Chemical Properties and Bacterial Community Reaction to Acidiﬁed Cattle Slurry Fertilization in Soil from Maize Cultivation

: Acidiﬁed slurry is a novel organic fertilizer that limits gaseous ammonia emissions and reduces nitrogen losses. Our research aimed to determine the effects of short-term fertilization with acidiﬁed slurry on the chemical properties and bacterial community of soil used for maize cultivation. In the months after spreading, raw slurry fertilization had a signiﬁcant impact on the increase in values of N-NO 3 . In contrast, soil fertilized with acidiﬁed slurry had lower N-NO 3 values when compared to raw slurry fertilization treatments. Bacterial sequencing using Illumina MiSeq showed no differences in the genetic diversity of bacterial communities. In all tested soil samples, dominants at the phylum level were Actinobacteria, Proteobacteria, and Acidobacteria, while dominants at the class level were Actinobacteria, Alphaproteobacteria, Thermoleophilia, Gammaproteobacteria, and Acidimicrobiia. The values of biodiversity indices (Shannon index, Simpson index) in tested samples were similar. Our results suggest that short-term fertilization with acidiﬁed slurry does not adversely affect the biodiversity and structure of the bacterial communities and has a slight impact on soil chemical properties.


Introduction
The ongoing process of agricultural intensification has an increasing impact on changes taking place in the natural environment [1]. One solution to this problem is the introduction of new fertilizer management systems which, in addition to minimizing harmful effects on the environment, also increase agricultural efficiency [2,3].
Organic fertilizers are responsible for nearly 2% of the world's greenhouse gas emissions [4]. Ammonia (NH 3 ) emission is one of the more important problems associated with organic fertilizer management. Air pollution by ammonia causes a possible risk to human and animal health [5]. NH 3 evaporation from liquid slurry represents more than 70% of the total emissions of this gas from agricultural sources [6]. Liquid animal slurry used as a fertilizer, in the first 24 h after application, loses up to 50% of its NH 3 content in the form of gaseous emissions, which also contributes to a decrease in fertilization value. Volatilization of ammonia derived from fertilizers depends on the specific soil, and atmospheric conditions, such as high soil alkalinity, high moisture, and low soil buffering capacity [7]. The use of slurry acidification techniques (SAT) is one of the proposed solutions to reduce nitrogen losses and ammonia emissions [8][9][10]. The main aim of this procedure is to lower the slurry's pH to a value below 6.4 so that the NH 3 -NH 4 equilibrium shifts towards the ionized form. SAT turn harmful and gaseous ammonia into aqueous ammonium that remains in the soil, while also contributing to the reduction in methane (CH 4 ) and nitrous oxide (N 2 O) emissions [11][12][13].
Maize (Zea mays L.), being one of the most cultivated cereals around the world, makes up an acreage of up to 196 million hectares, and in Europe approximately 11% of all crops are maize fields [14,15]. It is not only a food crop but also an industrial crop, with only 12-13% of its production used for direct human consumption. In Poland, maize acreage has increased since 2007 by over 146% and now makes up 8% of the total crop acreage in the European Union [16]. Maize, being an important and widely cultivated crop, is also one of the biggest recipients of organic and mineral fertilizers. Therefore, it is important to balance its sustainability and economic factors, while limiting negative environmental impact.
Microorganisms living in soil are abundant and highly diverse, and they are also sensitive to changes in environmental conditions [17]. Soil microorganisms take part in the soil's metabolism, thus playing a key role in maintaining its fertility and determining the long-term sustainable productivity of agricultural lands [18,19]. Because of their crucial importance in maintaining soil functions, there has been considerable effort invested in understanding the bacterial community response to disturbance or environmental change [20,21]. Studies across several types of soil demonstrated changes in microbial abundance depending on changes in the soil environment [22]. Proteobacteria, Firmicutes, Actinobacteria, and Bacteroidetes are often found to increase their overall share in bacterial communities due to N fertilization, while the relative abundance of Acidobacteria and Chloroflexi decrease [23,24].
So far, little attention has been given to the impact of acidified slurry on chemical properties, bacterial abundances, and biodiversity in soil [25]. Moreover, there is no research on bacterial community reactions to acidified slurry fertilization using highthroughput sequencing.
For these reasons, our study aimed to evaluate the possible consequences of acidified slurry fertilization on maize cultivation. We attempted to answer the following questions: does the use of acidified cattle slurry decrease the soil's pH, and will any potential changes in the soil's pH have a consequence for the soil bacterial community?

Study Site and Weather Conditions
The field experiment that involved maize cultivation (Zea mays) was carried out between the 2017 and 2018 cropping seasons in Falenty, Mazovia, Poland (52 • 08 37.4 N 20 • 55 42.1 E). According to World Reference Base [26] soil classification, the soil belonged to Chernozems, had a sandy loam texture, and the following chemical properties before fertilization (measured in 0-20 cm soil layers, March 2017): pH 6.4, N-NO 3 7.74 mg kg −1 , N-NH 4 4.0 mg kg −1 .
The experiment was set up as a completely randomized plot design with four replicates. Each plot had an area of 15 m 2 with a buffer zone of 1.5 m. Maize was harvested in September in both experimental years (2017 and 2018).
Meteorological data were gathered from the Warsaw-Okęcie weather station (EPWA 12375) located approximately 4 km away from the field experiment. The average annual air temperature in the experimental years was 10.9 • C and the annual rainfall was 433.4 mm ( Figure 1).

Cattle Slurry Properties and Application
Modified cattle slurry was obtained via slurry acidification technique according to Pedersen et al. [27]. The chemical properties of raw and acidified slurry are presented in Table 1.
In April 2017 and 2018, plots were fertilized with slurry and acidified slurry. The following fertilization treatments were applied: control (no fertilization); 84 N kg·ha −1 raw slurry (RS84); 84 N kg·ha −1 acidified slurry (AS84); 105 N kg·ha −1 raw slurry (RS105); and 105 N kg·ha −1 acidified slurry (AS105). The slurries were surface-spread using 10 L plastic Agronomy 2021, 11, 601 3 of 10 watering cans in lines to imitate band spreading with a trailing hose. The fertilization rate of cattle slurries was calculated according to the slurry properties. Slurry pH values were measured directly in 50 mL of undiluted slurry that was previously stirred until homogenous. Slurry dry matter content was evaluated using a gravimetric method [28], total organic content (TOC) following ISO 10694 guidelines [29], total nitrogen (TN) using the Kjeldahl method [28], and NH 4+ content using molecular absorption spectrophotometry [30].

Cattle Slurry Properties and Application
Modified cattle slurry was obtained via slurry acidification technique according to Pedersen et al. [27]. The chemical properties of raw and acidified slurry are presented in Table 1. Table 1. Chemical properties of raw slurry and acidified slurry. In April 2017 and 2018, plots were fertilized with slurry and acidified slurry. The following fertilization treatments were applied: control (no fertilization); 84 N kg⋅ha -1 raw slurry (RS84); 84 N kg⋅ha -1 acidified slurry (AS84); 105 N kg⋅ha -1 raw slurry (RS105); and 105 N kg⋅ha -1 acidified slurry (AS105). The slurries were surface-spread using 10 L plastic watering cans in lines to imitate band spreading with a trailing hose. The fertilization rate of cattle slurries was calculated according to the slurry properties. Slurry pH values were measured directly in 50 mL of undiluted slurry that was previously stirred until homogenous. Slurry dry matter content was evaluated using a gravimetric method [28], total organic content (TOC) following ISO 10694 guidelines [29], total nitrogen (TN) using the Kjeldahl method [28], and NH 4+ content using molecular absorption spectrophotometry [30].

Soil Sampling and Chemical Analysis
Soil samples for chemical analysis were taken from a depth of 0-20 cm in May (I) and August (II) in 2017 and in 2018. Samples for bacterial sequencing were collected in August 2018. Samples for sequencing were collected with a soil sampler probe from three random points from each treatment plot and mixed to form a composite sample. Soil samples for chemical analysis were taken from three random points from each of the four replicated treatment plots and then mixed to form a composite sample. Samples for chemical

Soil Sampling and Chemical Analysis
Soil samples for chemical analysis were taken from a depth of 0-20 cm in May (I) and August (II) in 2017 and in 2018. Samples for bacterial sequencing were collected in August 2018. Samples for sequencing were collected with a soil sampler probe from three random points from each treatment plot and mixed to form a composite sample. Soil samples for chemical analysis were taken from three random points from each of the four replicated treatment plots and then mixed to form a composite sample. Samples for chemical analysis were air-dried, ground, and passed through a 2 mm sieve. Samples for DNA extraction were stored at −80 • C.

Sequencing
Soil DNA isolation was carried out using a modified method based on the Genomic Mini AX Bacteria+ kit (A&A Biotechnology). Additional mechanical lysis of the samples was done using a FastPrep-24 device with zirconium balls. After isolation, DNA was further purified using an Anti-Inhibitor kit (A&A Biotechnology).
Sequencing of PCR amplicons of 16S rRNA genes was carried out on the V3-V4 hypervariable regions. Specific 314F (5 -CCTACGGGNGGCWGCAG-3 ) and 785R (5 -GACTACHVGGGTATCTAATCC-3 ) primer sequences were used to amplify the selected region and prepare the library. The PCR reaction was carried out using NEBNext ® High-Fidelity 2X PCR Master Mix. Sequencing took place using an Illumina MiSeq sequencer in paired-end (PE) technology 2 × 250 nt using Illumina v2 kit.
Sequencing data were deposited in NCBI Sequence Read Archive under BioProject accession PRJNA659882. The 16S rRNA gene sequence reads were processed with mothur [31]. Paired-end reads were merged and assembled using the make.contigs command. Pairs shorter than 400 bp, longer than 470 bp, or with an average Phred score quality below 25 were discarded. Chimeras were removed using the vsearch algorithm. The final reads were clustered into operational taxonomic units (OTUs) using the dist.seqs and cluster.split commands (opticlust algorithm) with a 0.03 distance cut off. A taxonomic identity was attributed to each OTU via the SILVA 132 rRNA database [32] using an 80% homogeneity cut off. Reads were normalized by subsampling to the lowest number of obtained sequences in a sample. Rarefaction curves along with data used for figure generation is available in the supplementary materials ( Figure S1, Tables S1 and S2).

Statistics and Data Visualization
Statistical analysis and visualization of sequencing data were done using R 3.5.3 statistical programming language [33] and the phyloseq package [34]. One-way ANOVA was used for the analysis of soil chemical parameters among treatments. Variance homogeneity was examined using the Levene test. The Wilk-Shapiro normality statistic was calculated to determine if residual values conformed to a normal distribution. Differences between treatments were tested using the Tukey-Kramer (HSD) test at α = 0.05 [35].

Chemical Properties
The values of pH in the tested soils, regardless of treatment and year, did not differ statistically ( Table 2). After slurry fertilization, values persisted at a similar level in all experimental variants.
Similar dependencies in the soil after the use of acidified slurry were noted by Edesi et al. [25], and the authors noted no significant changes in pH values regardless of the date, year, and variant of the experiment.
In the first year of research, N-NO 3 content in tested soil was significantly higher in all fertilized treatments in May and much lower in August. In 2018, similar results were observed. Significantly less N-NO 3 was recorded in AS84 and AS105 treatments, compared to the control plot and acidified treatments. The results noted are similar to those previously obtained by Edesi et al. [25] for pig slurry. However, in our study, favourable weather conditions in both years allowed a fast-paced conversion of N-NH 4 to N-NO 3 in the weeks after spreading. The lower values of N-NO 3 in acidified slurry treatments can be caused by a decrease in the intensity of nitrification processes after fertilization with acidified slurry [8,36].
Regarding N-NH 4 content in the soil, there were no major differences in soils from different treatments in the first year of research. The values of this parameter in the first year of the study were very similar and ranged from 3.95 to 4.98 mg kg −1 . In 2018, values of N-NH 4 were significantly higher compared to 2017, and higher in soils fertilized with raw and acidified slurry when compared to the control plot. The lack of differences in the first year could have been caused by the rapid rate of the N-NH 4 to N-NO 3 conversion process carried out by soil microorganisms, and immediate nitrate uptake [23].   Treatment  I  II  I  II  I  II  I  II  I  II  I

Bacterial Community Composition
A total of 960,393 16S bacterial raw reads were obtained from five samples. After quality filtering a total of 733,630 reads were obtained (123590 unique), with an average of 146,726 reads per sample. A total of 36,595 unique OTUs were formed after binning with a 97% similarity rate.
Bacterial biodiversity indexes in tested soil samples were similar, as both the Shannon and Simpson values did not differ (Table 3). The Shannon index ranged from 7.18 (AS105) to 7.43 (RS84), and the Simpson index from 0.002321 (0) to 0.004599 (AS84). Similar regularities were noted by Das et al. [37], who, after applying composted cattle manure and swine manure in submerged rice paddy, did not find significant differences in bacterial diversity indices compared to the control. In addition, Hartmann et al. [38], studying soil from crop rotations (winter wheat and grass clover), did not observe significant differences in bacterial richness and evenness in conventional manure fertilization in comparison with control plots.
One of the most important aspects of next generation sequencing (NGS) is the ability to identify key taxa carrying out important biochemical reactions in the soil. The analysis of the taxonomic composition of the microbiota derived from the tested soil showed that the dominants at the phylum level included Actinobacteria (37.32-41.80%), Proteobacteria (24.50-25.61%), and Acidobacteria (9.86-11.61%) (Figure 2).
(24.50-25.61%), and Acidobacteria (9.86-11.61%) (Figure 2). Dominants at the class level, among others, included Actinobacteria (13.97-21.08%), Alphaproteobacteria (11.36-12.54%), Thermoleophilia (9.79-12.80%), Gammaproteobacteria (8.41-9.62%), and Acidimicrobiia (6.40-7.83%). In addition, some OTUs were assigned to groups of microorganisms that remained unclassified. The most numerous such group in the samples tested was Subgroup_6_or (6.10-7.22%), which included non-culturable bacteria belonging to the phylum Acidobacteria (Figure 3).   The results obtained on the percentage share of dominates at the phylum level are often reproduced in other papers. Based on terminal restriction fragment length polymorphism (T-RFLP), Hartmann and Widmer [39] stated that the dominants in agricultural soils include bacteria from the phyla Actinobacteria (35-39%), Proteobacteria (26-33%), and Acidobacteria (11-13%), and less commonly, Firmicutes (3-6%), Bacteroidetes (2-3%), and Chloroflexi (1-2%). In later years, many studies created thanks to next-generation sequencing confirmed this thesis [40]. Most of these studies examined the effects of longterm mineral and organic fertilization on soil microbiota [41][42][43][44]. Fertilization with manure, liquid manure, or slurry may stimulate the development of copiotrophic bacteria, such as those belonging to the phyla Actinobacteria, Proteobacteria, or Firmicutes. These bacteria multiply rapidly when large amounts of organic matter are introduced. However, larger amounts of carbon and nitrogen do not affect the development of oligotrophic bacteria, which include mainly Acidobacteria. This group of bacteria has a more economical The results obtained on the percentage share of dominates at the phylum level are often reproduced in other papers. Based on terminal restriction fragment length polymorphism (T-RFLP), Hartmann and Widmer [39] stated that the dominants in agricultural soils include bacteria from the phyla Actinobacteria (35-39%), Proteobacteria (26-33%), and Acidobacteria (11-13%), and less commonly, Firmicutes (3-6%), Bacteroidetes (2-3%), and Chloroflexi (1-2%). In later years, many studies created thanks to next-generation sequencing confirmed this thesis [40]. Most of these studies examined the effects of longterm mineral and organic fertilization on soil microbiota [41][42][43][44]. Fertilization with manure, liquid manure, or slurry may stimulate the development of copiotrophic bacteria, such as those belonging to the phyla Actinobacteria, Proteobacteria, or Firmicutes. These bacteria multiply rapidly when large amounts of organic matter are introduced. However, larger amounts of carbon and nitrogen do not affect the development of oligotrophic bacteria, which include mainly Acidobacteria. This group of bacteria has a more economical life strategy than copiotrophic bacteria and occurs in soils poor in nutrients [45][46][47]. Therefore, the introduction of large amounts of organic matter into the soil can interfere with the natural O:K ratio and thus negatively affect biological soil homeostasis [48].
No changes in the structure of bacterial communities were noted due to the use of acidified and non-acidified slurry. Our results are supported by the study from Edesi et al. [25]. Despite using different methods, these authors noted no significant impact of pig slurry on the bacterial community using phospholipid fatty acid (PLFA) extraction [49]. Because of the short-term nature of the fertilization experiment, the bacterial community quickly returned to its initial state, and permanent changes in the bacterial community are usually observed only after long-term fertilization [50]. However, some publications suggest a significant effect of short-term fertilization on the soil microbiome [51,52]. The discrepancy between the possible effects of short-term fertilization on soil microbiota can be attributed to the wide range of applied agricultural treatments and practices [53,54].
No differences were noted between the applied fertilization doses. Several papers present similar results. Stark et al. [55], after applying fertilization for 91 days, found no change in the bacterial community compared to control treatment. Similar regularities were also found in the case of research conducted by Roberts et al. [56], who used nitrogen fertilization at 20 or 130 kg ha −1 for 3 years.

Conclusions
Fertilization with acidified slurry did not cause a major shift in the chemical properties of the tested soil. However, it lowered the content of NO 3 -N one month after spreading with acidified slurry in both years, in comparison to control and raw slurry treatments. Moreover, our findings based on soil sequencing show that short-term fertilization with acidified slurry did not cause an immediate shift in bacterial community structure. According to data available in other studies about the effects of long-term fertilization, it can be only hypothesized that prolonged use of acidified slurry might cause significant changes in the abundances and presence of certain bacterial taxa. Our paper is one of the first to use NGS (DNA) taxonomic profiling in the study of acidified slurry and its influence on the soil. However, it did not answer questions regarding the RNA profile, and did not cover all of the soil's chemical properties.
There is a need for further research on this issue, but at this time, the use of acidified slurry as fertilizer seems to be an environmentally safe way of increasing maize yields while limiting ammonia and other potentially harmful gaseous emissions.