- Email: [email protected]
Response of hydrolytic enzyme activities and nitrogen mineralization to fertilizer and organic matter application in subtropical paddy soils
European Journal of Soil Biology 80 (2017) 27e34
Contents lists available at ScienceDirect
European Journal of Soil Biology journal homepage: http://www.elsevier.com/locate/ejsobi
Original article
Response of hydrolytic enzyme activities and nitrogen mineralization to fertilizer and organic matter application in subtropical paddy soils M.A. Kader a, c, e, *, S. Yeasmin a, Z.M. Solaiman d, S. De Neve b, S. Sleutel b a
Department of Soil Science, Bangladesh Agricultural University, Mymensingh 2202, Bangladesh Department of Soil Management, Ghent University, Coupure Links 653, 9000 Gent, Belgium c School of Veterinary and Life Sciences, Murdoch University, South St, Murdoch, WA 6150, Australia d UWA School of Agriculture and Environment, and The UWA Institute of Agriculture, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia e School of Agriculture and Food Technology, The University of the South Pacific, Alafua Campus, Samoa b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 July 2016 Received in revised form 24 December 2016 Accepted 15 March 2017
Drivers of nitrogen (N) mineralization in paddy soils, especially under anaerobic soil conditions, are elusive. The influences of exogenous organic matter (OM) and fertilizer application on the activities of five relevant enzymes (b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase) were measured in two long-term field experiments. Of the two field experiments, the 18-year field experiment was established in a weathered terrace soil with rice-wheat crop rotation at the Bangabandhu Sheikh Mujibur Rahman Agricultural University (BSMRAU) farm with five OM treatments and two levels of mineral N fertilizer. The 30-year experiment was established in a young floodplain soil with rice-rice crop rotation at the Bangladesh Agricultural University (BAU) farm with five mineral fertilizer treatments including one with farm yard manure. At BSMRAU, N fertilizer and OM amendments significantly increased all enzyme activities, suggesting the availability of primarily substrate for microbial activity. Whereas at BAU, non-responsiveness of b-glucosidase activity, suggesting that fertilizer and OM amendments had little effect on overall soil microbial activity. Nevertheless the microbial demand for N, b-glucosaminidase and L-glutaminase activities differed among the treatments (P < 0.05) and showed opposite trends with soil N mineralization. Hence enzymatic pathways to acquire N differed with the treatment at BAU site, indicates differences in soil N quality and bio-availability. L-glutaminase activity was the sole investigated variable that positively correlated to both the aerobic and anaerobic N mineralization rates in both field experiments. Combined with a negative correlation between b-glucosaminidase activity and N mineralization rate, it appears that terminal amino acid NH2 hydrolysis was a rate-limiting step for soil N mineralization at the BAU site. Future investigations with joint quantification of polyphenol accumulation and binding of N alongside an array of extracellular enzymes, including oxidases for phenols and hydrolases for N-compounds, would enable verification of the hypothesized binding and stabilization of N with accumulating polyphenols at BAU site under SOM storing management. © 2017 Elsevier Masson SAS. All rights reserved.
Handling editor: Christoph Tebbe Keywords: Enzyme activity Fertilization Long-term experiment N mineralization Organic amendment Paddy soil
1. Introduction A better understanding of the factors controlling nitrogen (N) mineralization in paddy soils and development of practical
* Corresponding author. Department of Soil Science, Bangladesh Agricultural University, Mymensingh 2202, Bangladesh; School of Veterinary and Life Sciences, Murdoch University, South St, Murdoch, WA 6150, Australia; School of Agriculture and Food Technology, The University of the South Pacific, Alafua Campus, Samoa. E-mail address: [email protected] (M.A. Kader). http://dx.doi.org/10.1016/j.ejsobi.2017.03.004 1164-5563/© 2017 Elsevier Masson SAS. All rights reserved.
indicators of soil N supply is essential to improve N-use efficiency in South-east Asian paddy rice production, thereby reducing the application of relatively expensive N fertilizers. Several biological and chemical methods have been proposed as N mineralization indices [1,2]. However, limited progress has been made on reliable prediction of paddy soil N mineralization. In our previous work, basic soil properties as well as an array of physicochemical soil organic matter (SOM) fractions have been tested for the prediction of potential N mineralization from Bangladeshi paddy soils under laboratory incubations [3e5]. While some of these soil variables
28
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
correlated with the potential aerobic N mineralization rate, relationships with anaerobic N mineralization were mostly insignificant or negative. Hence it was inferred that neither SOM quantity, nor SOM quality dominantly determine the anaerobic N mineralization process. A multitude of other biotic or abiotic constraining factors, which are not expressed in readily measurable soil variables or SOM fractions, could control the anaerobic N mineralization in paddy soils. As organic N mineralization is mediated by microbial extracellular enzymes, assays of their activity should provide insight into key intermediate soil biochemical processes [6] and if successful, could be used as sensitive N mineralization indexes. Enzyme activities are the end result of the interaction of SOM biochemistry and physical soil conditions, both being shaped by management. The activities of urease and L-asparaginase and therefore the soil amidohydrolase activities in general have the potential to evaluate mineralizable N [7]. Tabatabai et al. [6] proposed N-acetyl-ß-Dglucosaminidase activity as an index of soil N mineralization among six amidohydrolases enzymes involved in N cycling and four glycosidase enzymes involved in carbon cycling in soils. However, such conclusions are not yet supported for flooded paddy soils by the lack of experimental data. Specific practices, i.e., wet cultivation, puddling, alternate wetting and drying make paddy soils distinct in physical, chemical and biological properties and role of enzymes to more frequently studied soil types. The present study considered five relevant ecoenzymes, covering initial and terminal steps in organic matter and N mineralization and urea hydrolysis. b-glucosaminidase and arylamidase are selected as representatives for extracellular enzymatic breakdown of complex organic N compounds into amides, amino sugars, and amino acids (aminization). This is assumed to be an initial rate-limiting step in soil N mineralization [8]. L-glutaminase was selected as representative for an array of enzymes involved in the production of NHþ 4 from amino acids through ammonification [9]. Microbial nutrient demand is determined by the elemental stoichiometry of microbial biomass in relation to environmental nutrient availability [10]. b-glucosidase activity, involved in cellobiose hydrolase and carbon (C) and energy supply, was therefore included as well, as this allowed further evaluation of the eco-enzymatic C:N-ratio as an indication of the tendency of microbial activity to be determined for nutrient or energy flow. Lastly, urease is an enzyme that catalyses the hydrolysis of urea, hydroxyurea, dihydroxyurea and semicarbazid into CO2 and NH3 [11]. The principal aim of this study was to evaluate the relative control of these selected enzymatic steps on aerobic and anaerobic N mineralization in young floodplain paddy soils. We interpret strong correlations between N mineralization rate and enzyme activity as likelihood that the mediated OM-transformation step would be limiting N mineralization. The secondary aim was to elucidate if enzyme activity was either determined by product demand or by substrate availability. It was hypothesized that: N fertilizer addition would reduce differences in enzyme activities between exogenous OM treatments due to a lifting of product demand, in casu mineral N; exogenous OM application would promote activity of all enzymes because of a generally enhanced substrate availability; a similar effect in mineral fertilizer treatments that would promote crop growth, and logically larger OM inputs from root exudation and incorporation of crop residues; and a higher demand for mineral N in treatments with exogenous OM with higher C:N ratio to result in a specific promotion of the activity of hydrolytic enzymes mediating N-transformations relative to bglucosidase.
2. Materials and methods 2.1. Site description and soil sampling Soil samples were collected from two long-term field experiments. One experiment was established in 1989 at the Bangabandhu Sheikh Mujibur Rahman Agricultural University (BSMRAU) farm at Salna (24 050 N, 90160 E), Bangladesh on a clayey, kaolinitic, Ultic Ustocrept soil [12] developed from Madhupur clay. The soil texture was silty clay loam (15:46:39) [13]. A yearly Wheat (variety Akbar)-Fallow-Transplanted rice (variety T. Aman) cropping pattern was practised. The BSMRAU field trial involved five OM application treatments (1 no-application, 2 rice straw (air dry) at 2 Mg ha1, 3 green manure at 7.5 Mg ha1 fresh biomass of Sesbania rostrata, 4 compost at 25 Mg ha1 made from cow dung and rice straw, and 5 cow dung at 25 Mg ha1 as fresh manure). These were combined with 2 levels of inorganic N fertilizer amendments (0 and 100 kg N ha1 for rice and 0 and 120 kg N ha1 for wheat). Triple super phosphate (TSP) was applied at 44 kg P ha1 for both crops and muriate of potash (KCl) was applied at 55 and 45 kg K ha1 for rice and wheat, respectively, during the final land preparation. The experiment was laid out in a 5 3 factorial design with replications. The dimension of each plot was 12 m 7 m with a plot-to-plot distance of 1.5 m. Details on the agronomic management can be found in Ref. [13]. The second field experiment was established in 1978 at Bangladesh Agricultural University (BAU) farm at Mymensingh (24 430 N, 90 250 E), Bangladesh on a loamy, mixed, non-acidic Aeric Haplaquept [12], developed from old Brahmaputra alluvium. The soil texture was silt loam (19:63:18). The treatments all had a yearly Boro rice (irrigated winter rice transplanted on midJanuary and harvested mid-May)-Fallow-T. Aman rotation and included treatments with application of mineral fertilizer (control, N, NP, NPK) and one with application of mineral N and farmyard manure (NþFYM). F The application rates of N, P, K, S, and Zn per crop were 90, 20, 19, 30, and 5 kg ha1, respectively applied as urea, triple super phosphate, potassium chloride, gypsum, and zinc oxide. Cow dung was mixed with rice straw applied once a year 10e15 days prior transplantation of Boro rice at a rate of 5 Mg ha1 fresh material. The experiment was conducted in a randomized block design with three replications (12 m 6 m). Details on the agronomic management can be found in Ref. [13]. Surface soil samples (0e15 cm) were collected from 15 locations per replicate plot by means of a 2.5 cm inner diameter auger in May 2008 at BSMRAU, and in July 2008 at BAU. These samples were bulked into one composite sample and thoroughly mixed. The field moist soil was gently broken apart by hand and air-dried and ground to pass a 2-mm sieve prior to the assessment of N mineralization, and enzyme activities. 2.2. Nitrogen mineralization Fourteen-week laboratory incubations were carried out to determine both aerobic and anaerobic N mineralization rates, as described in detail by Kader et al. [4]. Three replicate plot soils per treatment at BAU and two replicate plot soils per treatment at BSMRAU were used to quantify potential N mineralization. In total 42 tubes were filled for each BAU treatment such as 21 tubes (3 replicates 7 dates) for aerobic and 21 tubes for anaerobic incubation; and 28 tubes per BSMRAU treatment such as 14 tubes (2 replicates 7 dates) for aerobic and 14 for anaerobic incubation. Removal of mineral N from soil solution by denitrification or immobilization was not considered and only net N mineralization was measured in the incubation experiments. The net aerobic N mineralization data were best described by a zero-order kinetic
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
model: N(t) ¼ N0 þ k0t, where t is the time (in days), N(t) is the amount of mineral N at time t, N0 is the initial amount of mineral N (mg N kg1), and k0 is the zero-order N mineralization rate (mg N kg1 day1). The anaerobic N mineralization data were best described with a first-order model: N(t) ¼ NA(1-e-k1t), with NA the mineralizable N and k1 a first-order rate parameter. Average mineralization rates (in mg N kg1 day1) were then derived for the entire 14-week incubation period. 2.3. Enzyme activity analysis Five enzyme activities were measured in one-week pre-incubated soil (at 27 C, 60% water-filled pore space and at field bulk density). The b-glucosaminidase (EC 3.2.1.30), b-glucosidase (EC 3.2.1.21), L-glutaminase (EC 3.5.1.2), urease (EC 3.5.1.5) and arylamidase (EC 3.4.11.2) activities were quantified according to Parham & Deng [14], Eivazi & Tabatabai [15], Frankenberger & Tabatabai [16], Tabatabai & Bremner [11] and Acosta-Martinez & Tabatabai [8], respectively. For determination of all enzymes, controls were included, in which the substrate was added only after incubation and termination of the enzyme reaction. All results reported were averages of duplicate assays and expressed on a moisture-free soil mass basis. Moisture was determined from weight loss after drying at 105 C for 48 h. 2.4. Statistical analysis All statistical analyses were carried out in SPSS Version 19. Nonlinear regression analysis with the Levenberg-Marquardt algorithm was used to estimate the first-order model for N mineralization kinetic parameters. ANOVA and Duncan's post-hoc mean comparison tests were performed to assess significant differences in the mean enzyme activities of the OM and fertilizer application treatments. Pearson's correlation analysis was used to investigate the relationship between the tested enzyme activities and the aerobic and anaerobic N mineralization rates. 3. Results 3.1. Nitrogen mineralization The influence of OM amendment and N fertilizer application on the evolution of mineral N in soil and N mineralization rates in both field experiments was described before in detail by Kader [3]. Briefly, there was an overall significant positive effect (P<0.05) of mineral N application on both aerobic and anaerobic N mineralization rates (mg N kg1 day1) in the BSMRAU field experiment (Table 1). N mineralization rates did not significantly differ between the various exogenous OM application treatments. Both anaerobic as well as aerobic N mineralization rates correlated positively with soil N content (r¼0.46, P<0.05 and r¼0.60, P<0.01, respectively) and SOC content (r¼0.60, P<0.01 and r¼0.68, P<0.01, respectively). The aerobic N mineralization rate correlated negatively with soil C:Nratio (r¼0.61, P<0.05). In the BAU floodplain soil experiment, the aerobic N mineralization rate differed significantly (P<0.01) among the mineral fertilizer treatments (Table 1). Aerobic N mineralization rates were statistically identical in NP and control treated plots (1.97e2.24 mg N kg1 day1) and higher than in the N, NPK and NþFYM treated plots (1.22e1.54 mg N kg1 day1). Anaerobic N mineralization rate exhibited a very strong variation and did not differ significantly between the treatments (0.98e1.98 mg N kg1 day1) (Table 1). Neither aerobic nor anaerobic N mineralization rate correlated with SOC, soil N content, C:N-ratio or pHKCl.
29
3.2. Enzyme activities For the BSMRAU site, all the measured enzymes activities were significantly influenced by OM treatment (b-glucosaminidase (P<0.01), b-glucosidase (P<0.01), L-glutaminase (P<0.01), urease (P<0.01) and arylamidase (P<0.05)) and there were interaction effects of OM-application and mineral N supply (b-glucosaminidase (P<0.01), b-glucosidase (P<0.01), L-glutaminase (P<0.01) and arylamidase (P<0.05)). An exception was a nil interaction effect of OM and N fertilizer addition on urease activity (Table 2). Among the OM treatments, the b-glucosaminidase activity followed the order: cow dung > compost > green manure > no application > rice straw application (Table 2). The lowest b-glucosidase activity was observed in the control followed by cow dung and compost treatments while green manure and rice straw treated soils had significantly higher activities (Table 2). The control and rice straw treatments had significantly lower L-glutaminase activities compared with the compost, green manure and cow dung treatments (Table 2), with an exceptionally higher activity in the latter. Urease activities of the BSMRAU soil varied widely from 1.8 to 1 14.0 mg NHþ dry soil h1, with a particularly lower activity for 4 -N g the green manure treatment. Activities of arylamidase differed by an order of magnitude with a significantly lower activity in compost treated soil compared to the control and cow dung treatments and intermediate values for the rice straw and green manure treatments (Table 2). Activities of all studied enzymes, except urease, increased significantly with application of mineral N fertilizer (b-glucosaminidase (P<0.05), b-glucosidase (P<0.01), Lglutaminase (P<0.01) and arylamidase (P<0.01)). Activities of urease and arylamidase were lowered by green manure and rice straw amendment in 220 kg N ha1 treated plots (Fig. 1b). Similarly, in the BAU field experiment, fertilizer application significantly altered b-glucosaminidase (P<0.05), L-glutaminase (P<0.05) and urease activity (P<0.05), but unlike at BSMRAU, not always positively affected (Table 3). The order of variation among BAU treatments was similar to that of BSMRAU. For instance bglucosaminidase activity ranged between 42.03 and 62.08 mg PNP g1 dry soil h1, in line with previous reports of 20e80 mg PNP g1 dry soil h1 by Parham & Deng [14] and Ekenler & Tabatabai [17] for North American soils, however, with contrasting cropping pattern. A significantly higher b-glucosaminidase activity was measured in the NþFYM treated plots followed by the NPK and NP treatments, and then N and control treatments (Table 3). Unlike at BSMRAU, bglucosidase activity did not significantly differ among the BAU treatments, and closely matched observations in paddy soils of 14e32 mg PNP g1 dry soil h1 by Das et al. [18]. L-glutaminase activity was significantly higher in NP treatment at the BAU site, when compared to the unfertilized control, NPK and NþFYM treatments (Table 3).
3.3. Correlations between enzyme activities and soil properties In the BSMRAU experiment, only L-glutaminase correlated positively with soil N and SOC (P<0.01), and negatively with pHKCl (P<0.05). The other enzyme activities did not correlate significantly to any of the soil properties except for a negative correlation between arylamidase and pHKCl (P<0.05). However, in the BAU experiment, b-glucosaminidase activity positively correlated with soil N and SOC (P<0.01) and b-glucosidase activity with SOC (P<0.01) and the soil C:N ratio (P<0.05). In contrast to the BSMRAU experiment, L-glutaminase activity was not correlated with any of the investigated soil properties. Arylamidase activity again showed a negative correlation with pHKCl (P<0.05).
30
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
Table 1 N mineralization rates, total N, total SOC, C:N ratios and pHKCl(±standard error) of the two experimental sites (BSMRAU and BAU) (adapted from Kader [3]). Treatments
Total N (g kg1)
SOC (g kg1)
C:N ()
pHKCl ()
N mineralization rate (mg N kg1 day1) Aerobic
Anaerobic
BSMRAU Nitrogen 0 kg N ha1 220 kg N ha1
b
0.52 ± 0.06a 0.62 ± 0.09b
7.30 ± 0.69a 8.54 ± 1.00b
14.2 ± 0.7 13.9 ± 0.7
4.7 ± 0.1b 4.5 ± 0.1a
0.31 ± 0.05a 0.71 ± 0.25b
0.84 ± 0.18a 1.22 ± 0.28b
ANOVA
**a
*
NS
**
*
*
Organic matter Control Compost Cowdung Green manure Rice straw
0.55 0.62 0.55 0.57 0.52
ANOVA
NS
Nitrogen x OM 0 kg N ha1 Control Compost Cowdung Green manure Rice straw 220 kg N ha1 Control Compost Cowdung Green manure Rice straw
7.73 8.40 7.62 8.24 7.54
± ± ± ± ±
0.22 0.45 0.13 0.69 0.23
NS
14.1 13.6 13.9 14.5 14.5
± ± ± ± ±
1.2 0.8 0.2 0.8 0.5
NS
4.6 4.7 4.7 4.5 4.6
± ± ± ± ±
0.0 0.3 0.3 0.1 0.2
NS
0.29 0.64 0.54 0.52 0.59
± ± ± ± ±
0.00 0.25 0.26 0.26 0.24
NS
0.82 1.12 1.21 1.19 0.80
± ± ± ± ±
0.14 0.07 0.51 0.35 0.28
NS
± ± ± ± ±
0.11 0.03 0.01 0.01 0.16
7.57 7.72 7.37 7.76 6.11
± ± ± ± ±
0.74 0.08 0.57 0.26 0.87
13.3 13.8 14.2 15.1 14.9
± ± ± ± ±
0.7 0.5 0.3 0.4 1.9
4.6 4.9 4.9 4.6 4.8
± ± ± ± ±
0.0 0.1 0.1 0.0 0.1
0.29 0.39 0.28 0.33 0.28
± ± ± ± ±
0.04 0.16 0.13 0.02 0.09
0.76 1.10 0.90 1.00 0.64
± ± ± ± ±
0.04 0.13 0.19 0.27 0.03
0.54 0.67 0.73 0.64 0.57
± ± ± ± ±
0.06 0.04 0.05 0.12 0.08
7.53 8.73 9.43 8.98 8.04
± ± ± ± ±
0.82 0.61 0.32 0.81 0.33
14.1 13.4 13.0 14.1 14.2
± ± ± ± ±
0.1 0.9 0.2 0.7 0.7
4.6 4.4 4.5 4.4 4.4
± ± ± ± ±
0.0 0.0 0.0 0.0 0.1
0.29 0.89 0.79 0.70 0.90
± ± ± ± ±
0.03 0.06 0.12 0.04 0.17
1.01 1.22 1.65 1.50 1.03
± ± ± ± ±
0.03 0.17 0.04 0.27 0.16
± ± ± ± ±
0.48 0.83 0.04 0.68 0.44
NS
BAU Control N NP NPK NþFYM
1.60 1.64 1.75 1.78 1.77
ANOVA
*
b
0.03 0.14 0.02 0.08 0.16
0.57 0.56 0.52 0.52 0.41
ANOVA
a
± ± ± ± ±
NS ± ± ± ± ±
0.1a 0.0ab 0.0bc 0.1c 0.0c
15.1 15.3 15.8 17.0 16.6
NS ± ± ± ± ±
0.1a 0.2ab 0.9abc 1.2c 0.7bc
*
9.4 9.3 9.0 9.6 9.4 NS
* ± ± ± ± ±
0.6 0.1 0.3 0.2 0.4
5.7 5.9 5.7 5.7 5.7 **
NS ± ± ± ± ±
0.0a 0.0b 0.1a 0.0a 0.1a
1.97 1.54 2.24 1.22 1.23
NS ± ± ± ± ±
0.34b 0.12a 0.13b 0.04a 0.20a
**
1.54 1.98 1.92 1.51 0.98 NS
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
3.4. Correlations between N mineralization and soil enzyme activities Pearson correlation coefficients were calculated between the N mineralization rates (aerobic and anaerobic) and the enzyme activities for both experimental sites, separately (Table 5). In the BSMRAU experiment, only L-glutaminase activity correlated positively with the aerobic (P<0.05) and anaerobic (P<0.01) N mineralization rates. Likewise at the BAU experimental site, Lglutaminase activity correlated positively with both aerobic (P<0.05) and anaerobic (P<0.01) N mineralization rates. In addition, b-glucosaminidase activity was negatively correlated with both aerobic and anaerobic N mineralization rates (P<0.05). Lastly, bglucosidase and urease activities correlated negatively with aerobic (P<0.05) and anaerobic N mineralization rate (P<0.05).
with positive correlations among these variables (P < 0.05). However, across the BAU experimental treatments there was an unexpected negative correlation between SOC content and aerobic N mineralization rate (when expressed relatively to soil N content) reported by Kader [3]. There was no correlation with soil N level, in line with previously reported poor or unexpected dependency of anaerobic N mineralization and soil properties of Bangladeshi floodplain soils [4,5]. The BSMRAU terrace soil is inundated only during the rainy season for cultivation of rainfed Aman rice (one rice crop per year; 5 months) while the BAU floodplain soil is inundated for a longer period during the cultivation of irrigated Boro rice in winter season in addition to rainfed Aman rice in rainy season (two rice crops per year; 9 months). A longer inundation period at BAU site could have led to accumulation of organic N as previously reported [19,20], of which, an apparent substantial part was not readily mineralizable.
4. Discussion 4.2. Effect of OM and fertilizer application on soil enzyme activities 4.1. Effect of management on N mineralization Soil management affected N mineralization with a significant promotion of aerobic and anaerobic N mineralization with N fertilizer addition at the BSMRAU site while at the BAU site this was not the case. Soil management directly induced build-up of soil C and N at BSMRAU also led to elevate soil N mineralization rates
At both sites, the studied soil enzyme activities were significantly affected by the management practices used. Changes in soil enzyme activities were expressed in relative to the unfertilized controls to compare the responses of the different enzyme activities (Figs. 1 and 2). As stated by Ekenler & Tabatabai [17] and Dick [21], differences in the activity of enzymes result from differences
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
31
Table 2 Soil enzyme activities (±standard error) in incubated soil from the BSMRAU field experiment.
b-glucosaminidase (mg PNP g1 dry soil h1)
b-glucosidase (mg PNP g1 dry
L-glutaminase (mg N g1dry soilh1)
Urease (mg NH4-Ng1 dry soil h1)
Arylamidase (mg b-napthylamine g1 dry soil h1)
Nitrogen 0 kg N ha1 220 kg N ha1
b 43.0 ± 4.7a 45.5 ± 1.2b
21.6 ± 1.5a 27.5 ± 2.1b
1.7 ± 0.4a 4.4 ± 0.5b
8.1 ± 1.1 8.8 ± 1.7
5.4 ± 0.9a 9.5 ± 1.5b
ANOVA
*a
**
**
NS
**
Organic matter Control Compost Cowdung Green manure Rice straw
39.2 44.1 61.1 40.3 36.5
11.4 ± 2.2b 10.5 ± 1.7b 9.2 ± 1.8b 2.2 ± 0.4a 8.8 ± 2.1b
8.8 ± 2.8b 3.6 ± 1.5b 10.5 ± 2.9a 7.3 ± 1.6ab 6.9 ± 0.9ab
ANOVA
**
**
*
Treatments
± ± ± ± ±
2.9ab 1.0c 5.1d 2.5b 3.3a
soil h1)
18.8 23.0 24.3 28.9 27.8
± ± ± ± ±
1.8a 3.5b 2.7b 3.6c 2.8c
**
1.6 3.8 3.4 4.3 2.2
± ± ± ± ±
0.7a 0.3b 0.9b 0.2b 0.4a
**
Nitrogen x OM 0 kg N ha1 Control Compost Cowdung Green manure Rice straw 220 kg N ha1 Control Compost Cowdung Green manure Rice straw
34.4 43.7 69.8 36.0 30.9
± ± ± ± ±
1.8c 1.9b 1.9a 1.1bc 0.3c
16.4 17.4 29.1 23.0 22.9
± ± ± ± ±
1.8b 1.3b 1.4a 0.4ab 0.0ab
0.4 4.0 1.4 2.1 0.8
± ± ± ± ±
0.2c 0.1a 0.4ab 0.2b 0.2ab
8.8 ± 1.8ab 9.6 ± 0.8ab 7.0 ± 1.8ab 2.6 ± 0.8b 12.3 ± 0.0a
4.8 1.6 5.6 9.3 5.6
44.0 44.4 52.5 44.6 42.2
± ± ± ± ±
0.7b 1.5b 1.3a 0.8b 0.3b
21.3 28.8 20.0 35.7 32.8
± ± ± ± ±
1.8b 2.7ab 1.4b 0.9a 0.1a
2.8 3.6 5.5 6.6 3.7
± ± ± ± ±
0.1b 0.1b 0.6ab 0.9a 0.0b
14.0 ± 1.8a 11.4 ± 2.6a 11.4 ± 2.6a 1.8 ± 0.0a 5.3 ± 1.8a
12.8 ± 3.9a 5.6 ± 2.2a 15.5 ± 1.3a 5.4 ± 1.4a 8.2 ± 0.7a
ANOVA
**
NS
*
a b
**
**
± ± ± ± ±
0.9a 0.3a 1.3a 2.5a 0.9a
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
in quantity and quality of crop residues and resulting soil microbial activity. The relationship is complex, however, because enzyme activities represent the cumulative effects of substrate availability, demand for the product, and connected environmental conditions and resulting microbial activity [22]. Utilizing this framework, the effects of fertilizer and OM application on enzyme activities were interpreted. 4.2.1. Substrate availability It is not surprising that repeated OM application at the BSMRAU site increased activity of b-glucosidase (Table 2) as it catalyses biodegradation of b-glucosides in plant debris [23,24]. Amendment with rice straw and green manure, i.e. carbohydrate rich substrates, relatively promoted b-glucosidase activities compared to application of compost and cow dung, likely because these already have gone through intensive microbial transformation. In line, compost and cow dung amendments promoted b-glucosaminidase activity, which may be attributed to their higher content of animal glucosamine polymers and microbial cell wall N (peptidoglycan, chitin), targeted by b-glucosaminidase [25]. With multiple sources of C and N in the soil, Geisseler & Horwath [26] found that enzymes participating in protein degradation would also be controlled primarily by the concentration of nitrogen substrate. The higher Lglutaminase activity in compost, cow dung and green manure amended BSMRAU plots compared to rice straw and unamended plots may stem from differences in proteinaceous N availability. In the BSMRAU site, application of 220 kg N ha1 furthermore enhanced all studied enzyme activities, alongside crop growth, SOC and soil N. Most likely, N-application increased OM inputs from residual roots, stubble and root exudates, and thereby indirectly increased SOC and N levels (Table 1) as well as energy sources for microbial activity. While aminohydrolases show rapid responses to changing substrate levels [27], trends in urease and arylamidase
activities over the OM treatments did not follow the expected pattern of enhanced substrate availability (Fig. 1b). In fact, urease activity was lowered by OM application in the plots applied with 220 kg N ha1 as urea (Fig. 1b). This contradictory result may be explained by the much higher levels of NHþ 4 (Table 1), which suppresses urease activity [28]. In general, the BSMRAU experiment demonstrated that repeated (22 years) OM amendment without N fertilizer application promotes soil catalytic activity. In the BAU experiment, however, the variation of the enzyme activities over the OM treatments was less consistent. The NP, NPK and NþFYM application significantly increased soil N and SOC (P < 0.05) and positively correlated with b-glucosaminidase activity (Table 4), which is in agreement with the reasoning that substrate availability would dominantly control enzyme production. However, NPK and NþFYM application strongly lowered arylamidase and L-glutaminase activity relative to the unfertilized control, in spite of the elevated SOC and soil N levels (Fig. 2). Also, b-glucosidase activity did not respond to fertilizer application at the BAU site, which is in clear contrast to the BSMRAU experiment. These contrasting trends in soil N, SOC and studied N mineralizing enzymes (urease, arylamidase and L-glutaminase) demonstrate that a promotion of enzyme activity with increasing SOM levels is enzyme specific at the BAU site. In paddy soils with prolonged inundation N can be stabilized through binding with aromatic rings of anaerobically accumulated lignin and phenolic subunits [19,20,29] and this may be the case in the NPK and NþFYM treatments with elevated SOC. In summary, it is clear that environmental conditions and microbial nutrient demand should be used as additional frameworks to explain these observations. 4.2.2. Microbial nutrient demand and environmental conditions As extracellular enzyme activity mediates microbial nutrient acquisition from organic matter, these activities are alternatively,
32
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
Fig. 2. Relative changes of b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase enzyme activities in different treatments over control in a young floodplain soil of field experiment (BAU). Positive and negative values indicate higher and lower enzyme activities compared to the unfertilized control: e.g. þ47% for urease in case of the N treatment indicates a nearly half as high activity with mineral N applied vs. the unamended plots. Treatments indicated by different letters have statistically different (P < 0.05) means per enzyme activity according to Duncan's multiple range post-hoc test.
often interpreted as indicators of microbial nutrient demand. At the BSMRAU site, annual application of 220 kg N ha1 would then be expected to lower the microbial demand for N. However, activities of L-glutaminase, arylamidase and b-glucosaminidase were all significantly higher than in the 0 kg N ha1 counterpart soil plots. Hence, the demand for product, in this case mineral N, would consequently seem to hold lesser control on levels of these enzymes in the BSMRAU plots, corroborating the previously stated reasoning that instead substrate availability would. However, 220 kg N ha1 application did interact with the response of enzyme activities to OM-amendment (Table 2), rendering L-glutaminase and b-glucosaminidase activities less responsive to OM application (Fig. 1). Cow dung, green manure and rice straw application resulted in 40e80% lower arylamidase activity compared to the unamended control, unlike in the 0 kg N ha1 treated soils (Fig. 1). Possibly, the combined N-supply from mineral N and OM application rendered N non-limiting for microbes, relieving the need to produce hydrolytic enzymes. Hence, the established relationship where enzyme activity is dictated by N-demand by the microbial
Fig. 1. Relative changes of b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase enzyme activities in different exogenous OM treatments over the control treatment (a) without N (0 kg N ha1) and (b) with 220 kg N ha1 application in a highly weathered terrace soil of field experiment (BSMRAU). Positive and negative values indicate higher and lower enzyme activities compared to the unamended control at 0 kg N ha1 (a) and 220 kg N ha1 levels (b): e.g. þ102% for b-glucosaminidase in case of the compost treatment at 0 kg N ha1 indicates an approximately twice as high activity than in the unamended plots at 0 kg N ha1. Note the differing right axis for L-glutaminase activity in Fig. 1(a). Treatments indicated by different letters have statistically different (P < 0.05) means per enzyme activity according to Duncan's multiple range post-hoc test.
Table 3 Soil enzyme activities (±standard error) in incubated soil from the BAU field experiment. Treatments
b-glucosaminidase (mg PNP g1 dry soil h1)
b-glucosidase (mg PNP g1 dry soil h1)
Control N NP NPK NþFYM
b
42.0 ± 0.7a 44.9 ± 2.0a 50.4 ± 1.4b 55.4 ± 1.3c 62.1 ± 2.4d
17.3 16.8 15.0 21.4 17.6
ANOVA
*a
NS
a b
± ± ± ± ±
0.3 1.1 0.4 2.3 1.5
L-glutaminase (mg N g1 soilh1)
5.2 6.7 8.3 3.7 2.3 *
± ± ± ± ±
0.3bc 0.9cd 0.6d 0.6ab 0.7a
Urease (mg NH4-Ng1 dry soil h1) 3.5 5.3 2.6 2.9 8.2 *
± ± ± ± ±
0.0a 1.0ab 0.5a 1.2a 1.5b
Arylamidase (mg b-napthylamine g1 dry soil h1) 15.6 ± 5.1 16.1 ± 3.2 16.0 ± 3.1 5.6 ± 1.8 6.5 ± 1.6 NS
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
33
Table 4 Pearson's correlation coefficients between soil enzyme activities and general soil properties. BSMRAU (n ¼ 20)
Enzymes
b-glucosaminidase b-glucosidase L-glutaminase Urease Arylamidase a
BAU (n ¼ 15)
Soil N
SOC
C:N
pH
Soil N
SOC
C:N
pH
0.13 0.13 0.69**a 0.02 0.24
0.09 0.12 0.69** 0.18 0.27
0.20 0.11 0.06 0.14 0.13
0.11 0.30 0.45* 0.20 0.49*
0.72** 0.38 0.19 0.02 0.24
0.73** 0.70** 0.32 0.18 0.27
0.11 0.53* 0.23 0.14 0.13
0.22 0.11 0.01 0.20 0.49*
* Correlation is significant at the 0.05 level, ** Correlation is significant at the 0.01 level.
Table 5 Pearson's correlation coefficient between N mineralization rate (aerobic and anaerobic) and enzyme activities. Parameters
b-glucosaminidase b-glucosidase L-glutaminase
Urease Arylamidase
N Mineralization rate (mg N kg1 day1) BSMRAU (n ¼ 20)
BAU (n ¼ 15)
Aerobic
Anaerobic
Aerobic
Anaerobic
0.10 0.41 0.49*a 0.05 0.16
0.28 0.19 0.75** 0.11 0.37
0.59* 0.52* 0.77** 0.45 0.50
0.55* 0.06 0.54* 0.56* 0.33
a *Correlation is significant at the 0.05 level, ** Correlation is significant at the 0.01 level.
biomass, in turn determined by the elemental stoichiometry of microbial biomass relative to environmental nutrient availability [10], appears only partly relevant for the BSMRAU site. In case of the BAU experiment, only arylamidase and L-glutaminase responded similarly to the fertilizer applications, with a significantly lower activity in the NPK and NþFYM treatments (Fig. 2). Both NPK and NþFYM treatments differ from the other plots by the application of K, a frequently limiting nutrient for plant growth in double rice cropping with repeated large exports of K in rice straw. Indeed, the BAU site yield data shows (see supplementary material) that above-ground biomass and crop yields were significantly elevated in NPK and NþFYM treatments compared to non-K treated plots. This higher plant productivity resulted in build-up of soil N and SOC but anaerobic N mineralization was comparatively lower. Nourbakhsh & Alinejadian [30] state that anaerobic conditions induced a limiting situation for microbial activities and therefore, NH3 release rate during anaerobic incubation was greatly controlled by the microbial activities rather than bioavailability of N. There was no direct quantification of microbial activity in the present experiments but the b-glucosidase activity as a proxy for Cdemand could be considered. It is then of further interest to compare enzymes involved in N-mineralization vis- a-vis to bglucosidase activity. At BAU site, the latter was unresponsive to the fertilizer treatments, and so microbial activity and consequently microbial nutrient demand, are not suggested to differ strongly, compared to L-glutaminase and arylamidase activity. Therefore, in contrast to the BSMRAU site, it cannot be concluded that microbial activity primarily determines production of extra-cellular enzymes catalyzing N-mineralization at BAU site. Hence, it appears that at the BAU site, these activities were disconnected from microbial demand, but depend on substrate quality and content. Differences in SOM quality between the BAU fertilizer treatments might thus explain this disconnection between soil N mineralization and enzymatic activity and thus soil N level. Schmidt-Rohr et al. [29] found that significant amounts of the amide N were directly bound to aromatic rings in paddy soils, leading them to conclude that N bound to lignin accounts for its reduced bioavailability. Such
a specific reduction of amide N-availability with SOM accumulation in the NPK andNPKþFYM treatments might explain the lower arylamidase and L-glutaminase activities. b-glucosaminidase on the contrary did follow soil N content, which was expected as no specific stabilization of amino-sugars to phenolics is known. Alternatively, SOM content in organic amended soils can increase the capacity of the soils to protect extra-cellular pools of soil enzymes against proteolytic activities and hence, provide more functional enzyme molecules in soil [31]. If so, it could be hypothesized that less arylamidases and L-glutaminases were simply required with higher SOC content, however, the present experimental setup does not allow verification of this mechanism. 4.3. Enzymatically catalyzed SOM hydrolysis as bottleneck for anaerobic soil N mineralization A negative correlation between b-glucosidase activity and aerobic N mineralization rate (P<0.05) for the BAU site is in agreement with previously reported findings [6,32e34]. Higher b-glucosidase activities indirectly represent microbial growth, which temporarily immobilizes N from the soil as the C:N ratio of bacteria and fungi is smaller than the C:N ratio of SOM. Nitrogen acquisition from the aerobic system at BAU site followed the elemental stoichiometric theory [10] as observed by the positive correlation (r¼0.82) between the ratio of L-glutaminase and b-glucosidase enzymes as well as by aerobic N mineralization (Table 5). The anaerobic N mineralization rate, instead, was significantly correlated to L-glutaminase activity in both field experiments (r¼0.75, P < 0.01 and r¼0.54, P < 0.05 for BSMRAU and BAU, respectively), which is align with findings of Nourbakhsh & Alinejadian [30]. L-glutaminase is involved in the final step of N-mineralization where a terminal amino group of a peptide or amino acid is converted into NH3. Arylamidase, alternatively catalyses the hydrolysis of various amino acids from arylamides but does not directly induce the release of NH3 and was not significantly correlated to soil N mineralization. This suggests that at both BSMRAU and BAU sites, terminal amino acid hydrolysis may have formed a bottleneck in anaerobic N mineralization, as represented by soil NHþ 4 -N build-up. Earlier steps in soil N mineralization seem less rate limiting: b-glucosaminidase activity was involved in degradation of chitin and peptidoglucan, i.e. polymers in microbial cell walls, and even correlated negatively (Table 5) to aerobic and anaerobic N mineralization rate at BAU site. 5. Conclusion The dependency of soil N supply on SOM quality and SOM quantity and general soil properties of Bangladeshi floodplain paddy soils is complex [4,5]. This was also demonstrated by earlier study of two field experiments with increasing N mineralization at BSMRAU with soil N and an opposite trend at BAU site, suggesting instead, stabilization of N with OM build-up. An overarching positive relationship between anaerobic N mineralization and an
34
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
amidohydrolase enzyme activity (here L-glutaminase) may suggest that terminal amino acid cleavage and not preceding depolymerisation steps control anaerobic N mineralization in these paddy soils. A link between an amidohydrolase directly involved in the release of proteinaceous N, and anaerobic N mineralization did appear logical, and however, it is not clear which are the dominant controlling factors. Limited bio-availability of N, through association with accumulating phenols with prolonged annual inundation may be one of the constraints. Further research combining probing of SOM biochemistry, particularly by quantifying contents of polyphenols and bound-N, alongside amidohydrolase activity assays will be needed to confirm this hypothesized control of Nbioavailability on N-mineralization in young paddy soils found throughout Southeast Asia. References [1] D.R. Keeney, Nitrogendavailability indices, in: A.L. Page, R.H. Miller, D.R. Keeney (Eds.), Methods of Soil Analysis. Part 2, second ed., Agronomy, 9, SSSA and ASA, Madison, WI, 1982, pp. 711e733. Chemical and Microbiological Properties. [2] G. Stanford, Assessment of soil nitrogen availability, in: F.J. Stevenson (Ed.), Nitrogen in Agricultural Soils, Agronomy Monograph, 22, ASA, Madison, WI, 1982, pp. 651e688. [3] M.A. Kader, Nitrogen Mineralization in Subtropical Paddy Soils in Relation to Soil Properties, Organic Matter Fractions, and Fertilizer Management, Ph.D. thesis, Faculty of Bioscience Engineering, Ghent University, Ghent, Belgium, 2012, 213 pp.. [4] M.A. Kader, S. Sleutel, S.A. Begum, A.Z.M. Moslehuddin, S. De Neve, Nitrogen mineralization in sub-tropical paddy soils in relation to soil mineralogy, management, pH, carbon, nitrogen and iron contents, Eur. J. Soil Sci. 64 (2013) 47e57. [5] S. Sleutel, M.A. Kader, K. Demeestere, C. Walgraeve, J. Dewulf, S. De Neve, Subcritical water extraction to isolate kinetically different soil nitrogen fractions, Biogeosciences 10 (2013) 7435e7447. [6] M.A. Tabatabai, M. Ekenler, Z.N. Senwo, Significance of enzyme activities in soil nitrogen mineralization, Commun. Soil Sci. Plant Anal. 41 (2010) 595e605. [7] N. Khorsandi, F. Nourbakhsh, Prediction of potentially mineralizable N from amidohydrolase activities in a manure-applied, corn residue-amended soil, Eur. J. Soil Biol. 44 (2008) 341e346. [8] V. Acosta-Martínez, M.A. Tabatabai, Arylamidase activity of soils, Soil Sci. Soc. Am. J. 64 (2000) 215e221. [9] S. Muruganandam, D.W. Israel, W.P. Robarge, Activities of nitrogenmineralization enzymes associated with soil aggregate size fractions of three tillage systems, Soil Sci. Soc. Am. J. 73 (2009) 751e759. [10] R.L. Sinsabaugh, B.H. Hill, J.J. Follstad Shah, Ecoenzymatic stoichiometry of microbial organic nutrient acquisition in soil and sediment, Nature 462 (2009) 795e799. [11] M.A. Tabatabai, J.M. Bremner, Assay of urease activity in soils, Soil Biol. Biochem. 4 (1972) 479e487. [12] Soil Survey Staff, Soil Taxonomy: a Basic System of Soil Classification for Making and Interpreting Soil Surveys, second ed., Natural Resources Conservation Service. U.S. Department of Agriculture Handbook 436, 1999. [13] A.J.M.S. Karim, K. Egashira, Y. Yamada, J. Haider, K. Nahar, Long-term application of organic residues to improve soil properties and to increase crop yield
in terrace soil of Bangladesh, J. Fac. Agric. Kyushu U 39 (1995) 149e165. [14] J.A. Parham, S.P. Deng, Detection, quantification and characterization of bglucosaminidase activity in soil, Soil Biol. Biochem. 32 (2000) 1183e1190. [15] F. Eivazi, M.A. Tabatabai, Glucosidases and galactosidases in soils, Soil Biol. Biochem. 20 (1988) 601e606. [16] W.T. Frankenberger, M.A. Tabatabai, L-Glutaminase activity of soils, Soil Biol. Biochem. 23 (1991) 869e874. [17] M. Ekenler, M.A. Tabatabai, b-Glucosaminidase activity of soils: effect of cropping systems and its relationship to nitrogen mineralization, Biol. Fertil. Soils 36 (2002) 367e376. [18] S.K. Das, A. Varma, Role of enzymes in maintaining soil health, in: G. Shukla, A. Varma (Eds.), Soil Enzymology, Springer, Berlin, 2011, pp. 293e312. [19] P.J. Colberg, Anaerobic microbial degradation of cellulose, lignin, oligolinols, and monaromatic lignin derivatives, in: A.J.B. Zender (Ed.), Biology of Anaerobic Microorganisms, Wiley, New York, 1988, pp. 333e372. [20] D.C. Olk, K.G. Cassman, E.W. Randall, P. Kinchesh, L.J. Sanger, J.M. Anderson, Changes in chemical properties of organic matter with intensified rice cropping in tropical rice soil, Eur. J. Soil Sci. 47 (1996) 293e303. [21] R.P. Dick, Soil enzyme activities as indicators of soil quality, in: J.W. Doran, D.C. Coleman, D.F. Bezdicek, B.A. Stewart (Eds.), Defining Soil Quality for a Sustainable Environment, Soil Science Society of America Journal, Madison, WI, 1994, pp. 108e123. [22] S.D. Allison, M.N. Weintraub, T.B. Gartner, M.P. Waldrop, Evolutionary-economic principles as regulators of soil enzyme production and ecosystem function, in: G. Shukla, A. Varma (Eds.), Soil Enzymology, Springer, Berlin, 2011, pp. 229e244. [23] H.A. Ajwa, M.A. Tabatabai, Decomposition of different organic materials in soils, Biol. Fert. Soils 18 (1994) 175e182. [24] C.E. Martinez, M.A. Tabatabai, Decomposition of biotechnology byproducts in soils, J. Environ. Qual. 26 (1997) 625e632. [25] R.G. Burns, J.L. DeForest, J. Marxsen, R.L. Sinsabaugh, M.E. Stromberger, M.D. Wallenstein, M.N. Weintraub, A. Zoppini, Soil enzymes in a changing environment: current knowledge and future directions, Soil Biol. Biochem. 58 (2013) 216e234. [26] D. Geisseler, W.R. Horwath, Relationship between carbon and nitrogen availability and extracellular enzyme activities in soil, Pedobiologia 53 (2009) 87e98. [27] S.P. Deng, M.A. Tabatabai, Effect of tillage and residue management on enzyme activities in soils. I. Amidohydrolases, Biol. Fert. Soils 22 (1996) 202e207. [28] D. Pandey, M. Agrawal, J.S. Bohra, Effects of conventional tillage and no tillage permutations on extracellular enzyme activities and microbial biomass under rice cultivation, Soil till. Res. 136 (2014) 51e60. [29] K. Schmidt-Rohr, J.D. Mao, D.C. Olk, Nitrogen-bonded aromatics in soil organic matter and their implications for a yield decline in intensive rice cropping, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 6351e6354. [30] F. Nourbakhsh, A. Alinejadian, Arginine ammonification and L-glutaminase assays as rapid indices of corn nitrogen availability, J. Plant Nutr. Soil Sci. 172 (2009) 127e133. [31] P. Nannipieri, P. Sequi, P. Fusi, Humus and enzyme activity, in: A. Piccolo (Ed.), Humic Substances in Terrestrial Ecosystems, Elsevier, Amsterdam, 1996, pp. 293e327. [32] D.L. Burton, W.B. McGill, Spatial and temporal fluctuation in biomass, nitrogen mineralization reactions, and mineral nitrogen in a soil cropped to barley, Can. J. Soil Sci. 72 (1992) 31e42. [33] D.E. Dodor, M.A. Tabatabai, Amidohydrolases in soils as affected by cropping systems, Appl. Soil Ecol. 24 (2003) 73e90. [34] N. Khorsandi, F. Nourbakhsh, Effect of amendment of manure and corn residues on soil N mineralization and enzyme activity, Agron. Sustain. Dev. 27 (2007) 139e143.
Contents lists available at ScienceDirect
European Journal of Soil Biology journal homepage: http://www.elsevier.com/locate/ejsobi
Original article
Response of hydrolytic enzyme activities and nitrogen mineralization to fertilizer and organic matter application in subtropical paddy soils M.A. Kader a, c, e, *, S. Yeasmin a, Z.M. Solaiman d, S. De Neve b, S. Sleutel b a
Department of Soil Science, Bangladesh Agricultural University, Mymensingh 2202, Bangladesh Department of Soil Management, Ghent University, Coupure Links 653, 9000 Gent, Belgium c School of Veterinary and Life Sciences, Murdoch University, South St, Murdoch, WA 6150, Australia d UWA School of Agriculture and Environment, and The UWA Institute of Agriculture, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia e School of Agriculture and Food Technology, The University of the South Pacific, Alafua Campus, Samoa b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 July 2016 Received in revised form 24 December 2016 Accepted 15 March 2017
Drivers of nitrogen (N) mineralization in paddy soils, especially under anaerobic soil conditions, are elusive. The influences of exogenous organic matter (OM) and fertilizer application on the activities of five relevant enzymes (b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase) were measured in two long-term field experiments. Of the two field experiments, the 18-year field experiment was established in a weathered terrace soil with rice-wheat crop rotation at the Bangabandhu Sheikh Mujibur Rahman Agricultural University (BSMRAU) farm with five OM treatments and two levels of mineral N fertilizer. The 30-year experiment was established in a young floodplain soil with rice-rice crop rotation at the Bangladesh Agricultural University (BAU) farm with five mineral fertilizer treatments including one with farm yard manure. At BSMRAU, N fertilizer and OM amendments significantly increased all enzyme activities, suggesting the availability of primarily substrate for microbial activity. Whereas at BAU, non-responsiveness of b-glucosidase activity, suggesting that fertilizer and OM amendments had little effect on overall soil microbial activity. Nevertheless the microbial demand for N, b-glucosaminidase and L-glutaminase activities differed among the treatments (P < 0.05) and showed opposite trends with soil N mineralization. Hence enzymatic pathways to acquire N differed with the treatment at BAU site, indicates differences in soil N quality and bio-availability. L-glutaminase activity was the sole investigated variable that positively correlated to both the aerobic and anaerobic N mineralization rates in both field experiments. Combined with a negative correlation between b-glucosaminidase activity and N mineralization rate, it appears that terminal amino acid NH2 hydrolysis was a rate-limiting step for soil N mineralization at the BAU site. Future investigations with joint quantification of polyphenol accumulation and binding of N alongside an array of extracellular enzymes, including oxidases for phenols and hydrolases for N-compounds, would enable verification of the hypothesized binding and stabilization of N with accumulating polyphenols at BAU site under SOM storing management. © 2017 Elsevier Masson SAS. All rights reserved.
Handling editor: Christoph Tebbe Keywords: Enzyme activity Fertilization Long-term experiment N mineralization Organic amendment Paddy soil
1. Introduction A better understanding of the factors controlling nitrogen (N) mineralization in paddy soils and development of practical
* Corresponding author. Department of Soil Science, Bangladesh Agricultural University, Mymensingh 2202, Bangladesh; School of Veterinary and Life Sciences, Murdoch University, South St, Murdoch, WA 6150, Australia; School of Agriculture and Food Technology, The University of the South Pacific, Alafua Campus, Samoa. E-mail address: [email protected] (M.A. Kader). http://dx.doi.org/10.1016/j.ejsobi.2017.03.004 1164-5563/© 2017 Elsevier Masson SAS. All rights reserved.
indicators of soil N supply is essential to improve N-use efficiency in South-east Asian paddy rice production, thereby reducing the application of relatively expensive N fertilizers. Several biological and chemical methods have been proposed as N mineralization indices [1,2]. However, limited progress has been made on reliable prediction of paddy soil N mineralization. In our previous work, basic soil properties as well as an array of physicochemical soil organic matter (SOM) fractions have been tested for the prediction of potential N mineralization from Bangladeshi paddy soils under laboratory incubations [3e5]. While some of these soil variables
28
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
correlated with the potential aerobic N mineralization rate, relationships with anaerobic N mineralization were mostly insignificant or negative. Hence it was inferred that neither SOM quantity, nor SOM quality dominantly determine the anaerobic N mineralization process. A multitude of other biotic or abiotic constraining factors, which are not expressed in readily measurable soil variables or SOM fractions, could control the anaerobic N mineralization in paddy soils. As organic N mineralization is mediated by microbial extracellular enzymes, assays of their activity should provide insight into key intermediate soil biochemical processes [6] and if successful, could be used as sensitive N mineralization indexes. Enzyme activities are the end result of the interaction of SOM biochemistry and physical soil conditions, both being shaped by management. The activities of urease and L-asparaginase and therefore the soil amidohydrolase activities in general have the potential to evaluate mineralizable N [7]. Tabatabai et al. [6] proposed N-acetyl-ß-Dglucosaminidase activity as an index of soil N mineralization among six amidohydrolases enzymes involved in N cycling and four glycosidase enzymes involved in carbon cycling in soils. However, such conclusions are not yet supported for flooded paddy soils by the lack of experimental data. Specific practices, i.e., wet cultivation, puddling, alternate wetting and drying make paddy soils distinct in physical, chemical and biological properties and role of enzymes to more frequently studied soil types. The present study considered five relevant ecoenzymes, covering initial and terminal steps in organic matter and N mineralization and urea hydrolysis. b-glucosaminidase and arylamidase are selected as representatives for extracellular enzymatic breakdown of complex organic N compounds into amides, amino sugars, and amino acids (aminization). This is assumed to be an initial rate-limiting step in soil N mineralization [8]. L-glutaminase was selected as representative for an array of enzymes involved in the production of NHþ 4 from amino acids through ammonification [9]. Microbial nutrient demand is determined by the elemental stoichiometry of microbial biomass in relation to environmental nutrient availability [10]. b-glucosidase activity, involved in cellobiose hydrolase and carbon (C) and energy supply, was therefore included as well, as this allowed further evaluation of the eco-enzymatic C:N-ratio as an indication of the tendency of microbial activity to be determined for nutrient or energy flow. Lastly, urease is an enzyme that catalyses the hydrolysis of urea, hydroxyurea, dihydroxyurea and semicarbazid into CO2 and NH3 [11]. The principal aim of this study was to evaluate the relative control of these selected enzymatic steps on aerobic and anaerobic N mineralization in young floodplain paddy soils. We interpret strong correlations between N mineralization rate and enzyme activity as likelihood that the mediated OM-transformation step would be limiting N mineralization. The secondary aim was to elucidate if enzyme activity was either determined by product demand or by substrate availability. It was hypothesized that: N fertilizer addition would reduce differences in enzyme activities between exogenous OM treatments due to a lifting of product demand, in casu mineral N; exogenous OM application would promote activity of all enzymes because of a generally enhanced substrate availability; a similar effect in mineral fertilizer treatments that would promote crop growth, and logically larger OM inputs from root exudation and incorporation of crop residues; and a higher demand for mineral N in treatments with exogenous OM with higher C:N ratio to result in a specific promotion of the activity of hydrolytic enzymes mediating N-transformations relative to bglucosidase.
2. Materials and methods 2.1. Site description and soil sampling Soil samples were collected from two long-term field experiments. One experiment was established in 1989 at the Bangabandhu Sheikh Mujibur Rahman Agricultural University (BSMRAU) farm at Salna (24 050 N, 90160 E), Bangladesh on a clayey, kaolinitic, Ultic Ustocrept soil [12] developed from Madhupur clay. The soil texture was silty clay loam (15:46:39) [13]. A yearly Wheat (variety Akbar)-Fallow-Transplanted rice (variety T. Aman) cropping pattern was practised. The BSMRAU field trial involved five OM application treatments (1 no-application, 2 rice straw (air dry) at 2 Mg ha1, 3 green manure at 7.5 Mg ha1 fresh biomass of Sesbania rostrata, 4 compost at 25 Mg ha1 made from cow dung and rice straw, and 5 cow dung at 25 Mg ha1 as fresh manure). These were combined with 2 levels of inorganic N fertilizer amendments (0 and 100 kg N ha1 for rice and 0 and 120 kg N ha1 for wheat). Triple super phosphate (TSP) was applied at 44 kg P ha1 for both crops and muriate of potash (KCl) was applied at 55 and 45 kg K ha1 for rice and wheat, respectively, during the final land preparation. The experiment was laid out in a 5 3 factorial design with replications. The dimension of each plot was 12 m 7 m with a plot-to-plot distance of 1.5 m. Details on the agronomic management can be found in Ref. [13]. The second field experiment was established in 1978 at Bangladesh Agricultural University (BAU) farm at Mymensingh (24 430 N, 90 250 E), Bangladesh on a loamy, mixed, non-acidic Aeric Haplaquept [12], developed from old Brahmaputra alluvium. The soil texture was silt loam (19:63:18). The treatments all had a yearly Boro rice (irrigated winter rice transplanted on midJanuary and harvested mid-May)-Fallow-T. Aman rotation and included treatments with application of mineral fertilizer (control, N, NP, NPK) and one with application of mineral N and farmyard manure (NþFYM). F The application rates of N, P, K, S, and Zn per crop were 90, 20, 19, 30, and 5 kg ha1, respectively applied as urea, triple super phosphate, potassium chloride, gypsum, and zinc oxide. Cow dung was mixed with rice straw applied once a year 10e15 days prior transplantation of Boro rice at a rate of 5 Mg ha1 fresh material. The experiment was conducted in a randomized block design with three replications (12 m 6 m). Details on the agronomic management can be found in Ref. [13]. Surface soil samples (0e15 cm) were collected from 15 locations per replicate plot by means of a 2.5 cm inner diameter auger in May 2008 at BSMRAU, and in July 2008 at BAU. These samples were bulked into one composite sample and thoroughly mixed. The field moist soil was gently broken apart by hand and air-dried and ground to pass a 2-mm sieve prior to the assessment of N mineralization, and enzyme activities. 2.2. Nitrogen mineralization Fourteen-week laboratory incubations were carried out to determine both aerobic and anaerobic N mineralization rates, as described in detail by Kader et al. [4]. Three replicate plot soils per treatment at BAU and two replicate plot soils per treatment at BSMRAU were used to quantify potential N mineralization. In total 42 tubes were filled for each BAU treatment such as 21 tubes (3 replicates 7 dates) for aerobic and 21 tubes for anaerobic incubation; and 28 tubes per BSMRAU treatment such as 14 tubes (2 replicates 7 dates) for aerobic and 14 for anaerobic incubation. Removal of mineral N from soil solution by denitrification or immobilization was not considered and only net N mineralization was measured in the incubation experiments. The net aerobic N mineralization data were best described by a zero-order kinetic
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
model: N(t) ¼ N0 þ k0t, where t is the time (in days), N(t) is the amount of mineral N at time t, N0 is the initial amount of mineral N (mg N kg1), and k0 is the zero-order N mineralization rate (mg N kg1 day1). The anaerobic N mineralization data were best described with a first-order model: N(t) ¼ NA(1-e-k1t), with NA the mineralizable N and k1 a first-order rate parameter. Average mineralization rates (in mg N kg1 day1) were then derived for the entire 14-week incubation period. 2.3. Enzyme activity analysis Five enzyme activities were measured in one-week pre-incubated soil (at 27 C, 60% water-filled pore space and at field bulk density). The b-glucosaminidase (EC 3.2.1.30), b-glucosidase (EC 3.2.1.21), L-glutaminase (EC 3.5.1.2), urease (EC 3.5.1.5) and arylamidase (EC 3.4.11.2) activities were quantified according to Parham & Deng [14], Eivazi & Tabatabai [15], Frankenberger & Tabatabai [16], Tabatabai & Bremner [11] and Acosta-Martinez & Tabatabai [8], respectively. For determination of all enzymes, controls were included, in which the substrate was added only after incubation and termination of the enzyme reaction. All results reported were averages of duplicate assays and expressed on a moisture-free soil mass basis. Moisture was determined from weight loss after drying at 105 C for 48 h. 2.4. Statistical analysis All statistical analyses were carried out in SPSS Version 19. Nonlinear regression analysis with the Levenberg-Marquardt algorithm was used to estimate the first-order model for N mineralization kinetic parameters. ANOVA and Duncan's post-hoc mean comparison tests were performed to assess significant differences in the mean enzyme activities of the OM and fertilizer application treatments. Pearson's correlation analysis was used to investigate the relationship between the tested enzyme activities and the aerobic and anaerobic N mineralization rates. 3. Results 3.1. Nitrogen mineralization The influence of OM amendment and N fertilizer application on the evolution of mineral N in soil and N mineralization rates in both field experiments was described before in detail by Kader [3]. Briefly, there was an overall significant positive effect (P<0.05) of mineral N application on both aerobic and anaerobic N mineralization rates (mg N kg1 day1) in the BSMRAU field experiment (Table 1). N mineralization rates did not significantly differ between the various exogenous OM application treatments. Both anaerobic as well as aerobic N mineralization rates correlated positively with soil N content (r¼0.46, P<0.05 and r¼0.60, P<0.01, respectively) and SOC content (r¼0.60, P<0.01 and r¼0.68, P<0.01, respectively). The aerobic N mineralization rate correlated negatively with soil C:Nratio (r¼0.61, P<0.05). In the BAU floodplain soil experiment, the aerobic N mineralization rate differed significantly (P<0.01) among the mineral fertilizer treatments (Table 1). Aerobic N mineralization rates were statistically identical in NP and control treated plots (1.97e2.24 mg N kg1 day1) and higher than in the N, NPK and NþFYM treated plots (1.22e1.54 mg N kg1 day1). Anaerobic N mineralization rate exhibited a very strong variation and did not differ significantly between the treatments (0.98e1.98 mg N kg1 day1) (Table 1). Neither aerobic nor anaerobic N mineralization rate correlated with SOC, soil N content, C:N-ratio or pHKCl.
29
3.2. Enzyme activities For the BSMRAU site, all the measured enzymes activities were significantly influenced by OM treatment (b-glucosaminidase (P<0.01), b-glucosidase (P<0.01), L-glutaminase (P<0.01), urease (P<0.01) and arylamidase (P<0.05)) and there were interaction effects of OM-application and mineral N supply (b-glucosaminidase (P<0.01), b-glucosidase (P<0.01), L-glutaminase (P<0.01) and arylamidase (P<0.05)). An exception was a nil interaction effect of OM and N fertilizer addition on urease activity (Table 2). Among the OM treatments, the b-glucosaminidase activity followed the order: cow dung > compost > green manure > no application > rice straw application (Table 2). The lowest b-glucosidase activity was observed in the control followed by cow dung and compost treatments while green manure and rice straw treated soils had significantly higher activities (Table 2). The control and rice straw treatments had significantly lower L-glutaminase activities compared with the compost, green manure and cow dung treatments (Table 2), with an exceptionally higher activity in the latter. Urease activities of the BSMRAU soil varied widely from 1.8 to 1 14.0 mg NHþ dry soil h1, with a particularly lower activity for 4 -N g the green manure treatment. Activities of arylamidase differed by an order of magnitude with a significantly lower activity in compost treated soil compared to the control and cow dung treatments and intermediate values for the rice straw and green manure treatments (Table 2). Activities of all studied enzymes, except urease, increased significantly with application of mineral N fertilizer (b-glucosaminidase (P<0.05), b-glucosidase (P<0.01), Lglutaminase (P<0.01) and arylamidase (P<0.01)). Activities of urease and arylamidase were lowered by green manure and rice straw amendment in 220 kg N ha1 treated plots (Fig. 1b). Similarly, in the BAU field experiment, fertilizer application significantly altered b-glucosaminidase (P<0.05), L-glutaminase (P<0.05) and urease activity (P<0.05), but unlike at BSMRAU, not always positively affected (Table 3). The order of variation among BAU treatments was similar to that of BSMRAU. For instance bglucosaminidase activity ranged between 42.03 and 62.08 mg PNP g1 dry soil h1, in line with previous reports of 20e80 mg PNP g1 dry soil h1 by Parham & Deng [14] and Ekenler & Tabatabai [17] for North American soils, however, with contrasting cropping pattern. A significantly higher b-glucosaminidase activity was measured in the NþFYM treated plots followed by the NPK and NP treatments, and then N and control treatments (Table 3). Unlike at BSMRAU, bglucosidase activity did not significantly differ among the BAU treatments, and closely matched observations in paddy soils of 14e32 mg PNP g1 dry soil h1 by Das et al. [18]. L-glutaminase activity was significantly higher in NP treatment at the BAU site, when compared to the unfertilized control, NPK and NþFYM treatments (Table 3).
3.3. Correlations between enzyme activities and soil properties In the BSMRAU experiment, only L-glutaminase correlated positively with soil N and SOC (P<0.01), and negatively with pHKCl (P<0.05). The other enzyme activities did not correlate significantly to any of the soil properties except for a negative correlation between arylamidase and pHKCl (P<0.05). However, in the BAU experiment, b-glucosaminidase activity positively correlated with soil N and SOC (P<0.01) and b-glucosidase activity with SOC (P<0.01) and the soil C:N ratio (P<0.05). In contrast to the BSMRAU experiment, L-glutaminase activity was not correlated with any of the investigated soil properties. Arylamidase activity again showed a negative correlation with pHKCl (P<0.05).
30
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
Table 1 N mineralization rates, total N, total SOC, C:N ratios and pHKCl(±standard error) of the two experimental sites (BSMRAU and BAU) (adapted from Kader [3]). Treatments
Total N (g kg1)
SOC (g kg1)
C:N ()
pHKCl ()
N mineralization rate (mg N kg1 day1) Aerobic
Anaerobic
BSMRAU Nitrogen 0 kg N ha1 220 kg N ha1
b
0.52 ± 0.06a 0.62 ± 0.09b
7.30 ± 0.69a 8.54 ± 1.00b
14.2 ± 0.7 13.9 ± 0.7
4.7 ± 0.1b 4.5 ± 0.1a
0.31 ± 0.05a 0.71 ± 0.25b
0.84 ± 0.18a 1.22 ± 0.28b
ANOVA
**a
*
NS
**
*
*
Organic matter Control Compost Cowdung Green manure Rice straw
0.55 0.62 0.55 0.57 0.52
ANOVA
NS
Nitrogen x OM 0 kg N ha1 Control Compost Cowdung Green manure Rice straw 220 kg N ha1 Control Compost Cowdung Green manure Rice straw
7.73 8.40 7.62 8.24 7.54
± ± ± ± ±
0.22 0.45 0.13 0.69 0.23
NS
14.1 13.6 13.9 14.5 14.5
± ± ± ± ±
1.2 0.8 0.2 0.8 0.5
NS
4.6 4.7 4.7 4.5 4.6
± ± ± ± ±
0.0 0.3 0.3 0.1 0.2
NS
0.29 0.64 0.54 0.52 0.59
± ± ± ± ±
0.00 0.25 0.26 0.26 0.24
NS
0.82 1.12 1.21 1.19 0.80
± ± ± ± ±
0.14 0.07 0.51 0.35 0.28
NS
± ± ± ± ±
0.11 0.03 0.01 0.01 0.16
7.57 7.72 7.37 7.76 6.11
± ± ± ± ±
0.74 0.08 0.57 0.26 0.87
13.3 13.8 14.2 15.1 14.9
± ± ± ± ±
0.7 0.5 0.3 0.4 1.9
4.6 4.9 4.9 4.6 4.8
± ± ± ± ±
0.0 0.1 0.1 0.0 0.1
0.29 0.39 0.28 0.33 0.28
± ± ± ± ±
0.04 0.16 0.13 0.02 0.09
0.76 1.10 0.90 1.00 0.64
± ± ± ± ±
0.04 0.13 0.19 0.27 0.03
0.54 0.67 0.73 0.64 0.57
± ± ± ± ±
0.06 0.04 0.05 0.12 0.08
7.53 8.73 9.43 8.98 8.04
± ± ± ± ±
0.82 0.61 0.32 0.81 0.33
14.1 13.4 13.0 14.1 14.2
± ± ± ± ±
0.1 0.9 0.2 0.7 0.7
4.6 4.4 4.5 4.4 4.4
± ± ± ± ±
0.0 0.0 0.0 0.0 0.1
0.29 0.89 0.79 0.70 0.90
± ± ± ± ±
0.03 0.06 0.12 0.04 0.17
1.01 1.22 1.65 1.50 1.03
± ± ± ± ±
0.03 0.17 0.04 0.27 0.16
± ± ± ± ±
0.48 0.83 0.04 0.68 0.44
NS
BAU Control N NP NPK NþFYM
1.60 1.64 1.75 1.78 1.77
ANOVA
*
b
0.03 0.14 0.02 0.08 0.16
0.57 0.56 0.52 0.52 0.41
ANOVA
a
± ± ± ± ±
NS ± ± ± ± ±
0.1a 0.0ab 0.0bc 0.1c 0.0c
15.1 15.3 15.8 17.0 16.6
NS ± ± ± ± ±
0.1a 0.2ab 0.9abc 1.2c 0.7bc
*
9.4 9.3 9.0 9.6 9.4 NS
* ± ± ± ± ±
0.6 0.1 0.3 0.2 0.4
5.7 5.9 5.7 5.7 5.7 **
NS ± ± ± ± ±
0.0a 0.0b 0.1a 0.0a 0.1a
1.97 1.54 2.24 1.22 1.23
NS ± ± ± ± ±
0.34b 0.12a 0.13b 0.04a 0.20a
**
1.54 1.98 1.92 1.51 0.98 NS
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
3.4. Correlations between N mineralization and soil enzyme activities Pearson correlation coefficients were calculated between the N mineralization rates (aerobic and anaerobic) and the enzyme activities for both experimental sites, separately (Table 5). In the BSMRAU experiment, only L-glutaminase activity correlated positively with the aerobic (P<0.05) and anaerobic (P<0.01) N mineralization rates. Likewise at the BAU experimental site, Lglutaminase activity correlated positively with both aerobic (P<0.05) and anaerobic (P<0.01) N mineralization rates. In addition, b-glucosaminidase activity was negatively correlated with both aerobic and anaerobic N mineralization rates (P<0.05). Lastly, bglucosidase and urease activities correlated negatively with aerobic (P<0.05) and anaerobic N mineralization rate (P<0.05).
with positive correlations among these variables (P < 0.05). However, across the BAU experimental treatments there was an unexpected negative correlation between SOC content and aerobic N mineralization rate (when expressed relatively to soil N content) reported by Kader [3]. There was no correlation with soil N level, in line with previously reported poor or unexpected dependency of anaerobic N mineralization and soil properties of Bangladeshi floodplain soils [4,5]. The BSMRAU terrace soil is inundated only during the rainy season for cultivation of rainfed Aman rice (one rice crop per year; 5 months) while the BAU floodplain soil is inundated for a longer period during the cultivation of irrigated Boro rice in winter season in addition to rainfed Aman rice in rainy season (two rice crops per year; 9 months). A longer inundation period at BAU site could have led to accumulation of organic N as previously reported [19,20], of which, an apparent substantial part was not readily mineralizable.
4. Discussion 4.2. Effect of OM and fertilizer application on soil enzyme activities 4.1. Effect of management on N mineralization Soil management affected N mineralization with a significant promotion of aerobic and anaerobic N mineralization with N fertilizer addition at the BSMRAU site while at the BAU site this was not the case. Soil management directly induced build-up of soil C and N at BSMRAU also led to elevate soil N mineralization rates
At both sites, the studied soil enzyme activities were significantly affected by the management practices used. Changes in soil enzyme activities were expressed in relative to the unfertilized controls to compare the responses of the different enzyme activities (Figs. 1 and 2). As stated by Ekenler & Tabatabai [17] and Dick [21], differences in the activity of enzymes result from differences
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
31
Table 2 Soil enzyme activities (±standard error) in incubated soil from the BSMRAU field experiment.
b-glucosaminidase (mg PNP g1 dry soil h1)
b-glucosidase (mg PNP g1 dry
L-glutaminase (mg N g1dry soilh1)
Urease (mg NH4-Ng1 dry soil h1)
Arylamidase (mg b-napthylamine g1 dry soil h1)
Nitrogen 0 kg N ha1 220 kg N ha1
b 43.0 ± 4.7a 45.5 ± 1.2b
21.6 ± 1.5a 27.5 ± 2.1b
1.7 ± 0.4a 4.4 ± 0.5b
8.1 ± 1.1 8.8 ± 1.7
5.4 ± 0.9a 9.5 ± 1.5b
ANOVA
*a
**
**
NS
**
Organic matter Control Compost Cowdung Green manure Rice straw
39.2 44.1 61.1 40.3 36.5
11.4 ± 2.2b 10.5 ± 1.7b 9.2 ± 1.8b 2.2 ± 0.4a 8.8 ± 2.1b
8.8 ± 2.8b 3.6 ± 1.5b 10.5 ± 2.9a 7.3 ± 1.6ab 6.9 ± 0.9ab
ANOVA
**
**
*
Treatments
± ± ± ± ±
2.9ab 1.0c 5.1d 2.5b 3.3a
soil h1)
18.8 23.0 24.3 28.9 27.8
± ± ± ± ±
1.8a 3.5b 2.7b 3.6c 2.8c
**
1.6 3.8 3.4 4.3 2.2
± ± ± ± ±
0.7a 0.3b 0.9b 0.2b 0.4a
**
Nitrogen x OM 0 kg N ha1 Control Compost Cowdung Green manure Rice straw 220 kg N ha1 Control Compost Cowdung Green manure Rice straw
34.4 43.7 69.8 36.0 30.9
± ± ± ± ±
1.8c 1.9b 1.9a 1.1bc 0.3c
16.4 17.4 29.1 23.0 22.9
± ± ± ± ±
1.8b 1.3b 1.4a 0.4ab 0.0ab
0.4 4.0 1.4 2.1 0.8
± ± ± ± ±
0.2c 0.1a 0.4ab 0.2b 0.2ab
8.8 ± 1.8ab 9.6 ± 0.8ab 7.0 ± 1.8ab 2.6 ± 0.8b 12.3 ± 0.0a
4.8 1.6 5.6 9.3 5.6
44.0 44.4 52.5 44.6 42.2
± ± ± ± ±
0.7b 1.5b 1.3a 0.8b 0.3b
21.3 28.8 20.0 35.7 32.8
± ± ± ± ±
1.8b 2.7ab 1.4b 0.9a 0.1a
2.8 3.6 5.5 6.6 3.7
± ± ± ± ±
0.1b 0.1b 0.6ab 0.9a 0.0b
14.0 ± 1.8a 11.4 ± 2.6a 11.4 ± 2.6a 1.8 ± 0.0a 5.3 ± 1.8a
12.8 ± 3.9a 5.6 ± 2.2a 15.5 ± 1.3a 5.4 ± 1.4a 8.2 ± 0.7a
ANOVA
**
NS
*
a b
**
**
± ± ± ± ±
0.9a 0.3a 1.3a 2.5a 0.9a
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
in quantity and quality of crop residues and resulting soil microbial activity. The relationship is complex, however, because enzyme activities represent the cumulative effects of substrate availability, demand for the product, and connected environmental conditions and resulting microbial activity [22]. Utilizing this framework, the effects of fertilizer and OM application on enzyme activities were interpreted. 4.2.1. Substrate availability It is not surprising that repeated OM application at the BSMRAU site increased activity of b-glucosidase (Table 2) as it catalyses biodegradation of b-glucosides in plant debris [23,24]. Amendment with rice straw and green manure, i.e. carbohydrate rich substrates, relatively promoted b-glucosidase activities compared to application of compost and cow dung, likely because these already have gone through intensive microbial transformation. In line, compost and cow dung amendments promoted b-glucosaminidase activity, which may be attributed to their higher content of animal glucosamine polymers and microbial cell wall N (peptidoglycan, chitin), targeted by b-glucosaminidase [25]. With multiple sources of C and N in the soil, Geisseler & Horwath [26] found that enzymes participating in protein degradation would also be controlled primarily by the concentration of nitrogen substrate. The higher Lglutaminase activity in compost, cow dung and green manure amended BSMRAU plots compared to rice straw and unamended plots may stem from differences in proteinaceous N availability. In the BSMRAU site, application of 220 kg N ha1 furthermore enhanced all studied enzyme activities, alongside crop growth, SOC and soil N. Most likely, N-application increased OM inputs from residual roots, stubble and root exudates, and thereby indirectly increased SOC and N levels (Table 1) as well as energy sources for microbial activity. While aminohydrolases show rapid responses to changing substrate levels [27], trends in urease and arylamidase
activities over the OM treatments did not follow the expected pattern of enhanced substrate availability (Fig. 1b). In fact, urease activity was lowered by OM application in the plots applied with 220 kg N ha1 as urea (Fig. 1b). This contradictory result may be explained by the much higher levels of NHþ 4 (Table 1), which suppresses urease activity [28]. In general, the BSMRAU experiment demonstrated that repeated (22 years) OM amendment without N fertilizer application promotes soil catalytic activity. In the BAU experiment, however, the variation of the enzyme activities over the OM treatments was less consistent. The NP, NPK and NþFYM application significantly increased soil N and SOC (P < 0.05) and positively correlated with b-glucosaminidase activity (Table 4), which is in agreement with the reasoning that substrate availability would dominantly control enzyme production. However, NPK and NþFYM application strongly lowered arylamidase and L-glutaminase activity relative to the unfertilized control, in spite of the elevated SOC and soil N levels (Fig. 2). Also, b-glucosidase activity did not respond to fertilizer application at the BAU site, which is in clear contrast to the BSMRAU experiment. These contrasting trends in soil N, SOC and studied N mineralizing enzymes (urease, arylamidase and L-glutaminase) demonstrate that a promotion of enzyme activity with increasing SOM levels is enzyme specific at the BAU site. In paddy soils with prolonged inundation N can be stabilized through binding with aromatic rings of anaerobically accumulated lignin and phenolic subunits [19,20,29] and this may be the case in the NPK and NþFYM treatments with elevated SOC. In summary, it is clear that environmental conditions and microbial nutrient demand should be used as additional frameworks to explain these observations. 4.2.2. Microbial nutrient demand and environmental conditions As extracellular enzyme activity mediates microbial nutrient acquisition from organic matter, these activities are alternatively,
32
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
Fig. 2. Relative changes of b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase enzyme activities in different treatments over control in a young floodplain soil of field experiment (BAU). Positive and negative values indicate higher and lower enzyme activities compared to the unfertilized control: e.g. þ47% for urease in case of the N treatment indicates a nearly half as high activity with mineral N applied vs. the unamended plots. Treatments indicated by different letters have statistically different (P < 0.05) means per enzyme activity according to Duncan's multiple range post-hoc test.
often interpreted as indicators of microbial nutrient demand. At the BSMRAU site, annual application of 220 kg N ha1 would then be expected to lower the microbial demand for N. However, activities of L-glutaminase, arylamidase and b-glucosaminidase were all significantly higher than in the 0 kg N ha1 counterpart soil plots. Hence, the demand for product, in this case mineral N, would consequently seem to hold lesser control on levels of these enzymes in the BSMRAU plots, corroborating the previously stated reasoning that instead substrate availability would. However, 220 kg N ha1 application did interact with the response of enzyme activities to OM-amendment (Table 2), rendering L-glutaminase and b-glucosaminidase activities less responsive to OM application (Fig. 1). Cow dung, green manure and rice straw application resulted in 40e80% lower arylamidase activity compared to the unamended control, unlike in the 0 kg N ha1 treated soils (Fig. 1). Possibly, the combined N-supply from mineral N and OM application rendered N non-limiting for microbes, relieving the need to produce hydrolytic enzymes. Hence, the established relationship where enzyme activity is dictated by N-demand by the microbial
Fig. 1. Relative changes of b-glucosaminidase, b-glucosidase, L-glutaminase, urease and arylamidase enzyme activities in different exogenous OM treatments over the control treatment (a) without N (0 kg N ha1) and (b) with 220 kg N ha1 application in a highly weathered terrace soil of field experiment (BSMRAU). Positive and negative values indicate higher and lower enzyme activities compared to the unamended control at 0 kg N ha1 (a) and 220 kg N ha1 levels (b): e.g. þ102% for b-glucosaminidase in case of the compost treatment at 0 kg N ha1 indicates an approximately twice as high activity than in the unamended plots at 0 kg N ha1. Note the differing right axis for L-glutaminase activity in Fig. 1(a). Treatments indicated by different letters have statistically different (P < 0.05) means per enzyme activity according to Duncan's multiple range post-hoc test.
Table 3 Soil enzyme activities (±standard error) in incubated soil from the BAU field experiment. Treatments
b-glucosaminidase (mg PNP g1 dry soil h1)
b-glucosidase (mg PNP g1 dry soil h1)
Control N NP NPK NþFYM
b
42.0 ± 0.7a 44.9 ± 2.0a 50.4 ± 1.4b 55.4 ± 1.3c 62.1 ± 2.4d
17.3 16.8 15.0 21.4 17.6
ANOVA
*a
NS
a b
± ± ± ± ±
0.3 1.1 0.4 2.3 1.5
L-glutaminase (mg N g1 soilh1)
5.2 6.7 8.3 3.7 2.3 *
± ± ± ± ±
0.3bc 0.9cd 0.6d 0.6ab 0.7a
Urease (mg NH4-Ng1 dry soil h1) 3.5 5.3 2.6 2.9 8.2 *
± ± ± ± ±
0.0a 1.0ab 0.5a 1.2a 1.5b
Arylamidase (mg b-napthylamine g1 dry soil h1) 15.6 ± 5.1 16.1 ± 3.2 16.0 ± 3.1 5.6 ± 1.8 6.5 ± 1.6 NS
**, * denote statistical significance at P 0.01 and P 0.05, respectively; NS, not significant. Treatments indicated by different letters have statistically different (P < 0.05) means according to Duncan's multiple range post-hoc test.
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
33
Table 4 Pearson's correlation coefficients between soil enzyme activities and general soil properties. BSMRAU (n ¼ 20)
Enzymes
b-glucosaminidase b-glucosidase L-glutaminase Urease Arylamidase a
BAU (n ¼ 15)
Soil N
SOC
C:N
pH
Soil N
SOC
C:N
pH
0.13 0.13 0.69**a 0.02 0.24
0.09 0.12 0.69** 0.18 0.27
0.20 0.11 0.06 0.14 0.13
0.11 0.30 0.45* 0.20 0.49*
0.72** 0.38 0.19 0.02 0.24
0.73** 0.70** 0.32 0.18 0.27
0.11 0.53* 0.23 0.14 0.13
0.22 0.11 0.01 0.20 0.49*
* Correlation is significant at the 0.05 level, ** Correlation is significant at the 0.01 level.
Table 5 Pearson's correlation coefficient between N mineralization rate (aerobic and anaerobic) and enzyme activities. Parameters
b-glucosaminidase b-glucosidase L-glutaminase
Urease Arylamidase
N Mineralization rate (mg N kg1 day1) BSMRAU (n ¼ 20)
BAU (n ¼ 15)
Aerobic
Anaerobic
Aerobic
Anaerobic
0.10 0.41 0.49*a 0.05 0.16
0.28 0.19 0.75** 0.11 0.37
0.59* 0.52* 0.77** 0.45 0.50
0.55* 0.06 0.54* 0.56* 0.33
a *Correlation is significant at the 0.05 level, ** Correlation is significant at the 0.01 level.
biomass, in turn determined by the elemental stoichiometry of microbial biomass relative to environmental nutrient availability [10], appears only partly relevant for the BSMRAU site. In case of the BAU experiment, only arylamidase and L-glutaminase responded similarly to the fertilizer applications, with a significantly lower activity in the NPK and NþFYM treatments (Fig. 2). Both NPK and NþFYM treatments differ from the other plots by the application of K, a frequently limiting nutrient for plant growth in double rice cropping with repeated large exports of K in rice straw. Indeed, the BAU site yield data shows (see supplementary material) that above-ground biomass and crop yields were significantly elevated in NPK and NþFYM treatments compared to non-K treated plots. This higher plant productivity resulted in build-up of soil N and SOC but anaerobic N mineralization was comparatively lower. Nourbakhsh & Alinejadian [30] state that anaerobic conditions induced a limiting situation for microbial activities and therefore, NH3 release rate during anaerobic incubation was greatly controlled by the microbial activities rather than bioavailability of N. There was no direct quantification of microbial activity in the present experiments but the b-glucosidase activity as a proxy for Cdemand could be considered. It is then of further interest to compare enzymes involved in N-mineralization vis- a-vis to bglucosidase activity. At BAU site, the latter was unresponsive to the fertilizer treatments, and so microbial activity and consequently microbial nutrient demand, are not suggested to differ strongly, compared to L-glutaminase and arylamidase activity. Therefore, in contrast to the BSMRAU site, it cannot be concluded that microbial activity primarily determines production of extra-cellular enzymes catalyzing N-mineralization at BAU site. Hence, it appears that at the BAU site, these activities were disconnected from microbial demand, but depend on substrate quality and content. Differences in SOM quality between the BAU fertilizer treatments might thus explain this disconnection between soil N mineralization and enzymatic activity and thus soil N level. Schmidt-Rohr et al. [29] found that significant amounts of the amide N were directly bound to aromatic rings in paddy soils, leading them to conclude that N bound to lignin accounts for its reduced bioavailability. Such
a specific reduction of amide N-availability with SOM accumulation in the NPK andNPKþFYM treatments might explain the lower arylamidase and L-glutaminase activities. b-glucosaminidase on the contrary did follow soil N content, which was expected as no specific stabilization of amino-sugars to phenolics is known. Alternatively, SOM content in organic amended soils can increase the capacity of the soils to protect extra-cellular pools of soil enzymes against proteolytic activities and hence, provide more functional enzyme molecules in soil [31]. If so, it could be hypothesized that less arylamidases and L-glutaminases were simply required with higher SOC content, however, the present experimental setup does not allow verification of this mechanism. 4.3. Enzymatically catalyzed SOM hydrolysis as bottleneck for anaerobic soil N mineralization A negative correlation between b-glucosidase activity and aerobic N mineralization rate (P<0.05) for the BAU site is in agreement with previously reported findings [6,32e34]. Higher b-glucosidase activities indirectly represent microbial growth, which temporarily immobilizes N from the soil as the C:N ratio of bacteria and fungi is smaller than the C:N ratio of SOM. Nitrogen acquisition from the aerobic system at BAU site followed the elemental stoichiometric theory [10] as observed by the positive correlation (r¼0.82) between the ratio of L-glutaminase and b-glucosidase enzymes as well as by aerobic N mineralization (Table 5). The anaerobic N mineralization rate, instead, was significantly correlated to L-glutaminase activity in both field experiments (r¼0.75, P < 0.01 and r¼0.54, P < 0.05 for BSMRAU and BAU, respectively), which is align with findings of Nourbakhsh & Alinejadian [30]. L-glutaminase is involved in the final step of N-mineralization where a terminal amino group of a peptide or amino acid is converted into NH3. Arylamidase, alternatively catalyses the hydrolysis of various amino acids from arylamides but does not directly induce the release of NH3 and was not significantly correlated to soil N mineralization. This suggests that at both BSMRAU and BAU sites, terminal amino acid hydrolysis may have formed a bottleneck in anaerobic N mineralization, as represented by soil NHþ 4 -N build-up. Earlier steps in soil N mineralization seem less rate limiting: b-glucosaminidase activity was involved in degradation of chitin and peptidoglucan, i.e. polymers in microbial cell walls, and even correlated negatively (Table 5) to aerobic and anaerobic N mineralization rate at BAU site. 5. Conclusion The dependency of soil N supply on SOM quality and SOM quantity and general soil properties of Bangladeshi floodplain paddy soils is complex [4,5]. This was also demonstrated by earlier study of two field experiments with increasing N mineralization at BSMRAU with soil N and an opposite trend at BAU site, suggesting instead, stabilization of N with OM build-up. An overarching positive relationship between anaerobic N mineralization and an
34
M.A. Kader et al. / European Journal of Soil Biology 80 (2017) 27e34
amidohydrolase enzyme activity (here L-glutaminase) may suggest that terminal amino acid cleavage and not preceding depolymerisation steps control anaerobic N mineralization in these paddy soils. A link between an amidohydrolase directly involved in the release of proteinaceous N, and anaerobic N mineralization did appear logical, and however, it is not clear which are the dominant controlling factors. Limited bio-availability of N, through association with accumulating phenols with prolonged annual inundation may be one of the constraints. Further research combining probing of SOM biochemistry, particularly by quantifying contents of polyphenols and bound-N, alongside amidohydrolase activity assays will be needed to confirm this hypothesized control of Nbioavailability on N-mineralization in young paddy soils found throughout Southeast Asia. References [1] D.R. Keeney, Nitrogendavailability indices, in: A.L. Page, R.H. Miller, D.R. Keeney (Eds.), Methods of Soil Analysis. Part 2, second ed., Agronomy, 9, SSSA and ASA, Madison, WI, 1982, pp. 711e733. Chemical and Microbiological Properties. [2] G. Stanford, Assessment of soil nitrogen availability, in: F.J. Stevenson (Ed.), Nitrogen in Agricultural Soils, Agronomy Monograph, 22, ASA, Madison, WI, 1982, pp. 651e688. [3] M.A. Kader, Nitrogen Mineralization in Subtropical Paddy Soils in Relation to Soil Properties, Organic Matter Fractions, and Fertilizer Management, Ph.D. thesis, Faculty of Bioscience Engineering, Ghent University, Ghent, Belgium, 2012, 213 pp.. [4] M.A. Kader, S. Sleutel, S.A. Begum, A.Z.M. Moslehuddin, S. De Neve, Nitrogen mineralization in sub-tropical paddy soils in relation to soil mineralogy, management, pH, carbon, nitrogen and iron contents, Eur. J. Soil Sci. 64 (2013) 47e57. [5] S. Sleutel, M.A. Kader, K. Demeestere, C. Walgraeve, J. Dewulf, S. De Neve, Subcritical water extraction to isolate kinetically different soil nitrogen fractions, Biogeosciences 10 (2013) 7435e7447. [6] M.A. Tabatabai, M. Ekenler, Z.N. Senwo, Significance of enzyme activities in soil nitrogen mineralization, Commun. Soil Sci. Plant Anal. 41 (2010) 595e605. [7] N. Khorsandi, F. Nourbakhsh, Prediction of potentially mineralizable N from amidohydrolase activities in a manure-applied, corn residue-amended soil, Eur. J. Soil Biol. 44 (2008) 341e346. [8] V. Acosta-Martínez, M.A. Tabatabai, Arylamidase activity of soils, Soil Sci. Soc. Am. J. 64 (2000) 215e221. [9] S. Muruganandam, D.W. Israel, W.P. Robarge, Activities of nitrogenmineralization enzymes associated with soil aggregate size fractions of three tillage systems, Soil Sci. Soc. Am. J. 73 (2009) 751e759. [10] R.L. Sinsabaugh, B.H. Hill, J.J. Follstad Shah, Ecoenzymatic stoichiometry of microbial organic nutrient acquisition in soil and sediment, Nature 462 (2009) 795e799. [11] M.A. Tabatabai, J.M. Bremner, Assay of urease activity in soils, Soil Biol. Biochem. 4 (1972) 479e487. [12] Soil Survey Staff, Soil Taxonomy: a Basic System of Soil Classification for Making and Interpreting Soil Surveys, second ed., Natural Resources Conservation Service. U.S. Department of Agriculture Handbook 436, 1999. [13] A.J.M.S. Karim, K. Egashira, Y. Yamada, J. Haider, K. Nahar, Long-term application of organic residues to improve soil properties and to increase crop yield
in terrace soil of Bangladesh, J. Fac. Agric. Kyushu U 39 (1995) 149e165. [14] J.A. Parham, S.P. Deng, Detection, quantification and characterization of bglucosaminidase activity in soil, Soil Biol. Biochem. 32 (2000) 1183e1190. [15] F. Eivazi, M.A. Tabatabai, Glucosidases and galactosidases in soils, Soil Biol. Biochem. 20 (1988) 601e606. [16] W.T. Frankenberger, M.A. Tabatabai, L-Glutaminase activity of soils, Soil Biol. Biochem. 23 (1991) 869e874. [17] M. Ekenler, M.A. Tabatabai, b-Glucosaminidase activity of soils: effect of cropping systems and its relationship to nitrogen mineralization, Biol. Fertil. Soils 36 (2002) 367e376. [18] S.K. Das, A. Varma, Role of enzymes in maintaining soil health, in: G. Shukla, A. Varma (Eds.), Soil Enzymology, Springer, Berlin, 2011, pp. 293e312. [19] P.J. Colberg, Anaerobic microbial degradation of cellulose, lignin, oligolinols, and monaromatic lignin derivatives, in: A.J.B. Zender (Ed.), Biology of Anaerobic Microorganisms, Wiley, New York, 1988, pp. 333e372. [20] D.C. Olk, K.G. Cassman, E.W. Randall, P. Kinchesh, L.J. Sanger, J.M. Anderson, Changes in chemical properties of organic matter with intensified rice cropping in tropical rice soil, Eur. J. Soil Sci. 47 (1996) 293e303. [21] R.P. Dick, Soil enzyme activities as indicators of soil quality, in: J.W. Doran, D.C. Coleman, D.F. Bezdicek, B.A. Stewart (Eds.), Defining Soil Quality for a Sustainable Environment, Soil Science Society of America Journal, Madison, WI, 1994, pp. 108e123. [22] S.D. Allison, M.N. Weintraub, T.B. Gartner, M.P. Waldrop, Evolutionary-economic principles as regulators of soil enzyme production and ecosystem function, in: G. Shukla, A. Varma (Eds.), Soil Enzymology, Springer, Berlin, 2011, pp. 229e244. [23] H.A. Ajwa, M.A. Tabatabai, Decomposition of different organic materials in soils, Biol. Fert. Soils 18 (1994) 175e182. [24] C.E. Martinez, M.A. Tabatabai, Decomposition of biotechnology byproducts in soils, J. Environ. Qual. 26 (1997) 625e632. [25] R.G. Burns, J.L. DeForest, J. Marxsen, R.L. Sinsabaugh, M.E. Stromberger, M.D. Wallenstein, M.N. Weintraub, A. Zoppini, Soil enzymes in a changing environment: current knowledge and future directions, Soil Biol. Biochem. 58 (2013) 216e234. [26] D. Geisseler, W.R. Horwath, Relationship between carbon and nitrogen availability and extracellular enzyme activities in soil, Pedobiologia 53 (2009) 87e98. [27] S.P. Deng, M.A. Tabatabai, Effect of tillage and residue management on enzyme activities in soils. I. Amidohydrolases, Biol. Fert. Soils 22 (1996) 202e207. [28] D. Pandey, M. Agrawal, J.S. Bohra, Effects of conventional tillage and no tillage permutations on extracellular enzyme activities and microbial biomass under rice cultivation, Soil till. Res. 136 (2014) 51e60. [29] K. Schmidt-Rohr, J.D. Mao, D.C. Olk, Nitrogen-bonded aromatics in soil organic matter and their implications for a yield decline in intensive rice cropping, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 6351e6354. [30] F. Nourbakhsh, A. Alinejadian, Arginine ammonification and L-glutaminase assays as rapid indices of corn nitrogen availability, J. Plant Nutr. Soil Sci. 172 (2009) 127e133. [31] P. Nannipieri, P. Sequi, P. Fusi, Humus and enzyme activity, in: A. Piccolo (Ed.), Humic Substances in Terrestrial Ecosystems, Elsevier, Amsterdam, 1996, pp. 293e327. [32] D.L. Burton, W.B. McGill, Spatial and temporal fluctuation in biomass, nitrogen mineralization reactions, and mineral nitrogen in a soil cropped to barley, Can. J. Soil Sci. 72 (1992) 31e42. [33] D.E. Dodor, M.A. Tabatabai, Amidohydrolases in soils as affected by cropping systems, Appl. Soil Ecol. 24 (2003) 73e90. [34] N. Khorsandi, F. Nourbakhsh, Effect of amendment of manure and corn residues on soil N mineralization and enzyme activity, Agron. Sustain. Dev. 27 (2007) 139e143.