The distribution of functional N-cycle related genes and nitrogen in soil profiles fertilized with mineral and organic N fertilizer

Nitrogen (N) fertilizers applied to agricultural soils result in the release of nitrogen, mainly nitrate (NO3-) in addition to nitrous oxide (N2O) and ammonia (NH3), into the environment. Nitrogen transformation in soil is a complex process and the soil microbial population can regulate the potential for N mineralization, nitrification and denitrification. Here we show that agricultural soils under standard agricultural N-management are consistently characterized by a high presence of gene copies for some of the key biological activities related to the N-cycle. This led to a strong NO3- reduction (75%) passing from the soil surface (15.38 ± 11.36 g N-NO3 kg-1 on average) to 1 m deep layer (3.92 ± 4.42 g N-NO3 kg-1 on average), and ensured low nitrate presence in the deepest layer. Under these circumstances the other soil properties play a minor role in reducing soil nitrate presence in soil. However, with excessive N fertilization, the abundance of bacterial gene copies is not sufficient to explain N leaching in soil and other factors, i.e. soil texture and rainfall, become more important in controlling these aspects.


32
Anthropogenic activities are the major driver of changes in the global nitrogen (N) cycle since the 33 last century, resulting in N-flows being 3.3-fold higher than those due to natural processes, 34 achieving a total globally fixed nitrogen of 413 Tg N y -1 . Since nitrogen is one of the most 35 important nutrients for many life forms, such a strong change in its availability has important 36 effects on the balance of terrestrial and aquatic ecosystems [1,2]. reaching in 2050 levels equal to 150% of those of 2010 [9]. 48 Nitrogen fertilizers applied to agricultural soils result in the release of N into the environment, 49 both in the atmosphere (NH 3 , N 2 O, N 2 ), and in groundwater (NO 3 -) [10]. In many areas of the 50 world, agriculture has been acknowledged as the single largest source of N (mainly NO 3 -) to 51 environments [11][12][13][14], causing alterations in the carbon cycle, biodiversity reduction, 52 acidification, soil fertility reduction and air pollution [15]. Furthermore, human health problems 53 should be also considered; for example, in recent years many studies reported an increased 54 incidence of colorectal cancer in subjects who regularly drink water with a concentration of nitrate 55 above 8.6 mg L -1 [16,17]. All these problems lead to a social cost that has been estimated to be up 56 to 10 $ kg -1 N [18]. 57 Nitrogen in soils can be metabolized and transformed by soil microbial communities [3,[19][20][21][22][23], 58 returning to the atmosphere in gaseous form (N 2 O, N 2 ). Once the reduced forms of ammonia, 59 coming either from mineral or organic fertilization, are oxidized by nitrifiers, the denitrification 4 60 process can ensue which reduces nitrate and nitrite (NO 2 -) to nitric oxide (NO), nitrous oxide 61 (N 2 O), and dinitrogen (N 2 ). 62 Today there is general agreement that soil microorganisms play a central role in the N cycle 63 [3, 20,22] as they are responsible for the conversion of N in its various forms: the structure of soil 64 microbial populations is regarded as the major variable that regulates the potential for N fixation, 65 mineralization, nitrification and denitrification [24]. 66 In the recent years some findings have enriched the complexity of the known pathways ruling the 67 nitrogen cycling in the environment. One of such cases is the comammox [ of potential nitrification are the classic ammonia oxidizing bacteria and the newly discovered 81 comammox do not play a significant role in these pathways (P < 0.05).

82
Concerning the second nosZ clade (nosZ II) we chose to restrict the survey of the terminal gene of 83 denitrification to nosZ I based on the conclusions of Domeignoz-Horta and coworkers (2015) [28] 84 who reported that: (a) the nosZ I community was consistently more abundant than the nosZ II one 85 and (b) no significant differences between the two groups could be ascribed to the different 86 agricultural management practices, neither in relation to crops nor to fertilization regimes. The 87 same authors add that the lack of detectable variations between these subgroups is in line with the 88 fact that such differences have been reported only in long-term agronomical trials that had been 89 carried out for over 50 years.

90
Finally, regarding the known existence of two families of genes able to perform nitrite reductase 91 activity converting nitrite into nitrous oxide (nirK and nirS) we selected the former due to the 92 following reasons: (a) nirK-harboring bacteria mostly dominate in soils and rhizospheres over 93 nitrite reducers of the nirS kind [29]; (b) there is a tight correlation between nirS and nosZ [30] 94 which allows to infer indirect information on the abundance of the former by analyzing the latter.

95
These considerations are also confirmed by our prior work [31] in which we analyzed both nirK 96 and nirS as well as nosZ in Bermuda grass rhizospheres.

97
Enhanced N removal through optimization of denitrification has drawn much attention as an 98 effective approach towards N control because it is the only pathway, except for the process of   Ten of them (soil codes 1a, 1b, 2, 3a, 3b, 4a, 4b, 5, 6a, 6b) were fertilized by a regular farming 134 approach using different types and quantities of nitrogen up to a maximum of 450 kg N ha -1 (Stage 135 1 of the study). The last two soils (soil codes 7 and 8) received in 2016 an excess of N fertilizers 136 (Stage 2 of the study). In particular, Soil 7 was equivalent to Soil 4a (S4 Table) but it received an 137 extra N-fertilization in October (860 kg N Ha -1 , for a total annual N of 1,243 kg N Ha -1 ) by using 138 pig slurry (S1 Table). Soil 8 (S4 Table) represented a field cropped with maize and receiving, 139 during the season an excess of N (1,470 kg N Ha -1 ) by three N-fertilization events during the year, 140 using digestate in April, urea at the end of July and pig slurry in October (S1 Table).

141
Soils with the same code number but different letters were carried out at the same site (farm), but 142 in different fields and with different N fertilizers. The agronomic management of each site is 143 reported in S1 Table. Table. 8 150 For each soil sampling a composite sample was taken, formed by mixing 10 sub-samples taken 151 inside the plot. The collection points within each experimental plot were identified according to 152 an X distribution, taking care to avoid the borders of the plots.

153
Soils taken for chemical analyses were put into sealed containers and stored at 4°C; analyses were 154 performed starting the next day. Soil samples for DNA extraction and qPCR were processed in the 155 hours immediately following the sampling.  Table. Each sample was tested in triplicate, and the standard calibration 164 curve was built using five points in triplicate, equal to fifteen reactions. As templates for the 165 standard curves, amplicons for each of the target genes were cloned into purified plasmids (pGem-166 T; Promega Corp.) and inserted into E. coli JM101 by electroporation. Knowing the size of the 167 vector (3,015 bp) and those for each insert (data from literature, S3 Table), and measuring the    Ten agricultural soils (S1 Table, Table) 213 indicate that despite the soils differing for characteristics and management, with particular 214 reference to N dosed (range of 153-453 kg Ha -1 y -1 ) and N-fertilizers used (urea, animal slurries 215 and digestates from animal slurries) (S1 Table), the NO 3concentrations (Fig 1)  The abundance of gene copies related to the N-cycle 234 Nitrogen transformation in soil is a complex process and depends on many factors, but there is 235 agreement on the fact that that soil microorganisms play a central role [20,22] (Fig 2 and S6 Table).

241
Results show a strong spatial coincidence between the number of gene copies detected coding for 242 the different N transformations and mineral nitrogen content (r coefficients > 0.91; p<0.05; n=252) 243 (S7 Table). These results agree with recent indications highlighting the tendency of soil 244 microorganisms to form complex communities within which nitrogen is metabolized, processed 245 and transformed [22]. 246 Gene copies found per gram of soil decreased with depth for all soils studied (except for the gene 247 nifH, related to N fixation) (Fig 3, p<0.01; n=252). 248 Consequently, we found that the number of genes copies for the N cycle related genes were much 249 higher in the surface layer than in the other layers (Fig 3), in correspondence with the highest 250 concentrations of NO 3and NH 4 + (Fig 1). The difference observed was stronger in the case of 251 bacterial amoA genes, i.e. a number of gene copies (gene copies g -1 soil) for the layer 75-100 cm, 252 were 16 times lower than those detected in 0-25 cm layer.   (Fig 4). These soils all showed an alkaline pH 277 (8.45 ± 0.03; n= 5) and they are rich in clays and silt (0-50 cm: 28.73% ± 5.07% and 47.14% ± 278 5.36%, respectively; n=5) (S4 Table). These soils showed a strong reduction of the nitrate 279 concentration from the surface to 1 meter of depth (85% ± 19%; n=62). 280 In contrast, the soils which occurred in the right quadrants (group b, two-way ANOVA, p=0.0003; 281 F=3.61; DF= 18; n=59) (Fig 4) showed a NO 3concentration at a 1-meter depth (8.71 mg N-NO 3 -282 kg -1 ± 4.85; n=12) significantly greater than that of soils of group a. The pH measured for these 283 soils is neutral (6.93 ± 0.32; n=2), and sand is well represented in the first 50 cm (49.95% ± 6.99%; 284 n=2) (S4 Table). The nitrate reduction between the surface layer and the 75-100 cm deep layer 285 was less than that measured for the other soil group but still remarkable, i.e. 58% ± 20% (n=24).  (Fig 4). 289 Analysis performed (PCA in Fig 4) indicates that, in general, N dose does not affect nitrate 290 presence, unlike the soil texture. In particular clay and sand soil contents seem to affect nitrate 291 concentration along the soil profile (Fig 4), as confirmed by the Pearson correlation analysis for 292 the NO 3concentration in the 75-100 cm vs. percentage of sand and clay in the surface layer (0-50 293 cm), i.e. r=0.718, p<0.05, n=189 and r=-0.698, p<0.05, n=189, respectively. pH could also play a 294 role (Fig 4) as it is reported that alkaline pH stimulates biological activities [59], although other 295 authors reported the high adaptability of denitrifying bacteria to different pH [60]. In any case, 296 the reported pH effect contrasts with our results, which indicates that gene copy numbers are higher 297 for those soils characterized by a lower pH (right quadrant of Fig 4) than for soils having alkaline 298 pH (left quadrant of Fig 4). This controversial trend can be explained by considering that gene 299 copies presence is regulated by the amount of reactive N (r > 0.635, p<0.05; n=189), in agreement 300 with PCA results (Fig 4).

301
The application of the partial least square analysis (PLS) considering all factors included into PCA 302 analysis gives a regression (R 2 = 0.96, R 2 cv = 0.95; p<0.05; n=10; parameters = 26) (S8 Table) 303 that confirms all PCA parameters influencing nitrate presence such as before discussed. Indeed, 304 high clay and silt contents reduce nitrate concentration in the 75-100 cm soil layer. On the other 305 hand, soils characterized by light textures (sandy soil) are more exposed to nitrate leaching [34] 306 although in this case an increase in the presence of genes related to nitrifying/denitrifying activities 307 (copies of amoA-Eubacteria, amoA-Archea, nirK and nosZ genes) driven by the presence of 308 reactive N in the upper layers (S8 Table), is able to keep nitrate concentration in line with that of natural soils (9.6 mg N-NO 3kg -1 ) 35 , emphasizing the primary role of N-cycle related activities in 310 determining the control of nitrate concentration in soils profiles.

311
Our conclusions in this respect are that nitrate presence at 1-meter soil depth can be explained by 312 the abundance of gene copies for enzymes related to N-cycle, but that soil texture also plays an 313 important role. with pig slurry (S1 Table). Despite this, only a small part of the nitrate reached 75-100 cm depth 338 in soil, the NO 3concentration at the same depth in autumn being 6.85 mg N-NO 3kg -1 ± 1.91 339 (n=6). This value is much lower than that measured in the month of June (15.81 mg N-NO 3kg -1 ) 340 after normal N fertilization with urea (138 kg N Ha -1 ) and in the presence of the crop. In this case, 341 differences in nitrate concentration depended on rainfall which was double in June-July in 342 comparison with that for October-November (S9 Table). Soil 8 instead, showed NO 3 -343 concentrations at the surface which were much higher than those measured for soils fertilized at 344 the normal rate, in particular after high N-fertilization (June and August) (Fig 5c). Nitrate 345 concentration at 1-meter depth in this period was of 32.37 mg N-NO 3kg -1 ± 23.77 (n=3) and in 346 June, the NO 3content exceeded 50 mg N-NO 3kg -1 at 1-meter depth, that is five times higher than 347 values reported (on average) for the soils previously studied, including Soil 7 that was fertilized 348 with an excess of N similarly to Soil 8.

349
It is interesting to compare Soil 7 with Soil 8 in the autumn period. Indeed, in autumn Soil 7 350 received a large amount of N (860 kg N Ha -1 ) leading to the high NO 3presence in the surface 351 layer, which however did not correspond to high nitrate concentration at the 75-100 cm depth soil 352 layer (Fig 5). On the contrary, in the same period Soil 8 received much less N with similar N-353 fertilizer (580 kg N Ha -1 ) but showed high nitrate presence at 75-100 cm depth, although rainfall 354 registered was much less than that for Soil 7 (205 mm for Soil 7 and 124 mm for Soil 8). This 17 355 result appears more peculiar if we consider that the number of gene copies g -1 related to enzymes 356 implicated in nitrifying-denitrifying activities measured for Soil 8 is of 1 to 2 orders lower than 357 that measured for Soil 7 (Fig 2; S6 Table), despite an alkaline pH (S4 Table) which could stimulate 358 biological activities [59] and an organic carbon content which can support denitrifying activities 359 (S4 Table) [61].  [20]. We can therefore summarize results as follows: Soil 7 received a large amount of ammonia 367 in autumn that was transformed into nitrate by ammonia oxidation microorganisms, then nitrate 368 was concentrated above all in the surface layer, because it was not rapidly leached by rainwater.

369
In this way, the high residential time of ammonia and nitrate due to the abundant clay presence in 370 the soil allowed its denitrification, explaining the low NO 3content found at 1-meter depth [62].

371
This fact was confirmed by both the higher number of gene copies related to nitrifying/denitrifying 372 activities registered for Soil 7 with respect to Soil 8 (S6 Table), and by the very high positive 373 correlation found for genes coding for nitrification with those for denitrification (amoA Archaea 374 vs nirK: r = 0.814, p<0.05; amoA EUB vs nirK: r = 0.991, p<0.01; n=8; layer 0-25 cm). The fact 375 that the number of gene copies did not decrease along the soil profile seems to indicate that 376 nitrifying-denitrifying processes continue throughout all soil depths (S6 Table). Contrarily, Soil 8, 377 due to the high presence of sand, led to the rapid leaching of nitrate, limiting the proliferation of 18 378 microorganisms (i.e. gene copies number) related to N denitrification, that are, in effect, much 379 lower than those measured for Soil 7 (S6 Table). Results of this work suggest that with a normal N fertilization (up to 450 kg N Ha -1 ) the microbial 389 populations of the soil involved in the N cycle were able to completely metabolize the nitrogen 390 supplied with fertilization, despite soil characteristics, ensuring low nitrate content at one-meter 391 depth. However, for higher N fertilization rates (1,243 kg N Ha -1 and 1,470 kg N Ha -1 ), the activity 392 of soil microorganisms was not able to metabolize all the nitrogen. In this case the characteristics 393 of the soil, i.e. texture, and rain precipitation, also regulated the presence of nitrate in soil profiles.