Empirical model for mineralisation of manure nitrogen in soil.
Animal manures contain organic N that is released slowly and causes residual N effects in the years following application (Schroder et al. 2007). Accounting for the release dynamics of organic manure N is important for determining appropriate management measures. Due to the simultaneous mineralisation from soil organic N and inorganic N losses through gaseous emissions and leaching from soil, it is difficult to measure the mineralisation of organic manure N directly over longer periods. A model of the mineralisation of organic manure N can help better distinguish the contribution of manures from the total mineralisation in soil.
During the first weeks after pig and cattle slurry application, significant microbial immobilisation of inorganic N takes place due to decomposition and microbial assimilation of the organic carbon present in the manure (Kirchmann and Lundvall 1993; Sorcnscn and Amato 2002). Simultaneously, organic N in the manure is mineralised. Initially, immobilisation exceeds mineralisation and it takes approximately 3 months before the cumulated net mineralisation becomes positive. After this period there is a net release of N from the original organic compounds in the manure together with part of the initially immobilised N that is also released from decomposing microbial residues (Sorensen and Amato 2002).
Residual effects of a single manure application in the years after application are usually small and can be hard to detect (Bhogal etal. 2016). Therefore, [sup.15]N-labelled manures have been used to detect residual effects in the years after application (Jensen et al. 1999; Webb et al. 2013). Farmers often apply manure to the same field every year and, as a result of these repeated manure applications, a significant accumulated residual effect can be expected (Schroder et al. 2007).
Currently, a new nitrogen leaching model (NLES5) is being developed in Denmark, including a submodel of N leaching from animal manure inputs in the years after application. The new model is supposed to replace the existing NLES4 model (Kristensen et al. 2008). In the development of the new leaching model, we found that it would be necessary to include a robust calculation of the yearly N mineralisation rates from manures.
This paper presents a general and easy-to-use manure N mineralisation model describing the yearly net mineralisation during the first 5 years after manure application. The model is based on a 3-year field study of residual effects of cattle and pig slurry application on crop N uptake measured under North European conditions in a cool and moist climate, and the model outputs are compared with data from other experimental studies of manure N mineralisation and residual N effects under similar climatic conditions.
Materials and methods
A simple empirical model was developed based on data from a Danish field experiment reported previously by Sorcnscn and Amato (2002) and Sorenscn (2004). Framed plots were supplied with pig or cattle slurry or mineral fertiliser, with all plots receiving the same amount of inorganic N (100-103 kg N [ha.sup.-1] [year.sup.-1]), except in control plots without N application. The pig slurry was applied to loamy sand (9% clay) and sandy loam (18% clay) soils, whereas the cattle slurry was applied to loamy sand. In the experiment, three different slurry application methods were used: incorporation by mixing, simulated injection and surface banding. In the following, average data from the three application methods are used, because all three reflect common practices. The slurries included excreta and unused feed plus bedding. In the following years, all plots received mineral fertiliser at the same rate (120 kg N [ha.sup.-1] [year.sup.-1]), and spring barley and undersown ryegrass catch crops were established in all three years. The N uptake in barley and catch crops was measured and the additional N uptake was determined in manured plots and compared with plots receiving only mineral fertiliser in the first year. The slurry ammonium-N was labelled with [sup.15]N and a [sup.15]N balance showed similar recovery in soil and plants from slurry applied by incorporation or injection and labelled N from mineral N compared with the mineral fertiliser N treatment (Sorensen 2004). This indicates low ammonia emission after both incorporation and injection, whereas significant emission took place from the surface-banded slurry. Net N mineralisation due to manure was estimated using the assumption that the apparent N recovery (ANR) in barley of mineralised N was similar to the 60% ANR that was measured for mineral N fertiliser in the same experiment (Sorensen and Amato 2002). To calculate net mineralisation in autumn and winter after the spring barley harvest, we assumed that 49% of the mineralised N was recovered in the tops of the catch crop. This assumption is based on results obtained by Li et al. (2015), who recovered 49% of applied [sup.15]N-labelled nitrate (supplied as a split application over a period in autumn) in shoots of a ryegrass cover crop at the same location.
The mineralisation of manure N in Years 4 and 5 was based on the estimated mineralisation of residual manure N in Year 3. This was done by extrapolation using data from Hart et al. (1993), who measured wheat uptake of residual labelled N in soil for 5 years after labelling wheat plant residues with [sup.15]N in fertiliser in the first year. Jensen et al. (1999) and Sorensen and Jensen (1998) found that the N release rate from manure was similar to that from plant residues in the years after application. Therefore, we assumed that the interannual relative release of residual N from manure was similar to the findings of Hart et al. (1993) from plant residues, with residual N defined as the organic N remaining in soil at the start of a season or year. The crop uptake of residual [sup.15]N in soil reported by Hart et al. (1993) was 6.6%, 3.5%, 2.2% and 2.2% in the consecutive 2-5 years after application. The mineralisation of residual manure N in Years 4 and 5 in the present study was estimated accordingly by multiplying the mineralisation of residual N in Year 3 by the factor 2.2/3.5=0.63.
Manure N mineralisation model
The studies by Sorensen and Amato (2002) and Sorensen (2004) showed an initial net immobilisation phase and that the cumulated net mineralisation of slurry N was close to zero or slightly negative after the first 3 months, because crop N uptake in the first spring barley crop was similar or lower after slurry application compared with a mineral fertiliser reference receiving the same amount of inorganic N. Therefore, in the model we assumed an initial net N immobilisation phase within the first 14 days after application, as found in incubation studies. An initial net N immobilisation equivalent to 25% of the organic N in cattle slurry and 38% of organic N in pig slurry as measured by S0rensen and Amato (2002) and Sorensen (2004) was adopted in the model. The cumulated net mineralisation from slurry was assumed to be zero in the model after 3 months (Fig. 1) based on observations of an N uptake in the first barley crop when the slurry was injected being similar to an application of the same amount of mineral N in mineral fertiliser.
In the following ryegrass catch crop and in Years 2 and 3, N uptake was higher after manure application than in the mineral fertiliser reference (Table 1). The extra N uptake after manure application indicated that there was extra N mineralisation due to organic N in the manures and due to remineralisation of immobilised N.
Based on the additional N uptake (Table 1), net N mineralisation of manure N was estimated using the assumption that ANR in barley from mineralised N was similar to that from mineral fertiliser. The ANR of mineral fertiliser N was found to be 60% in the same crop (Table 2).
In Table 3, N mineralisation in the first 3 years is taken from Table 2, whereas mineralisation in the following 2 years was estimated by extrapolation based on data from Hart et al. (1993). This approach was supported by the fact that the modelled mineralisation in Year 3 could be well predicted from mineralisation in Year 2 using Hart et al. (1993). For instance, for cattle slurry, the mineralisation of residual organic N in the third year could be estimated as 20% x 3.5/6.6= 11%, whereas our model based on experimental data predicted 12% mineralisation of residual N in the third year (Table 3).
After 5 years, the model estimates cumulated net N mineralisation to be 71% of the organic N input in pig slurry and 51% of the organic N in cattle slurry (Fig. 1).
Soil texture effects on N mineralisation
van Faassen and van Dijk (1987) studied net N mineralisation from different manure types over 18 months after application to two contrasting soil types containing approximately 3% clay (sandy soil) and approximately 18% clay (sandy loam), van Faassen and van Dijk (1987) applied different faeces samples from cattle, pig and poultry. The overall N mineralisation was studied outdoor in unplantcd pots under a cool and moist climate in the Netherlands. The manures were applied in spring (April) and the experiment was repeated in the following year, van Faassen and van Dijk (1987) observed very low or even negative N mineralisation in the winter period, so the observed mineralisation during the first 0-6 months was nearly equal to the mineralisation in the first year, and mineralisation in the second period (6-18 months) was nearly equal to that in the second year. In the first year, the measured N mineralisation in the sandy loam soil was nearly the same to our model calculation (18% and 31% for cattle and pig slurry respectively vs 17% and 27% in our model), whereas it was higher in the sandy soil (Table 4). In the second year, the manure N mineralisation measured by van Faassen and van Dijk (1987) was significantly lower than that calculated by our model (8-14% and 11-17% for cattle and pig slurry respectively vs 17% and 27% in our model). Our model was based on soils with 9-18% clay. Thomsen and Olesen (2000) found no clear relationship between soil clay content (ranging from 11% to 45%) and manure N mineralisation in soil in an incubation study. In the experiment reported by Sarensen (2004), higher residual effects of cattle slurry on crop N uptake were not observed in a sandy soil with only 4% clay compared with loamy sand with 9% clay (P. Sorensen, unpubl. data). In a field study with [sup.15]N-labelled manures applied to two different soil types with 10% and 17% clay, Jensen et al. (1999) found no clear difference in plant uptake of mineralised labelled manure N in the second year, whereas in the third year they found an approximately 25% higher uptake from labelled manure N on the sandy soil. Because some studies have indicated that soil texture plays a role in net N mineralisation from manures in both the short term (Sorensen and Jensen 1995, 1998) and in the following years (van Faassen and van Dijk 1987; Jensen et al. 1999; Sorensen and Amato 2002), it would probably be relevant to refine the model further to distinguish between soil types, even though not all observed soil texture effects on N mineralisation are consistent.
Effects of manure application method
The distribution of slurry in soil affects N mineralisation and immobilisation, causing a higher net release of N when manure is not well mixed with soil, as with slurry injection (Sorensen and Jensen 1995). In the studies of Sorensen and Amato (2002) and S0rensen (2004), this was reflected in the higher N uptake in the first barley crop when using injection compared with manure incorporation, despite the negligible ammonia loss expected in both situations. In the present model we do not account for manure distribution in the soil and we have used average data across the three application methods (incorporation by mixing, injection and surface banding) in Table 1. All three manure application methods are used in Denmark.
Effects of manure composition on N mineralisation
The initial manure N mineralisation is affected by manure composition, which, in turn, is affected, among other things, by livestock diet (Sorensen and Fernandez 2003; Sorensen et al. 2003). Sorensen and Fernandez (2003) and Sorensen et al. (2003) found a significant negative relationship between the slurry C : N ratio and N availability (i.e. the sum of ammonium N and mineralised N in soil), but they could not identify any significant relationships between manure composition and net mineralisation of organic manure N. Sorensen et al. (2003) investigated N release from slurry obtained from cows fed differently and observed both positive and negative net N mineralisation from cattle slurries 3 months after application to soil. However, the average N release from the cattle slurries investigated was close to zero 3 months after application. Similarly, Sercnsen and Fernandez (2003) observed both positive and negative net mineralisation from different pig slurries. In most cases, they observed a positive net mineralisation after 3 months in soil. In addition, Kirchmann and Lundvall (1993) observed a positive net mineralisation from anaerobically stored pig slurry 10 weeks after application to soil. For the model presented herein we used data from the study of Sarensen and Amato (2002), who found a net mineralisation close to zero 3 months after pig slurry application. It should be noted that the temporal development in net mineralisation from slurry is uncertain, and that the time until net N mineralisation becomes positive is variable.
After the initial months of decomposition, most of the organic manure compounds have been used by soil micro-organisms and incorporated in microbes and microbial residues, and, at this stage, the release rate from residual N is only little affected by the composition of the original N sources (Jensen et al. 1999). The model we present herein is not able to take into account variable manure compositions and it is only intended to give a robust average estimate of manure N mineralisation. However, the model makes a distinction between pig and cattle slurry to take due account of the differences in recalcitrance of faeces of monogastric animals and ruminants (Chadwick et al. 2000). Chadwick et al. (2000) determined net N mineralisation from different manure types sampled at different farms from N uptake measured in grass grown in pots under controlled temperature conditions after manure application (Table 5). These authors found a quite high variation in net N mineralisation during the first period, even within the same manure types, and a negative relationship between C : organic N ratios in manures and the net release from organic N in the manures. Chadwick et al. (2000) also found that the mineralisation became low and nearly constant from all manure types after 4 months. In that study, the manures were stripped of ammonium-N before application, which reduced the ammonium content but may also have triggered a fast decomposition of decomposable compounds, like volatile fatty acids, before the manures were applied to the soil. Possibly, this may have reduced the initial N immobilisation in soil and thereby increased the net N mineralisation from the manures slightly.
In the study of Chadwick et al. (2000), there was no clear difference in average N mineralisation between slurries and solid manures from the same species (Table 5), and a similar model can therefore probably be used for both organic N in slurry and solid manure originating from the same species. The average mineralisation rate in the first year was nearly twice as high from organic N in pig manures as from cattle manures after 3184 degree-days of decomposition, which is equal to approximately 1 year under Danish conditions. The average mineralisation in the first year was close to the organic N mineralisation estimated in our model when also considering the high variation observed between different manure samples (Table 5).
Longer-term studies of manure N release
Schroder et al. (2007) estimated the accumulated N fertiliser replacement value from repeated injection of cattle slurries to grassland for periods of up to 4 years. By relating the N fertiliser value to the organic N applied with the manure and assuming that approximately 90% of the mineralised N was available for grass uptake, an approximate yearly mineralisation rate of organic N in cattle slurry of 21% was calculated for both the first and second years (Table 6). This mineralisation rate was higher than in our model (17% in both Years 1 and 2), but there was good agreement between measurements and our model estimates in Years 3 and 4 (8% and 6% respectively vs 8% and 5% in our model). The experiment reported by Schroder et al. (2007) was performed on a sandy soil and this may explain the higher N mineralisation rate in the first years compared with our model. In addition, we cannot exclude an effect of slightly higher average temperatures in the Netherlands causing faster mineralisation rates.
A new empirical model can be used to estimate the yearly mineralisation from organic N applied in pig and cattle slurries in arable soils under cool, moist climate conditions. The estimated net N mineralisation is consistent with results in several field studies with manure applied under moist and cool, temperate climate conditions in different countries. The model can be used to estimate N residual effects of manure both in terms of fertiliser value and as input in model calculations of long-term nitrate leaching loss of manure N.
The work was funded by the Gylle-IT project under the Danish research and development program Grant Udviklings- og Demonstrations-program (GUDP).
Bhogal A, Williams JR, Nicholson FA, Chadwick DR, Chambers KH, Chambers BJ (2016) Mineralization of organic nitrogen from farm manure applications. Soil Use and Management 32, 32 43. doi: 10.1111/sum. 12263
Chadwick DR. John F, Pain BF, Chambers BJ, Williams J (2000) Plant uptake of nitrogen from the organic nitrogen fraction of animal manures: a laboratory experiment. The Journal of Agricultural Science 134, 159-168. doi: 10.1017/S0021859699007510
Hart PBS, Powlson DS, Poulton PR, Johnston AE, Jenkinson DS (1993) The availability of the nitrogen in the crop residues of winter wheat to subsequent crops. The Journal of Agricultural Science 121, 355-362. doi: 10.1017/S0021859600085555
Jensen B, Sorensen P, Thomsen IK. Jensen ES, Christensen BT (1999) Availability of nitrogen in 15N-labeled ruminant manure components to successively grown crops. Soil Science Society of America Journal 63, 416-426. doi: 10.2136/sssaj 999.03615995006300020021x
Kirchmann H, Lundvall A (1993) Relationship between N immobilization and volatile fatty acids in soil after application of pig and cattle slurry. Biology and Fertility of Soils 15, 161-164. doi: 10.1007/BF00361605
Kristensen K, Waagepetersen J, Borgesen CD, Vinther FP, Grant R, Blicher-Mathiesen G (2008) Reestimation and further development in the N-LES, NLES3 to NLES4. DJF Plant Science no. 139, Aarhus University, Tjele, Denmark.
Li X, Sorensen P, Li FC, Petersen SO, Olesen JE (2015) Quantifying biological nitrogen fixation of different catch crops, and residual effects of roots and tops on nitrogen uptake in barley using in-situ [sup.15]N labelling. Plant and Soil 395, 273-287. doi: 10.1007/s11104-015-2548-8
Schroder JJ, Uenk D, Hilhorst GJ (2007) Long-term nitrogen fertilizer replacement value of cattle manures applied to cut grassland. Plant and Soil 299, 83-99. doi:10.1007/s11104-007-9365-7
Sorensen P (2004) Immobilisation, remineralisation and residual effects in subsequent crops of dairy cattle slurry nitrogen compared to mineral fertiliser nitrogen. Plant and Soil 267, 285-296. doi:10.1007/s11104005-0121-6
Sorensen P, Amato M (2002) Remineralisation and residual effects of N after application of pig slurry to soil. European Journal of Agronomy 16, 81 95. doi: 10.1016/S1161-0301(01)00119-8
Sorensen P, Fernandez JA (2003) Dietary effects on the composition of pig slurry and on the plant utilization of pig slurry nitrogen. The Journal of Agricultural Science 140. 343-355. doi: 10.1017/S0021859603003113
Sorensen P, Jensen ES (1995) Mineralization-immobilization and plant uptake of nitrogen as influenced by the spatial distribution of cattle slurry in soils of different texture. Plant and Soil 173, 283-291. doi:10.1007/BF00011466
Sorensen P, Jensen ES (1998) The use of [sup.15]N labelling to study the turnover and utilization of ruminant manure N. Biology and Fertility of Soils 28, 56-63. doi: 10.1007/s003740050463
Sorensen P, Weisbjerg MR, Lund P (2003) Dietary effects on the composition and plant utilization of nitrogen in dairy cattle manure. The Journal of Agricultural Science 141, 79-91. doi:10.1017/S002185 9603003368
Thomsen IK, Olesen JE (2000) C and N mineralization of composted and anaerobically stored ruminant manure in differently textured soils. The Journal of Agricultural Science 135, 151-159. doi: 10.1017/ S0021859699008096
van Faassen HG, van Dijk H (1987) Manure as a source of nitrogen and phosphorus in soils. In 'Animal manure on grassland and fodder crops. fertilizer or waste?'. (Eds HG Van Meer, RJ Unwin, TA Van Dijk, GC Eunik) pp. 27-45. (Martinus Nijhoff: The Hague, Netherlands)
Webb J, Sorensen P, Veithof G, Amon B, Pinto M, Rodhe L, Salomon E, Hutchings N, Burczyk P, Reid J (2013) An assessment of the variation of manure nitrogen efficiency throughout Europe and an appraisal of means to increase manure-N efficiency. Advances in Agronomy 119, 371 442. doi:10.1016/B978-0-12-407247-3,00007-X
Peter Sorensen (A,C), Ingrid K. Thomsen (A), and Jaap J. Schroder (B)
(A) Department of Agroecology, Aarhus University, Blichers Alle 20, 8830 Tjele, Denmark.
(B) Wageningen Plant Research, Wageningen University and Research Centre, PO Box 16, 6700 AP Wageningen, The Netherlands.
(C) Corresponding author. Email: firstname.lastname@example.org
Received 14 January 2017, accepted 19 May 2017, published online 23 June 2017
Caption: Fig. 1. Cumulative net mineralisation of cattle and pig slurry N in the soil over 5 years estimated using the new model.
Table 1. Additional N uptake recovered in barley plus catch crops (CC) from organic slurry N (given as a percentage) compared with uptake from mineral N-fertilised crops after application of cattle or pig slurry in the first year All plots received the same amount of inorganic N for a barley crop in the first year (100-103 kg N [ha.sup.-1]). Data are the mean [+ or -] s.e.m. Data are from Sorensen and Amato (2002) and Sorensen (2004) Manure 1st year 2nd year Ryegrass Spring CC barley Cattle slurry (81 kg organic N 8.5 [+ or -] 1.0 4.3 [+ or -] 1.7 [ha.sup.-1]), loamy sand (n = 12) (A) Pig slurry (51 kg organic N 12.6 [+ or -] 1.7 9.5 [+ or -] 2.2 [ha.sup.-1]), loamy sand and sandy loam soils (n = 24) (A) Manure 3rd year Ryegrass Spring CC barley Cattle slurry (81 kg organic N 4.7 [+ or -] 0.7 1.9 [+ or -] 3.5 [ha.sup.-1]), loamy sand (n = 12) (A) Pig slurry (51 kg organic N 5.3 [+ or -] 1.2 2.6 [+ or -] 3.6 [ha.sup.-1]), loamy sand and sandy loam soils (n = 24) (A) Manure Ryegrass CC Cattle slurry (81 kg organic N 2.5 [+ or -] 1.7 [ha.sup.-1]), loamy sand (n = 12) (A) Pig slurry (51 kg organic N 2.6 [+ or -] 1.6 [ha.sup.-1]), loamy sand and sandy loam soils (n = 24) (A) (A) Mean of three application (incorporation by mixing, simulated injection and surface banding) methods. Table 2. Estimated net mineralisation of organic N in pig and cattle slurry during the 3 years after application based on the observed apparent N recovery (ANR) in crops in Table I It is assumed that ANR in spring barley was similar to the ANR measured for mineral N and that ANR in the ryegrass catch crop (CC) was 49% of mineralised N in autumn-winter 1st year 2nd year Ryegrass Spring Ryegrass CC barley CC Estimated ANR of 49 60 49 mineralised N (%) Cattle slurry net mineralisation 17 (A) 7.2 10 (% organic N input) Pig slurry net mineralisation 26 16 11 (% organic N input) 3rd year Spring Ryegrass barley CC Estimated ANR of 60 49 mineralised N (%) Cattle slurry net mineralisation 3.1 5.0 (% organic N input) Pig slurry net mineralisation 4.4 5.4 (% organic N input) (A) Example: 8.5/0.49= 17 (8.5 taken from Table 1). Table 3. Model estimations of the yearly mineralisation of residual organic N in cattle and pig slurry N release in the first 3 years is based on data in Table 2. N release in the last 2 years is based on extrapolations using data from Hart et al. (1993) Manure 1st year 2nd year 3rd year Cattle slurry (% initial 17 17 8.1 organic N input) Pig slurry (% initial organic 26 27 9.8 N input) Cattle slurry (% remaining 17 20 12 residual organic N) (A) Pig slurry (% remaining 26 36 21 residual organic N) Manure 4th year 5th year Cattle slurry (% initial 4.5 4.1 organic N input) Pig slurry (% initial organic 4.9 4.3 N input) Cattle slurry (% remaining 7.8 7.8 residual organic N) (A) Pig slurry (% remaining 13.0 13.0 residual organic N) (A) Residual organic N is the organic manure N remaining in the soil at the start of each year when accounting for estimated N mineralised in previous years. Table 4. Net N mineralisation estimated as a percentage of added organic N from animal manures measured during field incubation in pots The manures were applied as fresh faeces or as anaerobically stored faeces in spring. Data are from van Faassen and van Dijk (1987) Manure type 0-6 months Sandy 6-18 months Sandy Sandy soil loam soil Sandy soil loam soil (3% clay) (18% clay) (3% clay) (18% clay) Cattle 37 18 14 8 Pig 41 31 17 11 Broiler manure 83 81 8 7 Table 5. Cumulative N mineralisation from different manure samples after 199 days (3184 accumulated degree-days above 0[degrees]C) estimated from the apparent N recovery in grass and expressed as a percentage of applied organic N taken up in grass in pots All ammonium-N in manure was stripped by drying before manure application. FYM, farmyard manure; Data are from Chadwick et al. (2000) % Organic N mineralised Manure type Range Mean Year 1 model estimate Cattle slurry 2-19 12 17 Cattle FYM (solid) 11-24 14 Pig slurry 21-37 27 26 Pig FYM (solid) 18-30 21 Poultry manure 16-56 29 Table 6. Nitrogen fertiliser replacement values (NFRV) measured after years of repeated applications of cattle slurry by injection to grassland in the Netherlands (S1UNi, cattle slurry from farm 1 untreated and injected and S2UNi, cattle slurry from farm 2 untreated and injected, data from Schroder et al. 2007) and estimated yearly mineralisation rate of organic manure N Year NFRV NFRV NFRV slurry NFRV S1UNi S2UNi (mean % yearly effect (% total N) (% total N) total N) (% total N) 1 54 66 60 9.5 (B) 2 66 73 70 9.5 3 70 76 73 3.5 4 74 77 76 2.5 Year NFRV yearly Estimated N effect (% organic mineralisation N input) (% organic N input (A)) 1 19 21 2 19 21 3 7 8 4 5 6 (A) Assuming that 90% of the yearly mineralisation was available for crop uptake in grass. (B) NFRV in the first year minus mineral N in the slurry (the N[H.sub.4]-N/total N ratio was 0.505).
|Printer friendly Cite/link Email Feedback|
|Author:||Sorensen, Peter; Thomsen, Ingrid K.; Schroder, Jaap J.|
|Date:||Aug 1, 2017|
|Previous Article:||Estimating nitrogen excretion and deposition by lactating cows in grazed dairy systems.|
|Next Article:||Modelling reduced coastal eutrophication with increased crop yields in Chinese agriculture.|