in the AP variant during the second sampling of 2022 (T2). The largest group of unclassified fungi was observed in the soil samples collected in 2021 on the second sampling date (T2) in AP (approximately 30% of the total population of the sample).

The top 15 genera in the bacteria communities under study are presented in Figure 4. Sequences assigned to Nocardioides were most abundant in all communities throughout the experiment, accounting for 2.10–6.23% of the total high-quality sequences. Candidatus_Solibacter was one of the dominant genera (2.43% relative abundance) in the control sample sourced on T1 in 2022. Pseudarthrobacter was abundant in AP samples from T2 in 2021 and 2022, as well as in both samples from T2 in 2022 samples, whereas Streptomyces was characteristic of the rhizosphere microbiomes in samples from T2 in 2022.

Heatmap analysis of the relative abundances of the most abundant genera showed clear differences in the fungal community structures between all samples (Figure 5). Genera whose abundance was above 2% in all samples were Exophiala, Fusarium, and Mortierella. Mortierella was the dominant genus, accounting for 9.69–23.56% of the total high-quality sequences. Botryotinia was abundant in the AK sample from T1 in 2021, whereas the genus Penicillium was abundant in the AK sample from T2 in 2022.

3.2. Total Abundance of Cultivable Bacteria and Fungi

Bacterial counts ranged from 1.6 × 10⁴ to 4.6 × 10⁶ CFUs in 1 g of soil, and fungal counts ranged from 1.0 to 5.0 × 10⁴ CFUs in 1 g of soil (

Table 3. Total abundance of bacteria and fungi in the field experiment.

Year	Period	Samples	Total Bacterial Count (CFU g−1)	Total Fungal Count (CFU g−1)
2021	T1	AK	2.4 ± 0.05 × 104 c	3.8 ± 0.05 × 104 abc
		AP	4.3 ± 0.05 × 104 c	3.3 ± 0.08 × 104 abc
	T2	AK	2.1 ± 0.17 × 106 ab	1.6 ± 0.01 × 104 cd
		AP	1.6 ± 0.18 × 106 b	1.0 ± 0.02 × 104 d
2022	T1	AK	2.9 ± 0.45 × 106 ab	1.8 ± 0.02 × 104 bcd
		AP	2.4 ± 0.51 × 106 ab	1.7 ± 0.03 × 104 cd
	T2	AK	3.1 ± 0.14 × 106 ab	5.0 ± 0.02 × 104 a
		AP	4.6 ± 0.27 × 106 a	3.8 ± 0.02 × 104 ab

AK—control site; AP—intercropping; T1, T2—the first and second soil sampling date; Values are the mean of four replicates of each sample. Values followed by different letters in columns indicate significant differences according to two-way ANOVA with post-hoc Tukey HSD test, ±SE—standard error.

). Compared to the control, intercropping (AP) did not significantly affect the number of bacteria or fungi in either the first or second year of the study. However, the number of CFUs was influenced by the sampling date and the year of study. At both T1 and T2 for both cultivation methods, a significant increase in bacterial counts was observed in the second year of the study compared to 2021. In 2021, fungal CFUs showed no statistically significant differences between the control sample and the soil sample where buckwheat was grown with Paulownia, on either sampling date. In contrast, there was a statistically significant increase in the number of fungal CFUs in T2 of 2022 compared to the previous year (Table 3).

3.3. Soil Biological Activity

Soil biological activity was determined in the soil samples based on DHA activity and T-GRSP. The results are shown in

Table 4. Dehydrogenase activity (DHA) and total glomalin concentration (T-GRSP) in air-dried soil.

Year	Period	Samples	DHA (µg TPF g−1 h−1)	T-GRSP (µg g−1)
2021	T1	AK	3.8 ± 0.34 c	1699.0 ± 0.53 a
		AP	7.9 ± 0.16 b	1522.9 ± 1.82 a
	T2	AK	2.9 ± 0.16 c	1492.1 ± 2.10 a
		AP	3.0 ± 0.14 c	1533.2 ± 1.91 a
2022	T1	AK	7.7 ± 0.53 b	1460.2 ± 1.13 ab
		AP	12.8 ± 0.40 a	1471.6 ± 1.33 ab
	T2	AK	2.0 ± 0.13 c	1088.0 ± 0.89 c
		AP	2.5 ± 0.08 c	1233.2 ± 1.10 bc

AK—control site; AP—intercropping; T1, T2—the first and second soil sampling dates; DHA—dehydrogenase activity expressed as the amount of released μg triphenylformazan (TPF) per hour per gram of soil; GRSP—glomalin-related soil protein content per gram of air-dried soil. Values followed by different letters in columns indicate significant differences according to two-way ANOVA with post-hoc Tukey HSD test, ±SE—standard error.

and Table S1.

Table S1. Summary of two-way analysis of variance (ANOVA) results testing the effects of cultivation (monoculture and intercropping), year (2021 and 2022) and sampling period (T1 and T2) on total bacterial count, total fungal count, dehydrogenase activity and T-GRSP. Data shown represent F-values and significance levels for each factor and interaction.

Dehydrogenase activity in the intercropping (AP) and control (AK) samples ranged from 2.0 to 12.8 µg TPF g⁻¹ h⁻¹. DHA activity was higher in the intercropping samples compared to the control samples, but statistically significant differences were observed for the first sampling date (T1). The highest DHA activity was observed in the intercropping samples in 2022 from the first sampling date (AP, T1). At the flowering stage of buckwheat (T2), enzyme activity decreased, regardless of the cultivation method.

The T-GRSP concentration ranged from 1039.2 to 1699.0 µg g⁻¹. The concentration of glomalin depended on the sampling date and not on the cultivation method. The highest concentration of T-GRSP was found in 2021 in the control sample (AK) from the first sampling date (T1). The following year, T-GRSP concentrations decreased compared to those in samples sourced on the first sampling date (T1).

3.4. Correlation of Microbiological Analysis (Bacteria, Fungi, DHA, and T-GRSP)

In the studies, a negative correlation was found between the number of bacteria and the glomalin content, which means that as the number of bacteria increased, the glomalin content in the samples decreased, and vice versa—that is, the fewer bacteria, the higher the glomalin content. Glomalin is a protein produced by mycorrhizal fungi, which plays a crucial role in soil structure, influencing its cohesion and water retention capacity. Additionally, a negative correlation between glomalin content and the number of fungi was also observed, indicating that higher glomalin content is associated with a lower number of fungi in the soil. In practice, these correlations indicate a complex interaction between microorganisms in the soil, where bacteria, fungi, and organic compounds, such as glomalin, have interdependent relationships (

Table 5. Correlation of microbiological analysis (bacteria, fungi, DHA, and T-GRSP).

	Correlations (Microbiology) The Indicated Correlation Coefficients Are Significant with p < 0.05000, N = 40
Variable	SD	Bacteria (CFU g−1)	Fungi (CFU g−1)	DHA	T-GRSP
Bacteria (CFU g−1)	1,763,107	1.00	0.09	−0.19	−0.54
Fungi (CFU g−1)	16,296	0.09	1.00	−0.21	−0.36
DHA	4	−0.19	−0.21	1.00	0.24
GRSP	212	−0.53	−0.36	0.24	1.00

CFU—colony-forming unit; DHA—dehydrogenase activity expressed as the amount of released μg of triphenylformazan (TPF) per hour per gram of soil; T-GRSP—glomalin-related soil protein content per gram of air-dried soil.

).

3.5. Buckwheat Root Colonisation by Fungi

We assessed the degree of colonisation of buckwheat roots by potentially pathogenic fungi.

Table 6. Percentage of filamentous fungi isolated from buckwheat roots using the culture method.

	2021		2022
	AK	AP	AK	AP
Total number of colonies (sum)	271	248	327	306
Colletotrichum sp.	0.4 (1 ± 0.22) c	–	–	–
Cylindrocarpon sp.	1.1 (3 ± 0.16) c	–	–	–
Exserohilum pedicellatum	–	1.6 (4 ± 0.21) c	5.9 (18 ± 0.11) b	9.5 (29 ± 0.12) b
Fusarium avenaceum	–	–	6.4 (20 ± 0.14) b	8.8 (27 ± 0.31) b
Fusarium culmorum	10.0 (27 ± 0.11) b	6.5 (169 ± 0.22) b	4.9 (15 ± 0.09) bc	3.9 (12 ± 0.16) c
Fusarium oxysporum	66.0 (179 ± 0.16) a	54.4 (135 ± 0.22) a	65.7 (201 ± 0.07) a	66.7 (204 ± 0.18) a
Fusarium solani	–	–	0.7 (2 ± 0.19) c	–
Fusarium sporotrichoides	–	0.4 (1 ± 0.22) c	2.3 (7 ± 0.22) c	–
Mucor mucedo	1.1 (3 ± 0.16) c	–	0.7 (2 ± 0.19) c	–
Penicillium notatum	–	12.9 (32 ± 0.10) b	–	–
Penicillium purpurogeum	0.7 (2 ± 0.19) c	0.8 (2 ± 0.19)	0.7 (2 ± 0.19) c	0.3 (1 ± 0.22) c
Penicillium vermiculatum	–	0.4 (1 ± 0.22)	–	–
Penicillium sp.	6.3 (17 ± 0.14) b	–	7.2 (22 ± 0.16) b	0.3 (1 ± 0.22) c
Phoma sp.	–	–	0.3 (1 ± 0.22) c	–
Pythium sp.	–	–	–	0.3 (1 ± 0.22) c
Rhizoctonia solani	–	–	0.3 (1 ± 0.22) c	0.7 (2 ± 0.19) c
Rhizopus nigricans	0.4 (1 ± 0.22) c	–	2.0 (6 ± 0.09) c	2.6 (8 ± 0.16) c
Trichoderma hamatum	1.8 (5 ± 0.22) c	7.3 (18 ± 0.18) bc	–	1.6 (5 ± 0.16) c
Trichoderma harzianum	7.4 (20 ± 0.14) bc	11.7 (29 ± 0.37) b	0.7 (2 ± 0.19) c	3.6 (11 ± 0.22) c
Trichoderma viride	0.4 (1 ± 0.22) c	0.4 (1 ± 0.22) c	2.3 (7 ± 0.17) c	1.6 (5 ± 0.16) c
Other fungi	4.4 (12 ± 0.16) bc	3.6 (9 ± 0.14) bc	–	–

AP—intercropping; AK—control; T1 and T2—the first and second buckwheat roots sampling date; symbol „–” indicates that no fungi were isolated from the sample. Values followed by different letters in columns indicate significant differences according to two-way ANOVA with post-hoc Tukey HSD test; ±SE, standard error; total number of fungi is given in brackets.

shows the percentage of fungi on buckwheat roots at flowering (T2). Buckwheat roots in both the first and second years of the study were most abundantly colonised by fungi of the genus Fusarium. The most abundant species isolated was F. oxysporum, accounting for 54–67% of the total number of fungi. We also isolated the potentially pathogenic species F. culmorum and F. avenaceum (between 4 and 10%). In both the first and second year of the study, buckwheat roots were colonised by saprotrophs of the genus Trichoderma, ranging from 3 to 20%. However, their number decreased in the second year of the study to 8% in the intercropping sample and to 3% in the control crop. In the first year of the study, Penicillium notatum was recorded in the intercropping sample (13%). An increased percentage of Exserohilum pedicellatum (2–10%) was also found on buckwheat roots in the 2022 crop year. Other potentially pathogenic fungi, such as F. solani, Rhizoctonia solani, Phoma spp., and Pythium spp., colonised buckwheat roots sporadically (<1%) in both AK and AP crops.

3.6. Biometric Traits and Yield of Buckwheat

The cultivation variant— intercropping of buckwheat with Paulownia and no intercropping—did not significantly affect any of the studied buckwheat traits, including plant height, number of branches on the main shoot, number of branches, inflorescences, number of kernels, and total kernel weight per plant. Despite the lack of significant statistical differences, many biometric traits were higher when buckwheat was grown with Paulownia, particularly the number of full kernels per plant and the total kernel weight per plant, compared to the control site. Over the years of the study, time significantly affected kernel yield and some traits studied, such as the number of branches per plant, number of kernels, and total kernel weight per plant (

Table 7. Select morphological and flowering biology-related traits of buckwheat and buckwheat yield (averages for specific combinations and years).

Specification	Plant Height (cm)	Number of Branches on Main Shoot	Number of Branches	Number of Inflorescences	Number of Full Seeds	Total Seed Mass per Plant (g)	Buckwheat Yield (t ha−1)
Cultivation
AP	46.7 ± 2.2	1.22 ± 0.2	2.99 ± 0.3	5.95 ± 0.5	22.5 ± 2.7	0.26 ± 0.03	0.65 ± 0.1
AK	47.2 ± 1.3	1.57 ± 0.1	3.59 ± 0.4	6.44 ± 0.5	21.8 ± 4.5	0.20 ± 0.05	0.51 ± 0.1
LSD	ns	ns	ns	ns	ns	ns	ns
Years
2021	47.3 ± 1.3	1.4 ± 0.1	2.4 ± 0.4	6.8 ± 0.5	15.5 ± 3.6	0.16 ± 0.04	0.39 ± 0.10
2022	47.1 ± 1.5	1.4 ± 0.1	4.2 ± 0.2	5.5 ± 0.3	28.8 ± 3.1	0.31 ± 0.04	0.77 ± 0.1
LSD	ns	ns	0.69	ns	9.1	0.100	0.25

AP—intercropping; AK—control; ns—statistically insignificant; LSD—least significant difference; ±SE, standard error.

). In 2022, we registered a nearly 100% higher yield of kernels per hectare.

The average height of the trees at the beginning of the growing season in 2021 was 145 cm, while at the end of the 2022 season, it was 311 cm. Tree crown sizes in October 2021 averaged 108 cm in height and 157 cm in width, and ny 2022, these values had increased to 156 cm and 209 cm, respectively. The trunk girths at breast height (130 cm from ground level) at the end of the 2021 and 2022 seasons were 16.4 and 17.9 cm, respectively.

4. Discussion

4.1. Effect of Soil Properties on Microbiological Changes

In the present multidisciplinary study, we determined the effects of TBI of Paulownia trees on soil physicochemical and microbial properties, as well as on the biometric traits of buckwheat. To the best of our knowledge, our study is the first to consider the microbiological aspects of soils in the intercropping of buckwheat and Paulownia under European conditions. In northern China, Paulownia elongata trees are intercropped with wheat or beans.

Soil pH is considered the main soil variable. It interacts with microorganisms, thus determining plant growth and biomass yield [28]. In the present study, we observed that both buckwheat monoculture and intercropping resulted in a lower soil pH. Ahmad et al. (2013) and Makinde et al. (2006) [29,30] demonstrated that intercropping systems lower soil pH. Hinsinger et al. (2006) [31] and Hagen-Thorn et al. (2004) [32] showed that tree roots acidify the soil through the release of acidic compounds and microbial respiration. Furthermore, a major mechanism ensuring the efficiency of phosphorus uptake by buckwheat may be the plant’s ability to acidify the rhizosphere [33].

Soil microbial diversity and activity are significant aspects of soil quality affected by TBI. Thus far, the relationship between soil microbial properties and the diversification of intercropping has remained understudied. Still, we lack comprehensive data involving both high-throughput metagenomic sequencing techniques of bacterial and fungal populations and traditional techniques to measure microbial abundance and enzymatic activity, as well as plant biometry in temperate zone agroforestry systems.

Land use patterns have a significant impact on the composition and diversity of soil microorganisms, which are extremely sensitive to changes in the soil environment [19,34]. Our sequencing of 16S rRNA gene fragments and ITS regions has provided unique data on bacterial and fungal population diversity and structure. Intercropping increased the diversity of rhizosphere bacterial populations, while the opposite trend was observed for fungal populations, as evidenced by the values of the Shannon and Simpson indices (Table S1). Increased bacterial diversity may be due to increased carbon and nutrient supply from litter, dead root cells, and tree root secretions [35]. Tree rows have a strong influence on soil microbial communities and provide a habitat for a microbiome that differs in composition from the microbiome of neighbouring crops. Consequently, the introduction of the soil microbiome associated with tree rows into arable land through agroforestry increases the overall diversity of the system. Our findings related to the diversity and number of OTUs of the mycobiome align with previous studies that have shown that fungal communities under trees gradually diversified. Young tree cultivation within an agroforestry system does not affect the rhizosphere fungal community, and no increase in fungal populations was detected in young agroforestry systems. Clivot et al. (2020) [36] and Beule and Karlovsky (2021) [37] only detected strong promotion of soil fungi after 10 years of poplar cultivation. This may be due to adaptation to the heterogeneous understorey space of tree biomass and understorey vegetation or stochastic phenomena as a result of limited exchange between fungal populations. We found that the microbiome of the rhizosphere soil from intercropping and buckwheat monoculture was dominated by bacteria classified as Actinobabateria, Proteobacteria, and Acidobacteria and the mycobiome by Ascomycota, Basidiomycota, and Mortiellomycota. In our previous study [38], we showed that the rhizosphere of Paulownia trees was dominated by the above-mentioned types of bacteria and fungi. Wang et al. (2022) [39] found that in the rhizosphere of buckwheat, the dominant phylum were Actinobacteria, Proteobacteria, and Acidobacteria, with fungi classified as Ascomycota, Basidiomycota and Mortierellomycota. The dominance of Actinobacteria and Proteobacteria is probably related to the nutrient-rich conditions of the rhizosphere [39,40]. However, Ascomycota and Basidiomycota play an important role in maintaining soil stability, plant biomass decomposition, and plant interactions [41]. The high relative abundance of Mortierellomycota can be evidence of good soil health [42]. We found an increased relative abundance of Candidatus solibacter, Nocardioides, Pseudarthrobacter, and Sphingomonas. These microorganisms are considered plant growth-promoting rhizobacteria (PGPR) [43,44,45,46,47]. The accumulation of these microorganisms in rhizosphere soil may positively affect buckwheat biometrics. Peng et al. (2022) [46] also showed that intercropping promotes the enrichment of PGPR. In addition, in our metagenomic study, we observed a decrease in the relative abundance of Fusarium, a common soilborne plant pathogen, in intercropping [47]. It is likely that intercropping effectively reduces the disease incidence in crop fields. Mechanisms supporting this effect include modification of the crop microclimate, secretion of allopathic compounds, positive effects on antagonistic microbial communities, and diversification of soil microbial communities [48].

In our study, growing buckwheat with Paulownia had a positive effect on the increase in the number of bacterial CFUs. It is likely that the higher abundance of bacteria and fungi is the result of direct contact between plant roots in the intercropping system, which stimulates plant roots to release more nutrients. Similar results have been obtained by Beule et al. 2020 [49], who observed an increase in bacterial abundance in poplar-based agroforestry systems compared to neighbouring monocultures in arable fields. Furthermore, Lee and Jose (2003) [50] indicated that the age of the agroforestry system influences the increase in microbial biomass. Li et al. (2013) [51] showed that the number of soil fungi, bacteria, and actinomycetes in the intercropping of two species, i.e., soybean and sugarcane, increased by 115.5, 43.6 and 57.3%, respectively, compared with monoculture. Our study showed that the CFUs of bacteria and fungi were dependent on the sampling date, being generally higher at the flowering stage in buckwheat (T2), which is consistent with Wang et al. (2019) [52]. Increasing soil microbial abundance is extremely important, as it can influence plant health and soil quality, thus ensuring the stability and productivity of natural ecosystems [52].

4.2. Effect of Intercropping System on DHA and T-GRSP Activity

Soil enzyme activity, including dehydrogenases, is an important indicator of organic matter decomposition and nutrient dynamics. [9]. In our study, intercropping increased the DHA activity, which is particularly evident at T1 in spring, when the optimum temperature under conditions of sufficient moisture may be a factor favouring higher enzymatic activity. In our previous study [9]^, we reported that young plantations of Paulownia trees had a positive effect on some soil microbial parameters, i.e., DHA activity. Similarly, Wan and Chen (2004) [53] observed higher enzymatic activity in tree-based intercropping, including Paulownia spp., probably due to increased carbon content, nutrients, leaf residues, and root secretions in the soil.

Many authors have pointed to GRSP as a good indicator of soil stability and fungal activity [35,42,54,55]. Our study showed that higher amounts of GRSP were recorded with intercropping at T2 compared to the control. This may be due to the plant roots secreting more C and energetic substances that promote AMF growth and reproduction, which may increase mycelial density and length and thus, the GRSP content [56]. Furthermore, the results of our study confirm previous findings that GRSP shows sensitivity to seasonal variation and land use change [54]. Our results are also consistent with those of Zhao et al. (2020) [57] on the impact of maize and soybean intercropping on GRSP. AMF can expand the uptake area of plant roots to improve water and mineral absorption, thereby promoting plant growth [57].

4.3. Effect of Intercropping Cultivation on the Yield of Plants

Intercropping affects crop yields depending on a number of factors, such as plant growth conditions, species, soil, and climatic conditions. The yield and biometric traits of intercropped plants depend largely on the species composition and the coexistence mechanisms developed by the plants. Dang et al. (2020) [58] found that intercropping led to significant improvements in biometric and compositional traits of millet grain yield, number of ears per plant and their length, grain mass per plant, and 1000-grain weight, with increases in grain yield of 5.6–20.7% in 2017, 7.9–53.9% in 2018, and 28.3–75.4% in 2019. Yin and He (1997) [59] found that intercropping 9-year-old Paulownia trees with wheat led to a 23% reduction in wheat yield in China. Similar results were obtained by Chirko et al. (1996) [60], who demonstrated that shading provided by 11-year-old Paulownia trees in a Paulownia–wheat intercropping system reduced yields by only 7%. However, the system had no effect on reducing the number of grains per square metre or dry matter per 1000 grains. Li et al. (2008) [61] showed that wheat yield was reduced by up to 50% in the Paulownia–wheat intercropping system, and a similar reduction in wheat yield was also found in the walnut–wheat system with 8-m spacing between tree rows [61]. Li et al. (2013) [51] found that sugarcane yields in intercropping were significantly higher than in monoculture.

Different results have been reported by Zhu et al. (2002) [33] and Shukla et al. (2019) [62]. Zhu et al. (2002) [33] showed that 2-year-old mulberry trees did not have a significant effect on millet yield, but mulberry tree leaf production increased by 30%. Shukla et al. (2019) [62] found that the yields of all crops tested in the agroforestry system were lower in shaded sites compared with sites with full light exposure. We found no statistically significant effect of intercropping on the biometric traits of buckwheat or its yield. Buckwheat yield was influenced by conditions that occurred in specific years of our research, with significantly better conditions for the yield and development of this plant occurring in the 2022 season. The close dependence of buckwheat yield on weather conditions has been confirmed in other studies [62].

5. Conclusions

Our research indicates that the Paulownia–buckwheat intercropping system positively affects soil microbial properties, offering a sensitive indicator of soil quality. However, further studies are needed to assess the long-term effects of intercropping and its impact on biodiversity over a wider range. Our findings, utilising both high-throughput sequencing and traditional methods for measuring microbial abundance and enzymatic activity, suggest that the TBI system in temperate climates enhances soil bacterial abundance, diversity, and function compared to monoculture. Additionally, bacteria were more responsive to cultivation methods than fungi, indicating that the mycobiome diversifies more gradually. Both the mycobiome and TBI microbiome promote microbes that benefit plant growth and yields. Temporal fluctuations are key factors influencing soil biological properties. We also found that buckwheat growth between Paulownia rows did not negatively affect crop yield. Our study provides a foundation for the development of innovative management strategies for tree cultivation for biomass energy production.