Production of 15N-Labelled Liquid Organic Fertilisers Based on Manure and Crop Residue for Use in Fertigation Studies

Large quantities of crop residue and animal manure from agricultural and livestock activities are annually produced worldwide. With proper management, these residues are potentially valuable sources of plant nutrients, mainly N. Recycling such subproducts in sustainably-based agricultural systems can minimise the use of mineral fertilisers, and hence reduce the potential risk of surface and groundwater pollution. Therefore, the purpose of this study was to obtain (small scale) two liquid labelled-organic fertilisers, an animal- and a vegetal-based organic (AO and VO, respectively) fertiliser, to be used as organic N sources in subsequent fertigation studies. Forage maize (Zea mays L.) grown under 15N-labelled fertiliser supply was used as raw material for VO fertiliser production, and also as 15N-labelled sheep feed to obtain 15N-labelled manure. The labelled faeces fraction was used as raw material for the AO fertiliser. The VO fertiliser was obtained after an acidic and an enzyme-driven hydrolysis. The AO fertiliser was obtained after acidic hydrolysis. The VO liquid fertiliser presented an N concentration of 330 mg·L-1, 85% of total N was organic, while ammonium and nitrate N accounted for 55% and 45% of the mineral nitrogen fraction, respectively. This fertiliser also exhibited high K, Ca and S concentrations and notable values for the remaining macro- and micronutrients. The AO liquid fertiliser had a similar total N concentration (496 mg·L-1, 82% of total N in an organic form) to that of VO, but its mineral N fraction significantly differed, which came in a predominantly (95%) ammonia form. It also had a high content of N, P, K and other macronutrients, and sufficient Fe, Zn, Mn, Cu and B levels, which suggests its suitability as a potential fertiliser. The percentage of 15N enrichment in both VO and AO liquid fertilisers exceeded 2% 15N atom excess, which enabled their use in subsequent assays run to assess nitrogen uptake efficiency.


Introduction
Large quantities of crop residue and animal manure generated by agriculture and livestock are annually produced worldwide. Today's annual USA production of agricultural residue is around 500 10 6 dry Mg [1]. In China, approximately 800Á10 6 Mg of agricultural residue are produced annually [2], while a similar value (approximately 700Á10 6 Mg) is obtained in the EU [3]. Regarding animal manure, the produced quantity depends on animal species, diet, age, the environment and productivity, among other factors, with values ranging from 30 to 25,500 kgÁyear -1 for broilers to lactating cows, respectively [4]. The total population in the EU amounts to 136 10 6 livestock units, of which 48% is bovine and 14% is poultry [5]. With proper management, this residue is a potentially valuable source of plant nutrients, mainly N. Its potential reuse as organic fertilisers could represent a sustainable approach to recycling nutrients and reintegrating organic matter into soil [6]. Recycling such nutrients in ecological and sustainably-based agricultural systems could not only minimise the use of mineral fertilisers, but could also reduce the potential risk of surface and groundwater pollution with nitrates. The objectives and targets of European legislation have been key drivers to improve residue management, to stimulate innovation in recycling and to limit landfilling use (EU, 2010). However in order to successfully recycle agricultural and livestock organic residue as a nutrient source, release of N and its uptake efficiency must be clearly identified. This is essential information because farmers are unable to consider organic residue a nutrient source when calculating fertiliser doses [7]. It would also help shed light onto the misleading perception that manure is an unreliable nutrient source [8]. Nevertheless, it is extremely complicated to predict the availability to crops of residue derived-N because it must be mineralised by soil microbes before being made available for its uptake [9]. This also varies vastly depending on the estimation method [8], and is dependent on the characteristics and "quality" of organic residue [10]. Some approaches based on the difference in N uptake in the presence and absence of added residue have been carried out, but poorly precise measurements (residue derived-N is a small fraction of total crop N uptake) and the impossibility of tracing N in soil pools greatly limit the information acquired from these indirect determinations [11].
In line with this, accurate residue-derived N tracing can only be achieved by using 15 N isotope-dilution techniques. The 15 N labelling technique has been previously used as a tool to assess the fate of manure N in long-term N cycling studies with daily cattle [8,12], goat or pig [13,14,15], poultry [16] and sheep [17,18,19] manure. Some research works have used this technique for the 15 N-labelling of plant residue [20,21,22]. However, these assays obtained organic residue in a solid form, which must be solubilised and mineralised before becoming plant-available [17]. Moreover, any nutrients present have a limited short-term effect on crop nutrition [23,24], but significantly contribute in crop N requirements during the year after applying manure. Given the implementation of more efficient irrigation systems (drip irrigation, subsurface irrigation or sprinkler systems), this solid organic residue cannot be used as a fertiliser without undergoing prior hydrolysis processes to increase its solubility to enable its use in fertigation.
Very few works on the production of liquid fertilisers using crop residue or animal manure have been published. Capulín-Grande et al. [25] assessed the production of liquid cattle manure extracts to study their subsequent contribution to Lolium perenne plant nutrition. Liquid fertilisers obtained from seaweed have been used in India and China for decades [26]. These liquid fertilisers have shown high levels of all trace elements, plant hormones and vitamins, but negligible amounts of N and P [27]. Similarly, a manure tea made by fermenting cow dung in water has been reported as a moderate NPK fertiliser and a major iron source [28,29] to evaluate the nutritional component of the liquid fertilisers that derived from three common plants. All liquid manures had a higher nutrient content than common solid organic fertilisers. Therefore, the purpose of this study was to obtain two liquid 15 N-labelled organic fertilisers (small scale), an animal and a vegetal-based organic fertiliser, to be used as a source of N and other essential elements to assess their efficiency as potential sources of N and other microand micronutrients in citrus fertigation in subsequent experiments.

Ethics statement
15N-labelled sheep manure production was carried out in the Institute of Animal Science and Technology (Valencia Polytechnic University, Spain). The experimental procedure was approved by the Animal Use and Care Committee of the Valencia Polytechnic University, and followed the codes of practice for animals used in experimental works proposed by the EU [30]. The remaining experiments were conducted in the Department of Citriculture and Vegetal Production of the Valencian Institute of Agrarian Research (IVIA, Moncada, Spain). Dr. Ana Quiñones was response for the experimental analyses in this manuscript and can be contacted in the future. The authors declare that this manuscript neither contains any ethic issue nor involves endangered or protected species. 15 N-labelled forage maize production Forage maize was used as raw material to obtain 15N-labelled animal and vegetal-based liquid organic fertilisers as it is one of the main cereals grown in the EU (EUROSTAT, 2015) and will, therefore, lead to larger amounts of crop residue.
In August 2009, forage maize (Zea mays L.) was grown in three plots, 10 m x 2 m each, on sandy-clay soil (67.5% sand, 10.6% silt, 21.9% clay; pH 7.9 and 0.58% organic matter content) at the IVIA centre (39°55´N, 0°39´W) in Moncada (Valencia, Spain). Plots were irrigated by microsprinklers and a nutrient solution that contained the necessary macro-and micronutrients. Six foliar extra-N applications were carried out fortnightly, three with a 0.5% (w/w) solution of (NH 2 ) 2 CO (urea) and three with a 1.0% (w/w) solution of (NH 4 ) 2 SO 4 (ammonium sulphate, AS). Solutions were 15 N-labelled (10 atom % 15 N in excess). As a result, plots were fertilised with 495 g urea and 1,220 g AS, which accounted for 483.9 g N (80.6 kg N ha -1 ) and 48.4 g 15 N in excess. Before applying the fertiliser, a plastic film was placed at a depth of 50 cm to avoid 15 N leaching loss.
In October, before plants started to flower, the aerial part of maize was cut and the root system was manually extracted. The whole root system was manually removed by pressurised water and by extracting small roots by hand. Plant tissues were separated into stems, both with and without leaves and roots. The water content of the leafy stems was reduced to 20% at room temperature to stabilise the plant matter to be used for hay for sheep feed. The other fractions were dried in an oven at 65°C and representative subsamples were stored at 4°C until the mineral concentration and nitrogen ( 15 N/ 14 N) isotopic compositions were further analysed.

The sheep-feed 15 N-labelling procedure
In order to increase the 15 N enrichment of sheep feed, 43.8 kg of stems and leaves were sprayed with a solution that contained 200 g of 15 N-labelled urea (10.2% atom 15 N excess). Extra labelled leafy stems were mixed and divided into 18 feed rations to ensure uniform 15 N enrichment in the daily feed supply [8]. 15

N-labelled sheep manure production
The forage method, by which ruminant livestock manure is labelled by feeding 15 N-enriched forage [12], although costly and laborious [17], it is the only available method for labelling urinary and faecal N components in the present-day, but ensures sufficient enrichment for subsequent organic fertiliser production.
Four dry sheep were used to produce labelled manure. Loss of 15 N in milk was thus avoided because between 20% and 30% of the N (protein) fed to a ruminant is converted into milk [8]. Sheep were housed in metabolic cages 7 days prior to the labelled feeding phase to give them time to acclimatise, and were fed with unlabelled straw to adapt them to the diet. Labelled maize (stem with leaf fractions) was administered for 18 days (from 22 February to 11 March 2010), offered ad libitum and renewed every 48 h. Uneaten maize was removed from the feeder before providing each new supply.
Urine and faeces were collected daily from cages separately since the 15 N enrichment of urine N was significantly lower than that of simultaneous faecal N [17]. Urine was collected from tubes that drained into containers fixed in ice to minimise gaseous N loss, and was then stored at 4°C. Faeces were hand-scraped from the perforated platform of the metal cages, weighed, oven-dried at 65°C and dry-weighed for dry matter determinations. Faeces and urine from each collection were subsampled and stored at 4°C for further analyses.

Obtaining liquid 15 N-labelled organic fertiliser from plant residue
Labelled maize fractions (500 g each) were mixed and ground in a vegetable shredder (Viking GE345), and further blended in a water-refrigerated mill (IKA M20, Staufen, Germany). The mixture was hydrolysed (cooking step) with 0.5% H 2 SO 4 at high pressure (1.5 MPa) and temperature (125°C). The sulphuric acid that came into contact with organic tissue destroyed it and facilitated its solubilisation by releasing the contained elements [25]. Finally it was washed to prevent the acidic solution from continuing the subsequent process, and was filtered to separate the solid and liquid (Liquid Fraction 1).
A 3-litre volume of water at 55°C was added to the solid fraction and pH was adjusted to 8.5 using NaOH. The resulting mix was subjected to action carbohydrate-degrading enzymes a multi-activer β-glucabase preparation produced by a selected strain of Humicola insolens, in which cellulase, xylanase, pentosanase and arabanase activities predominated. The hydrolysed fraction was filtered by producing Liquid Fraction 2 and a solid substrate. Protein hydrolysis was then performed in an alkaline medium in the solid fraction, which was resuspended in 4 L of water and adjusted to pH 3.0 with hydrochloric acid. Phytase (0.2 g and 6,000 unitsÁg -1 ) and an acid fungal protease (0.4 g), both produced from Aspergillus niger, were added and the solution was stirred at 50°C for 20 h. The entire slurry was heated to 95°C for 20 min, cooled down and adjusted to pH 5 with NaOH. Filtering led to a solid residue and Liquid Fraction 3. All the obtained extracts (Liquid Fractions 1, 2 and 3) were then mixed to achieve an extraction of around 75% of the original plant matter (Fig 1).
Obtaining liquid 15 N-labelled organic fertiliser from animal residue Labelled faeces were dried in an oven (65°C) and, after grinding as reported for the plant residue, were subjected to acid hydrolysis to enhance the N availability found in faeces for plants [24]. Faeces were digested in sulphuric acid (1.0%) at high pressure (1.5 MPa) and temperature (150°C), according to the method described by Capulín-Grande et al. [25]. These authors stated that addition of sulphuric acid to cattle manure led to higher P, Ca, Mg and S concentrations than natural liquid cattle manure extracts, or when stabilised by a warming-up process. After hydrolysis finished, extracts were decanted and the obtained fluid was filtered.

Analytical determinations
Subsamples of the 15 N-labelled maize and faeces were oven-dried (65°C) and ground in a water-refrigerated mill (IKA M20, Staufen, Germany) to pass through a 0.3 mm-diameter sieve. They were stored at 4°C until further analyses. In the vegetable samples, faeces and organic fertilisers, determinations of the total N concentration and 15 N abundance were made in an Elemental Analyzer (NC 2500, Thermo Finnigan, Bremen, Germany) coupled to an Isotope Ratio Mass Spectrometer (Delta Plus, Thermo Finnigan, Bremen, Germany). The N in the urea/ammonium and nitrite/nitrate forms present in urine was determined by the Kjeldahl method. Macro-and micronutrients were measured by simultaneous ICP emission spectrometry (iCAP-AES 6000, Thermo Scientific. Cambridge, UK). The results were expressed as a percentage (macronutrient) or parts per million (micronutrients) of dry weight (DW). All the determinations were made in duplicate. After each set of 10 (in the N analysis) or 15 (for the other elements) samples, a standard was run to ensure accuracy.
With the liquid organic fertilisers, an analysis of mineral nitrogen (nitrate and ammonium) was measured by a flow injection analysis (Aquatec 5400, Foss Tecator) following the methodology described by Raigón et al. [31]. The analysis of the mineral nitrogen fractions (N-NH 4 + -NO 3 -) was measured by the flow injection analysis (FIAstar 5000, Foss Tecator, Höganäs, Sweden). In order to determine isotopic composition, liquid fertilisers were steam-distilled (2200 Kjeltec, Auto Distillation Unit, Foss Tecator, Höganäs, Sweden). 15  -N were recovered in boric acid 0.32 N and reduced to dryness in an oven (P Selecta Barcelona, Spain) at 65°C. Aliquots were acidified with H 2 SO 4 before the analysis done in the above-mentioned mass spectrometer [32]. Organic nitrogen content was calculated as the difference between total and mineral N contents by the Kjeldahl method, as described by Bremner [32].

Calculations
The 15 N content in forage maize, organic residue and fertilisers obtained was calculated as follows: 15 N sample ðmgÞ ¼ Nð%; w=wÞ Á DWðgÞ Á atom % 15 N excess Á 10 À1 where atom % 15 N excess was calculated by subtracting the natural abundance of 15 N from the atmospheric N 2 , 0.3663 atom %, according to the International Atomic Energy Agency [33]. Total 15 N recovery represents the proportion of applied or fed 15 N (source), which is taken up by maize, is incorporated into faeces and urea, or is present in the fertilisers obtained.  Table 1.
The total maize biomass was 65.1 kg and leafy stems represented the main fraction (84% of the total DW). From the whole produced biomass, 25,743 mg of 15 N in excess were recovered from the labelled fertilisers, which represents 53.2% of the 15 N supplied (Table 1). This indicates that foliar N mineral fertilisation is a very efficient process to not only fertilise crops, but to also produce enough labelled plant matter for its subsequent use as raw material for organic fertilisers. In a field assay, Powell et al. [8] also found that 36-44% and 26-65% of 15 N-ammonium sulphate (10 atom % 15 N) was recovered in alfalfa and corn, respectively, in a 4-year experiment. Bosshard et al. [24] recovered 54% of the originally applied 15 N in ryegrass shoots when N was applied as K 15 NO 3 (15.9 atom % 15 N). The leafy stem fraction, mostly allocated from total plant 15 N (91%), resulted in an isotopic enrichment of the N pool in this compartment of 2.82 atom % 15 N excess. The root system showed remarkable higher 15 N enrichment (3.29 atom % 15 N excess), and its contribution to increase the average 15 N enrichment of the whole plant (2.82 atom % excess) was negligible given its low proportion of total biomass (8%). In a similar assay conducted to produce 15  . It is noteworthy that comparatively higher enrichments were obtained in these assays because ryegrass was cultivated in sand and no isotopic dilution effect by soil native N took place. Table 1. DW, N and 15 N content of the total maize biomass produced at the end of the labelling period, and the fractions used as sheep feed and as raw material for the vegetal-based organic fertiliser*. The macro-and micronutrient concentrations of the leafy stems fraction (Table 2) indicated that the source of vegetable organic fertiliser presented high potassium levels and sufficient quantities in the other nutrient elements.

Labelled sheep feed
Part of the leafy stem labelled fraction (43.8 kg) was used as sheep forage. Supplementary 15 Nurea labelled spray increased the 15 N enrichment of sheep feed on 3.75 atom % excess. This supplemental mineral N ensured greater isotopic enrichment of the subsequent sheep faeces, and also represented 13.7% of the total N in sheep feed. This was clearly below the threshold (20%) of the total dietary mineral N, which can cause toxicity. 15 N-labelled sheep manure production Extra labelled leafy stems were supplied as sheep forage; 38.9 kg were ingested by animals and 4.9 kg were collected from troughs as residual (Table 1). During the 15 N-labelling period, any uneaten hay was removed from the feeder before placing each new supply.
The ingest of labelled maize resulted in the increased 15 N enrichment of faecal-N over time (Fig 2). On the initial days of feeding with the 15 N-labelled feed, the 15 N in faeces scarcely exceeded natural abundance (0.3663 atom % excess). This was because it was diluted by unlabelled indigestible feed N, which remains in the digestive tract, and by endogenous N, which consists in microbial products and microorganisms from the rumen, intestine and hind gut and N that originate from the digestive tract [8,34,35]. Subsequently, the 15 N in faeces continues to be diluted by partly labelled endogenous N [17].
From day 3 of feeding on 15 N, the 15 N concentration considerably increased to reach a plateau value of 2.28 atom % excess between days 11 and 19, which coincided with the end of the feeding period on 15 N hay. Similar results have been obtained by Sørensen et al. [17] from sheep fed 15 N-labelled hay (4.52 atom % 15 N excess) and Chantigny et al. [14] by feeding pigs an 15 N-enriched diet (2.81 atom % 15 N). These authors obtained 15 N enrichment of 3.70 and 2.32 atom % after 9 and 12 days of labelled feeding, respectively. According to these authors, homogeneous labelling of sheep manure was achieved after a 7-day feeding period with 15 Nlabelled hay, which is in line with the results presented herein. At the end of isotope-enriched feeding, 58% of the 15 N initially present in the labelled leafy stem fraction was recovered in sheep faeces. Brosshard et al. [24] recovered about 77% of the 15 N in feed (ryegrass) in sheep manure.
From day 17 onwards, the labelled diet was replaced with normal unlabelled hay. Once the labelled feed was replaced with unlabelled fed, the 15 N percentage in faeces dramatically lowered and, 22 days after starting the labelled diet, 15 N enrichment barely exceeded natural abundance.
The manure obtained between days 7 and 19 of labelled feeding (Fig 2) accounted for 12.4 kg (DW), with 1.79% and 2.18% of N and 15 N excess, respectively (Table 3). These labelled Table 2. Macro-and micronutrient concentration of the maize used as raw material for obtaining the vegetal-based organic fertiliser at the end of the growing cycle*. faeces were subsequently used to obtain an animal-based organic fertiliser, which presented 15 N enrichment above 1.9%. These values (quantity and enrichment) enabled the use of the organic fertiliser obtained as a labelled product for subsequent trials based on the isotope dilution technique. The faeces that corresponded to previous and latter days (days1 to 6 and days 21 to 26) were discarded for their low 15 N enrichment (less than 1% on average).

Macronutrients
At the end of the 15 N-enriched feed period, the 15 N recovered in solid manure accounted for 18% of the 15 N present in sheep forage. This percentage increased to 29% when the discarded faeces and urine were counted ( Table 3). The rest of the labelled-N was probably in the N assimilated by sheep. Higher values were found in the urine and faeces of the cows (51-64%) fed with enriched forage [8].
At the end of the labelled feeding period, the average isotopic enrichment of sheep urine was lower than that obtained in faeces (Table 3), and was always below 2% (data not shown) during the collecting period. Urine was, therefore, ruled out for obtaining the animal organic fertiliser. Powell et al. [8] also reported greater enrichment in faeces than in urine in cows fed 15 N hay. Two days after starting 15 N feeding, the 15 N enrichment in faeces exceeded that in the urine 15 N excess in other studies conducted with sheep [24].
As in forage maize, macro-and micronutrient concentrations were also determined in the solid manure (Table 4). The faeces used for the animal-based organic fertiliser generally obtained higher concentration values for most sampled nutrients (N, P, Mg, Ca, S, Fe, Zn, Mn, Cu and B) compared to forage maize, with similar levels for the potassium element as in the labelled forage maize. Therefore, sheep faeces proved also to be a major potential source of plant nutrients.
Obtaining liquid 15 N-labelled organic fertiliser from plant residue Uneaten hay, this being labelled maize leafy stems that were not extralabelled with sprayed urea, leafless stems and roots were used to obtain the 15 N-labelled organic fertiliser from plant residue. The maize used as raw material presented 1.29% N and 3.09 atom % 15 N in excess ( Table 1).
The resulting vegetal-based liquid fertiliser presented high nitrogen concentrations (Table 5), mostly in an organic form (85%), while ammonium and nitrate-N respectively accounted for 55% and 45% of the mineral N fraction. The isotope enrichment of the resulting vegetal-based liquid fertiliser was sufficient for further uptake efficiency studies. The obtained liquid organic fertiliser also had high K, Ca and S concentrations, and showed notable values for the other macro-and micronutrients. The obtained plant-based liquid fertiliser retained 43% of the N originally present in the maize fractions used as raw material (Tables 1 and 5). Obtaining liquid 15 N-labelled organic fertiliser from animal residue As previously indicated, solid manure with enrichment above 1.9% atom 15 N excess was used as raw material to obtain the animal-based liquid fertiliser. The obtained liquid animal-based fertiliser was characterised; N total , ammonia, nitric and organic N concentration and atom % 15 N in excess, and the main macro-and microelements, were determined ( Table 6). The animal-based organic fertiliser showed a total N concentration of 496 mgÁL -1 , mostly (82%) as organic forms. The mineral-N fraction, which accounted for only 18% of the total N present in the fertiliser, mainly came in an ammonia form (95% of mineral N). These results support those obtained by Jackson-Smith et al. [36], who obtained much smaller amounts of ammonium and higher concentrations of organically bound N in semisolid dairy manure than in slurry.
The percentage of 15 N enrichment in the animal-based liquid fertiliser was 2.17% lower than that in the vegetal-based one (2.62%). In this cooking, the process to obtain liquid fertiliser was less efficient, and about 30% of the N present in labelled faeces was recovered in the final fertiliser, where the by-products of the decanting and filtering stages were the main points of 15 N loss. Carballo et al. [37] also managed to obtain a biofertiliser from a compost extract and their nitrogen recovery was about 55% of the 15 N applied as labelled faeces. In order to increase the efficiency of this process on an industrial scale, the use of appropriate enzymes or microbe inoculums for hydrolysis or fermentation processes should be considered. A concentration step would also be of interest in order to cut the expenses associated with transporting the final liquid fertiliser.
The animal-based organic fertiliser had a similar potassium concentration to that of the vegetal-based organic fertiliser. The animal manure extract showed higher concentrations of the other nutrients than those obtained in the liquid plant fertiliser, which was expected in view of the raw materials analysis results (Tables 3 and 4).

Conclusions
Processes to obtain two liquid fertilisers from crop residue and animal manure were assessed. Both fertilisers showed 15 N enrichment that enabled their use in a subsequent experiment to assess N uptake efficiency, and to hence evaluate them as potential N sources in citrus Table 5. Comparative volumes, total N and the N concentrations among fractions, 15 N excess and N recoveries of liquid vegetal-and animalbased organic fertilisers (mgÁL -1 )*.
doi:10.1371/journal.pone.0150851.t005 Table 6. Comparative values of macro-and micro concentrations in the liquid vegetal-and animal-based organic fertilisers (mgÁL -1 )*. fertigation. Their liquid presentation would also allow them to be used in more efficient irrigation systems, such as drip irrigation. The presence of other macro-(P, K) and micronutrients (Fe, Zn, Mn, Cu and B) suggests their suitability as potential multinutrient fertilisers.

Author Contributions
Conceived and designed the experiments: FL AQ. Performed the experiments: BMA MRMC CF FL AQ. Analyzed the data: BMA FL AQ. Contributed reagents/materials/analysis tools: BMA AQ. Wrote the paper: BMA AQ.