Plant N uptake in the periphery of a bovine urine patch: determining the ‘effective area’

ABSTRACT

The extent to which the wetted soil area of a urine patch influences surrounding pasture is relatively unknown. The study objective was to use ¹⁵N tracer to quantify pasture N uptake in the ‘wetted’ and periphery areas of a spring deposited bovine urine patch over 311 days. Ruminant ¹⁵N enriched urine was applied to soil creating a circular wetted area, ‘zone A’ (800 kg N ha⁻¹), with and without urea fertiliser (35 kg N ha⁻¹). Pasture yields, ¹⁵N recovery and soil inorganic-N dynamics were monitored from zone A and two peripheral zones, B and C. Fertiliser had no effect on cumulative urinary ¹⁵N recovery in pasture (50%–52%). Average cumulative pasture ¹⁵N recovery in zones A, B and C were 30.6%, 17.3% and 4.2%, respectively. Soil inorganic-¹⁵N recovery occurred in zones A and B, declining with distance from the wetted area. The results suggest an effective urine patch area of 0.95 m² or 3.4 times the wetted area.

KEYWORDS:

• Area
• bovine urine patch
• fertiliser
• ¹⁵N
• urine

Introduction

Nitrogen (N) excreta from grazing animals plays a key role in cycling N through pastoral systems (Haynes & Williams 1993; Selbie et al. 2015). Ruminant urine deposition accounts for nearly 70% of the N returned to pasture at rates of 600–1000 kg N ha⁻¹ (Haynes & Williams 1993; Selbie et al. 2015). Dairy cow urination events approximate 2 L in volume (Doak 1952), occur 10–12 times a day (Jarvis et al. 1995) and cover an average surface area of 0.26 m², with a range of 0.16–0.49 m² (Petersen et al. 1956; Davies et al. 1962; Hogg 1968; Richards & Wolton 1976; Haynes & Williams 1993). These values vary significantly and are influenced by grazing conditions and environmental factors (Haynes & Williams 1993). The volume of soil wetted by a urine patch varies with surface area affected, antecedent soil moisture, soil surface water repellency, surface compaction, microtopography, vegetation cover, slope and wind (Williams & Haynes 1994). Urine patches are estimated to cover 20%–30% of the grazed paddock area per year, with this percentage varying with the stocking rate (Silva et al. 1999; Moir et al. 2011).

Some of the urinary N returned to soil may be volatilised as NH₃, but much is eventually nitrified, resulting in the accumulation of nitrate () in the soil. Pasture N uptake from a urine patch has been estimated at 300–700 kg N ha⁻¹ yr⁻¹ (During & McNaught 1961; Ball et al. 1979; Ledgard 2001). Thus, the N loading under a urine patch far exceeds immediate plant demand. Surplus in the soil is vulnerable to leaching and this is recognised as a significant N loss pathway from dairy farming systems (Silva et al. 1999). Loss of such reactive N has environmental implications, since can potentially contribute to the eutrophication of surface waters, indirect losses of nitrous oxide, the formation of hypoxic zones and contamination of potable waters (Galloway et al. 2004). As a result, significant research has focused on understanding N loss dynamics under urine patches within dairy farming systems.

A urine patch is comprised of a ‘wetted area’ (where urine is directly deposited) and an ‘effective area’ which includes both the wetted area and the peripheral area outside it that is affected by the urinary N, primarily via plant root extension and soil N diffusion (Lantinga et al. 1987; Tinker & Nye 2000). Cattle urination events can create wetted areas of 0.014–0.49 m² (Selbie et al. 2015). The effective area for cattle is reported to range from 0.03–1.1 m² (Lotero et al. 1966; Moir et al. 2011). However, data are limited. Decau et al. (2003) measured urinary N diffusion up to 20 cm from the wetted area of 0.4 m² urine patches. Lotero et al. (1966) reported effective areas of 0.9–1.2 m² and observed a linear decrease in dry matter (DM) yields from the centre of a urine patch to the periphery of the effective area. Moir et al. (2011) also observed seasonal variation in the effective area whereby urine-affected areas tended to be larger from spring/summer deposited urine, and smaller in winter and autumn. These differences are mainly attributed to rapid winter leaching when soils are draining (less N available for plant uptake), but other factors may include seasonal differences in both animal N and water intake.

Many N leaching studies have used lysimeters to quantify losses from urine patches (Clough et al. 1996, 1998; Silva et al. 1999; Di & Cameron 2002, 2004, 2007; Decau et al. 2003, 2004; McDowell & Houlbrooke 2009; Ledgard et al. 2014; Shepherd et al. 2014; Buckthought et al. 2015). In many cases, urine treatments cover the whole area of the lysimeter. However, a lysimeter is a confined area of soil and its casing represents a physical barrier to the surrounding soil. Therefore, urine patch leaching data obtained under such conditions are reflective of the wetted area of a urine patch and do not account for the full effective area. This may have implications for extrapolating lysimeter study results to field situations and the use of such data to understand N cycling in farm system models. Clearly, further research is needed to compare the wetted and effective areas in terms of their relative responses to urinary-N deposition. Thus, the objective of this study was to use a ¹⁵N tracer methodology to quantify the pasture response, N uptake and ¹⁵N recovery in the ‘wetted’ and ‘effective’ urine patch areas after a simulated spring urine deposition event. This was assessed in combination with, and without, an application of urea fertiliser which often follows deposition on New Zealand dairy farms.

Materials and methods

Experimental site and preparation

This field study took place at Scott Farm, owned and operated by DairyNZ, and located in the Waikato Region of New Zealand (−37.70°S; 175.36°E). The soil type was a Horotiu Silt Loam (Typic Orthic Allophanic) (Hewitt 1998) with a pH of 5.6 and a bulk density of 0.87 Mg m⁻³. Four months before the start of the study, the plot area was fenced off and any visible dung patches removed. The site was regularly cut to a pasture residual of 1600 kg DM ha⁻¹ (c. 3 cm height). In early October 2011, due to insufficient late winter rain, the trial area was irrigated with 350 mm of water to remove any residual antecedent urine patch effects. The study site consisted of 16 circular plots (area 2.0 m²), each comprising three concentric rings (Figure 1). The central ring, termed ‘zone A', was 0.28 m² (0.6 m diameter); the middle ring, ‘zone B', was 0.67 m² (0.25 m from the edge of zone A) and the outer ring, ‘zone C', was 1.06 m² (0.25 m from the edge of zone B). Zone A was designated as the wetted area, and zones A, B and C, collectively, comprised the potential effective area (Figure 1).

Figure 1. Plot layout showing the concentric zones A, B and C and their dimensions and areas.

Experimental design and treatments

There were four treatments: (1) a single bovine urine application (800 kg N ha⁻¹); (2) a single bovine urine application (800 kg N ha⁻¹) with a single concurrently applied urea fertiliser application (35 kg N ha⁻¹); (3) a fertiliser-only application (35 kg N ha⁻¹); and (4) a control that received no urine or fertiliser. Treatments were applied only to zone A. Four replicates of each treatment were established in a Latin square design. All treatments were applied in spring on 17 October 2011. Herbage and soil samples were taken on 17 October prior to treatment application.

Fresh dairy cow urine was collected on 16 October 2011 during the morning milking at the AgResearch dairy farm at Tokanui. The urine was standardised to a concentration of 11.2 g N L⁻¹ and a ¹⁵N enrichment of 5 atom% using a combination of urea at natural abundance and 99 atom% ¹⁵N enriched urea. Two litres of urine were applied to appropriate plots with a watering can with particular care taken to avoid urine spreading outside zone A. Plots receiving fertiliser then had urea fertiliser prills applied. All treatments were washed in with 10 mm of irrigation to minimise NH₃ volatilisation (Black et al. 1987). An equivalent volume of water was applied to the control treatments.

Sample collection and analysis

Sampling began the day after treatment application (18 October 2011). To ensure the plots were sampled representatively, each concentric zone was divided into six equal segments that were designated as either soil or pasture sampling areas. This was to prevent anomalies in yield measurements occurring as a potential result of soil coring damaging plants, and to reduce spatial variability within each plot. Samples from each segment within a plot-zone were then bulked together to create a single sample for each zone for analysis. Pasture was cut to a residual of 1600 kg DM ha⁻¹ (c. 3 cm height) with electric clippers. A total of 10 pasture harvests were performed on days 14, 32, 58, 85, 106, 133, 165, 203, 239 and 311 post treatment applications.

Harvested pasture was oven dried at 55 °C for 48–72 h. Dried samples were then weighed and finely ground using a Cyclone Sample Mill with a 1 mm screen (UDY Corporation). Analyses for total N and ¹⁵N enrichment were performed on an isotope ratio mass spectrometer (EA-CF/IRMS; PDZ Europa GSL/20–20).

Soil sample collection was carried out weekly for the first 4 weeks. After this, sampling was reduced to fortnightly, monthly and then bimonthly intervals before ceasing at the end of January 2012. Soil cores (2.5 cm diameter) were collected from two depths (0–7.5 cm and 7.5–20 cm) from all zones. The number of cores taken from each segment was limited due to the small area available (particularly in zone A). Totals of two, four and six cores were taken from zones A, B and C, respectively. Soil samples were stored in plastic bags at 4 °C. Preparation of the soils for analysis began within 24 h of collection. All soil samples were passed through a 4 mm sieve and any herbage material was removed. Soil gravimetric moisture content (θ_g) was calculated as the difference between the field moist and oven dried (105 °C for 24 h) subsamples.

Soil inorganic N was determined using a potassium sulphate (K₂SO₄) extraction procedure (1:5 soil to extractant ratio) (Mulvaney 1996), with analyses for ammonium (NH₄-N) and nitrate-N + nitrite-N (NO₃-N + NO₂-N) performed on a Skalar SAN⁺⁺ segmented flow analyser (Skalar Analytical). Soil inorganic N extracts were prepared for ¹⁵N enrichment analysis using the diffusion method as described by Brooks et al. (1989). Soils were later dried and sieved and analysed for total N and ¹⁵N enrichment. Soil organic ¹⁵N recovery was calculated as the difference between total soil ¹⁵N recovery and inorganic ¹⁵N recovery. All soil ¹⁵N analysis was carried out using isotope ratio mass spectrometry (EA-CF/IRMS; PDZ Europa GSL/20–20).

Daily weather data were obtained over the duration of the experiment from the Ruakura climate station (via the NIWA Virtual Climate Station dataset [Tait & Turner 2005]), located 7 km from the field site.

Statistical analyses

Statistical analyses were performed using the ANOVA directive of GenStat (15th edition). The least significant difference (LSD) was used to determine statistical variation between the sample means at the 95% significance level.

When comparing treatment effects on pasture yield, N content and N uptake between whole plots, a weighted mean was calculated to account for the differences in surface area between zones A, B and C. (Equation 1 using pasture yield as an example).(1) Where Yield^a, Yield^b and Yield^c are the measured yields from zones A, B and C; Area^a, Area^b and Area^c are the surface areas (m²) of zones A, B and C; and Area^a+b+c is the area of the total plot.

Results

Climate data

Rainfall was evenly distributed and totalled 1088 mm for the experimental period (Figure S1A). Average air temperature over the experimental period was 13.9 °C and the average soil temperature (10 cm depth) was 14.8 °C (Figure S1B). There was a total of 15 frost days which occurred between 23 May and 19 Jul 2012. There were no significant differences in gravimetric moisture (θ_g) between zones or treatments.

Pasture dry matter yields and N uptake

Mean DM yields for individual harvests ranged from 530–2200 kg DM ha⁻¹ in zone A, 500–1800 kg  DM ha⁻¹ in zone B and 180–1500 kg DM ha⁻¹ in zone C (Figure 2). Yields of DM were greater in the urine treatments than the non-urine treatments at each harvest until day 85, however, these differences were only statistically significant (P < 0.001) on the second and third harvests (days 32 and 58). In the urine treatments, DM yields from zone A were greater than zone B (P < 0.001) up to day 58; and greater than zone C (P < 0.05) up to day 85. Pasture DM yields from zone B were greater than zone C (P < 0.05) on only two harvests (Figure 2). In the non-urine treatments, DM yields from zone A were greater than zones B and C on day 58 only. There was no fertiliser effect on DM yields in any treatment.

Figure 2. Mean pasture yields in zones A, B and C for treatments A, urine + fertiliser; B, urine only; C, fertiliser only; D, control. Error bars = LSD (95%), n = 4.

Cumulative DM yields ranged from 7500–13,000 kg DM ha⁻¹, 6300–8300 kg DM ha⁻¹ and 5600–6700 kg DM ha⁻¹ in zones A, B and C, respectively. In the urine treatments, cumulative DM yields from zone A were greater (P < 0.001) than those from zone B or C, with no effect of fertiliser addition (Figure 3). In the non-urine treatments, there were no significant differences in cumulative DM yields between zones A, B or C (Figure 3).

Figure 3. Cumulative pasture yields in zones A, B and C for each treatment. Different letters represent a statistically significant difference between yields within treatments. Error bar = LSD (95%), n = 4.

Pasture N concentrations ranged from 1.9%–4.4% in zone A, 2.1%–3.4% in zone B and 2.1%–3.1% in zone C (Figure 4). After urine application, pasture N concentrations in zone A were greater (P < 0.05) than in zones B and C on days 14 and 32 (Figure 4). Pasture N concentrations in zones A and B of the urine treatments were greater (P < 0.05) than in any of the zones within non-urine treatments from day 14 to 58 (zone A) and day 14 to 32 (zone B). Pasture N concentrations in zone C did not differ between treatments (Figure 4).

Figure 4. Mean pasture N content in zones A, B and C for the A, urine + fertiliser; B, urine only; C, fertiliser only; and D, control treatments. Error bars = LSD (95%), n = 4.

After urine application, pasture N uptake generally decreased with time in zones A and B, but remained relatively constant in zone C (Figure 5). In the urine treatments, pasture N uptake in zone A was greater (P < 0.001) than in zones B and C on days 14, 32 and 58 (Figure 5). Pasture N uptake from zone B was greater (P < 0.001) than from zone C on days 14, 32 and 58 in the urine + fertiliser treatment, and on days 14 and 32 in the urine-only treatment (Figure 5). In non-urine treatments, N uptake remained relatively constant over time with no differences in N uptake between zones, apart from day 58, when an increase in pasture N uptake was observed in zone A only of both non-urine treatments. Fertiliser had no effect on N uptake at the zone scale, neither was there any interaction between fertiliser and urine treatments (Figure 5).

Figure 5. Mean pasture N uptake in zones A, B and C for the A, urine + fertiliser; B, urine only; C, fertiliser only; and D, control treatments. Error bars = LSD (95%), n = 4.

Cumulative N uptake in the urine-affected treatments was greater in zone A than in zones B and C (P < 0.001); however, cumulative N uptake from zone B did not differ from zone C (Figure 6). There was no difference in cumulative N uptake between zones in the non-urine treatments (Figure 6).

Figure 6. Mean cumulative N uptake in zones A, B and C for each treatment. Different letters represent a statistically significant difference within treatments. Error bar = LSD (95%), n = 4.

¹⁵N recovery in urine-treated pasture

During individual harvests, the recovery of ¹⁵N in zone A of the urine-treated pasture was greatest following harvests on days 14 and 32, and then declined relatively quickly until day 106 (Figure 7A–B). By day 106, pasture ¹⁵N recovery at individual harvests had decreased to be less than 1.5% of ¹⁵N applied, and by day 311 was only 0.35% of the ¹⁵N applied (Figure 7A–B). In both urine treatments, pasture ¹⁵N recovery from zone A was greater (P < 0.001) than from zones B and C at every harvest. Pasture ¹⁵N recovery from zone B was greater (P < 0.001) than zone C at every harvest (Figure 7A–B). Pasture ¹⁵N recovery in zone C occurred until day 106. Within zones, there was no interaction between the urine and fertiliser treatments on pasture ¹⁵N recovery at any sampling occasion.

Figure 7. Pasture ¹⁵N recovery in zones A, B and C for the A, urine + fertiliser; B, urine only; C, fertiliser only; and D, control treatments. Error bars = LSD (95%), n = 4.

Cumulative pasture ¹⁵N recovery increased rapidly from day 14 to day 106 (Figure 7). Total cumulative pasture ¹⁵N recoveries from all zones did not differ between the urine plus fertiliser (52.1%) and urine-only (49.7%) treatments (Table 1). Similarly, urine treatment, with or without fertiliser, had no effect on cumulative ¹⁵N recoveries when comparing individual zones A, B and C (Table 1). However, within urine treatments, the cumulative ¹⁵N recovery from zone A was greater (P < 0.05) than from zones B and C, and the cumulative recovery from zone B was greater (P < 0.05) than from zone C at every harvest (Table 1).

Table 1. Cumulative recovery of ¹⁵N in pasture and soil at day 106. There was no effect of fertiliser on ¹⁵N recovery.

Fraction	Zone	Urine + fertiliser (% ^15N recovered)	Urine only (% ^15N recovered)
Cumulative plant ^15N	Zone A	30.6a	29.7a
	Zone B	17.3b	17.3b
	Zone C	4.2c	2.7c
	Subtotal	52.1	49.7
Soil total ^15N 0–7.5 cm depth	Zone A	12.0a	13.5a
	Zone B	2.7b	7.2b
	Zone C	1.8c	2.0c
	Subtotal	16.5	22.7
Soil total ^15N 7.5–20 cm depth	Zone A	3.2a	5.4a
	Zone B	0.7b	3.1b
	Zone C	0.7c	1.6c
	Subtotal	4.1	10.1
Total ^15N recovered	Zone A	45.8a	48.6a
	Zone B	20.7b	27.6b
	Zone C	6.7c	6.3c
	All zones	73.2	82.5

Letters indicate differences between zones (P < 0.05) within a given fraction.

Soil N concentrations and soil ¹⁵N recovery

At both soil depths (0–7.5 and 7.5–20 cm), soil concentrations were greater (P < 0.001) in the urine-affected treatments than the non-urine treatments from day 1 to 21. At both depths, soil concentrations under the urine treatments were greater (P < 0.001) in zone A than in zones B or C, from day 1 to 21; and were greater in zone B compared with zone C from day 1 to 14 (Figure 8A–B). After this time, soil concentrations equalled those in the control, with no difference between zones. In the fertiliser treatment, soil concentrations were greater (P < 0.05) in zone A, compared with zones B and C, only on day 1 (Figure 8A–C). There was no difference in soil concentrations between zones in the control at either depth (Figure 8D).

Figure 8. Soil concentration at two depths (0–7.5 and 7.5–20 cm) in zones A, B and C for the A, urine + fertiliser; B, urine only; C, fertiliser only; and D, control treatments. Error bars = LSD (95%), n = 4. Note different scales on the y axes.

In the 0–7.5 cm depth, soil concentrations peaked in zone A on day 21 in the two urine treatments, and on day 1 in the fertiliser-only treatment (Figure 9A–C). In the 7.5–20 cm depth, peaks in soil concentration were observed only in the urine treatments (also day 21) (Figure 9A–B). Under the urine treatments, at 0–7.5 cm depth, soil concentrations were greater (P < 0.001) in zone A compared with zones B and C until day 63. They were greater (P < 0.05) in zone B compared with zone C until day 43 (Figure 9A–B).

Figure 9. Soil concentrations at two depths (0–7.5 and 7.5–20 cm) in zones A, B and C for the A, urine + fertiliser; B, urine only; C, fertiliser only; and D, control treatments. Error bars = LSD (95%), n = 4. Note different scales on the y axes.

In the non-urine treatments, there was no difference in soil concentrations between zones, with the exception of day 1 in the fertiliser-only treatment, where there was a greater concentration in zone A at the 0–7.5 cm depth (Figure 9C–D).

Inorganic-¹⁵N was recovered up to day 43 (Figure 10) after which soil inorganic-N concentrations were too low to allow for analysis. Figure 10 shows the temporal pattern of ¹⁵N recovery, as inorganic-N , by treatment, zone and soil depth at each soil sampling event during the first 43 days.

Figure 10. Soil inorganic ¹⁵N recovery for the A, urine + fertiliser at 0–7.5 cm depth; B, urine only at 0–7.5 cm depth; C, urine + fertiliser at 7.5–20 cm depth; and D, urine-only treatments at 7.5–20 cm depth. Error bars = LSD (95%), n = 4.

At 0–7.5 cm depth, soil inorganic ¹⁵N recovery in zone A of the urine plus fertiliser treatment declined from 43% on day 1 to 13% on day 43. In zone A of the urine-only treatment, inorganic ¹⁵N recovery declined from 44% to 8% over the same time. Inorganic ¹⁵N recovery only occurred on days 1 and 8 in zone B. In this time, inorganic ¹⁵N recovery declined from 25% to 0% and 38% to 0% in the urine plus fertiliser and urine-only treatments, respectively (Figure 10A–C). No inorganic ¹⁵N recovery occurred in zone C at 0–7.5 cm depth.

At 7.5–20 cm depth, soil inorganic ¹⁵N recovery in zone A of the urine plus fertiliser treatment declined from 32% on day 1 to 10% on day 43. In zone A of the urine-only treatment, inorganic ¹⁵N recovery declined from 21% to 4% over the same time (Figure 10B–D). No inorganic ¹⁵N recovery occurred in zones B and C at 7.5–20 cm depth.

By day 106, at 0–7.5 cm depth, the recovery of total soil ¹⁵N (as organic-¹⁵N) in zone A was 12.0% and 13.5% for the urine plus fertiliser and urine-only treatments (Table 1). These recoveries were greater (P < 0.01) than those measured for both zone B and zone C (Table 1). At the 7.5–20 cm depth, the recoveries ranged from 0.7% to 3.2% in the urine + fertiliser treatment and 1.6% to 5.4% in the urine-only treatment (Table 1). There were no treatment or zone effects at this depth.

Discussion

The observed increase in pasture ¹⁵N recovery under the urine treatments was a function of increased pasture N uptake which in turn was a product of increases in pasture DM yield and N content.

Under the urine treatments, the total pasture ¹⁵N recoveries (49%–52%) from all zones were similar to those reported previously (Decau et al. 2003; Leterme et al. 2003), where urine-¹⁵N recoveries ranged from 30%–65% when urine was applied in spring or summer. Recovery of spring-applied urinary N is generally higher because conditions are usually more favourable for pasture growth for a longer period that precedes the following winter (Ball & Field 1982). Pasture ¹⁵N recoveries tend to be lower after autumn or winter urine application due to greater leaching and denitrification losses of N (Fraser et al. 1994; Williams & Haynes 1994; Clough et al. 1996, 1998; Di et al. 2002; Silva et al. 2005). Although summer rainfall in the current experiment was higher than average in December 2011 and January 2012, this occurred c. 1 month after urine application and did not appear to cause a significant reduction in ¹⁵N recovery by the pasture.

Greater pasture ¹⁵N recovery occurred in zone A because the urine application, and thus the total N loading, was confined to zone A. The fact that 22% of the urinary ¹⁵N (18% in zone B and 4.2% in zone C) was recovered by pasture outside the wetted area may have been due to NH₃ deposition on to soil or plants outside the wetted zone A, plant uptake of volatilised NH₃, lateral movement of soil inorganic-N, or N uptake by plants in zones B and C that had rooting structures within zone A.

Ammonia volatilised from the soil following ¹⁵N enriched urine application can be directly absorbed through the stomata of plants (Farquhar et al. 1980; Sommer & Jensen 1991; Schjoerring & Mattsson 2001; van Hove et al. 2002). A number of studies have assessed NH₃ absorption and emission from crop plant canopies such as maize (Denmead et al. 1982; Harper et al. 2000), sugarcane (Denmead et al. 2008), rapeseed, barley, wheat and peas (Schjoerring & Mattsson 2001). However, little is known about the extent of volatilised NH₃ that is redeposited and absorbed by pasture following urine deposition, or the distances that NH₃ travels from the urine patch (Ross & Jarvis 2001). Sommer & Jensen (1991) measured foliar absorption of NH₃ by ryegrass at various distances from a dairy farm yard and measured deposition of 3.0 and 0.7 g N m⁻², at 10 m and 130 m from the source, respectively. Ross & Jarvis (2001) measured NH₃ emission and deposition, using wind tunnels, from ryegrass swards affected by urine (227 kg N ha⁻¹) and found that 20%–60% of NH₃ emitted was deposited within a distance of 2 m. Thus NH₃ deposition rates and absorption by nearby pasture can be significant. The exchange of NH₃ between the plant canopy and the atmosphere may be bidirectional and pastures have been shown to be a source of NH₃ emissions (Harper et al. 1996; Plantaz 1998). However, under intensive farm management conditions, van Hove et al. (2002) found gaseous NH₃ concentrations inside ryegrass leaves were always lower than atmospheric concentrations, and concluded that the ryegrass pasture canopy was unlikely to be a significant NH₃ source. Negating the potential for plants to be a source of NH₃ in the current experiment is the fact that the pasture canopy's leaf area was relatively small due to the pasture having been harvested prior to urine application. Thus, emission of NH₃ from plants within zone A, with subsequent plant uptake in zones B or C, is unlikely to have made a significant contribution to plant ¹⁵N recovery in zones B and C. Direct emission of NH₃ from the soil and its subsequent deposition in zones B and C should also have been limited due to the fact that 10 mm of water was applied to suppress NH₃ volatilisation (Black et al. 1987; Fraser et al. 1994; Clough et al. 1996; Saarijarvi et al. 2006) thus volatilisation of NH₃ from soil in zone A, with subsequent plant absorption in zones B and C, would also be expected to be minimal. Plant ¹⁵N uptake resulting from potential diffusion of NH₃ through the soil pore network should also have been minimal due to water occupying soil pores. It is clear, however, that more research is required to determine the role that NH₃ absorption and emission from pastoral systems play in determining the fate of N in a urine patch.

Lateral subsurface movement of urine-derived inorganic-¹⁵N is a potential alternative explanation for pasture ¹⁵N recovery outside of zone A. Cumulative pasture ¹⁵N uptake was essentially complete by the time that soil inorganic-N forms were at concentrations too low to be analysed (106 days) and it is also within this time period that major differences between zones were observed with respect to pasture ¹⁵N recovery at individual harvests. Inorganic soil ¹⁵N recovery was, as expected, greatest on day 1. The fact that urinary ¹⁵N was recovered in zone B on day 1 in both the urine + fertiliser and urine-only treatments indicates that urinary N moved from zone A into zone B within 24 h of application, via lateral macropore flow (Williams et al. 1990). Diffusion of along a concentration gradient (Wild 1972) cannot explain the rapid recovery of plant ¹⁵N in zone C and is unlikely to have contributed significantly to the pasture ¹⁵N recovery in zone B because the diffusion coefficient of nitrate through soil is relatively slow at 0.05–1.2 cm² d⁻¹ (Hagin & Tucker 1982). Instead, the lack of any detectable soil inorganic ¹⁵N recovery in zone C infers that pasture ¹⁵N uptake in zone C was potentially due to pasture roots translocating ¹⁵N from zones A and/or B. The main root axis of perennial ryegrass plants can exceed 40 cm (Robin et al. 2010) and thus plants in zone C could have readily accessed inorganic ¹⁵N in either zones A or B.

As ¹⁵N was recovered only in the pasture and soil fractions, a ¹⁵N mass balance was not the focus of this study; however, the potential fate of the unaccounted for ¹⁵N could vary widely. The field trial was set up in late spring to minimise N leaching risk; however, very high rainfall in December 2011 and January 2012 resulted in drainage, so it is very likely that urinary ¹⁵N was leached as -¹⁵N. This rainfall may have also temporarily increased denitrification and the subsequent loss of ¹⁵N as nitrous oxide (N₂O-¹⁵N). Volatilisation of ¹⁵N as NH₃ was unlikely to be significant because all treatments were washed in with 10 mm water (Black et al. 1987) but this pathway cannot be ruled out.

The increase in pasture N uptake observed in zone A only of the non-urine treatments suggests the water applied to all treatments to prevent NH₃ emissions played a role, as this was the only difference in treatment between the zones. We suggest the antecedent soil moisture from the water applied to zone A on day 0, coupled with 12 mm rain on day 47 was enough to result in a flush of mineralisation and rapid plant uptake in zone A. The drier antecedent soil conditions in zones B and C resulted in a slower response to the rainfall that was observed the following harvest (Figure 5). Under the conditions of this study, the effective area of the urine patch, based on total urinary-¹⁵N recovery, may have been up to 2.01 m² (0.28 + 0.67 + 1.06 m²) or seven times the wetted area. However, as the results demonstrate, recovery of urinary-¹⁵N was not constant with distance from zone A. If we assume a total pasture ¹⁵N recovery of 51% with 30%, 17% and 4% from zones A, B and C, respectively, (Table 1) and express these pasture recoveries as rates of recovery on a per unit area basis (0.28, 0.67 and 1.06 m² for zones A, B and C, respectively) then pasture ¹⁵N recovery rates equate to 107, 25 and 3.8 % m⁻². Thus while pasture in zone C is responsible for recovering some of the ¹⁵N it is far less efficient than pasture in zones A and B. Thus, the effective area is perhaps best described by the sum of zones A and B (0.95 m²), which in total is still 3.4 times the size of zone A. Such an effective area falls within the upper end of the range previously reported (Lotero et al. 1966; Moir et al. 2011).

Urinary N uptake from outside the ‘wetted area’, as demonstrated in this study, may have implications for the interpretation of data from lysimeter studies. Ruminant urine is generally applied to the entire surface of a lysimeter (Silva et al. 1999, 2005; Di & Cameron 2002; Di et al. 2002; Menneer et al. 2008; McDowell & Houlbrooke 2009; Ledgard et al. 2014; Shepherd et al. 2014; Buckthought et al. 2015) but the lysimeter casing creates a physical barrier between the plant root systems inside and outside the lysimeter, which is not representative of field conditions. Consequently, the ‘effective area’ in such lysimeter studies corresponds to the wetted area. The lack of a peripheral ‘zone B’ in these lysimeter studies may result in enhanced leaching or gaseous losses if it is assumed that pasture N uptake in the wetted area is constant regardless of whether a peripheral zone B is present or not. Conversely, the pasture roots confined within the lysimeter casing may be exposed to higher amounts of available N than in a field situation where subsurface lateral flow of urine may occur. This might result in higher dry matter yields and higher N uptake in lysimeters compared with the field conditions, and reduced losses.

Implications of these results on lysimeter studies will be influenced by many factors, not least the season of urine deposition and climatic conditions. Therefore, further research is clearly needed to determine and compare seasonal urinary-N losses and plant uptake dynamics from lysimeters and in the field, with and without a peripheral zone present.

Conclusion

The greatest recovery of urinary ¹⁵N occurred in the pasture fraction (51%). Most of this was from within the urine-wetted area (zone A); however, 33% of the total ¹⁵N recovered in pasture was from outside the wetted area. The effective area of pasture partaking in urinary-¹⁵N uptake was therefore at least 3.4 times the wetted area. Soil inorganic-N dynamics showed urinary-¹⁵N present outside the initially wetted area within the first 15 days of the experiment. However, the mechanism for ¹⁵N recovery by pasture outside the wetted area was likely to be due to plant root uptake from outside the wetted area, translocating ¹⁵N to the unwetted pasture zone. These results have implications for understanding the fate of N under a urine patch and the extent of urinary-N uptake by pasture, both spatially and temporally, and the methodologies used in obtaining such information. Further research is required to determine the fate of urinary-N from lysimeters with and without a peripheral zone present.

Supplementary data

Figure S1. A, Rainfall over the experimental period; B, air and soil temperatures (10 cm depth) over the experimental period.

Acknowledgements

The statistical analyses were carried out by Dr John Waller, formerly of AgResearch Ltd. All ¹⁵N analyses were carried out by Roger Cresswell at Lincoln University.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was funded by the Pastoral 21 programme, a collaborative venture between DairyNZ, Fonterra, Dairy Companies Association of New Zealand, Beef + Lamb NZ and the Ministry of Business, Innovation and Employment.