Characterising the volumes and chemical characteristics of dairy sheep effluent in New Zealand

ABSTRACT

Dairy sheep produce an effluent stream from the milking parlour that is either applied directly to land via a sump and spray irrigator, or is stored in an effluent pond before application. In order to characterise this effluent stream, a monitoring programme was undertaken with effluent samples collected from two case study sheep milking farms over two lactation seasons (2014/2015 and 2015/2016), and a third over one lactation period (2015/2016). Typical effluent volumes generated ranged between 4 and 10 L ewe⁻¹day⁻¹ for the 130–200-day lactation. The mean physical and chemical attributes of effluent samples generated by the milking parlours were: 0.54% dry matter (DM), 220 g nitrogen (N) m⁻³, 32 g phosphorus (P) m⁻³, 150 g potassium (K) m⁻³ and 22 g sulphur (S) m⁻³. The mean nutrient concentrations of dairy sheep effluent are lower than values reported for dairy goat (P & S only) and dairy cow effluents (N, P, K and S).

KEYWORDS:

• Dairy sheep
• farm dairy effluent
• nitrogen
• phosphorus
• potassium
• E. coli

Introduction

Traditionally, New Zealand's sheep industry has been based on meat and fibre and is world renowned for producing high-quality meat and wool. However, there is growing interest in alternative income streams due to the continuing relatively low profitability of these dry-stock farm systems. This trend is reflected in a small but expanding dairy sheep industry in New Zealand. At present, there are three sheep milk processors operating at scale, and a number of smaller operators that make niche specialist sheep milk products (Peterson and Prichard 2015). Internationally, sheep milk currently has an estimated 1.4% share of the global dairy market, and accounts for 3.9% of milk production in South-East Asia, more than 4% in China and 7.5% in North Africa and the Middle East (FAO 2013).

As with dairy cows, dairy sheep produce an effluent stream associated with washing of the milking parlour. This effluent is either applied directly to land via spray irrigation, or is stored in an effluent pond for later application to land. Research has shown that farm dairy effluent (FDE) from dairy cow systems has contributed to losses of nitrogen (N), phosphorus (P) and faecal bacteria to water (Wang et al. 2004; Houlbrooke, Horne, Hedley, Snow, et al. 2004). A considerable investment in research and development has helped to identify appropriate management guidelines for safely returning cow FDE to land to limit such losses (Houlbrooke, Horne, Hedley, Hanly, et al. 2004; Houlbrooke and Monaghan 2010; Monaghan et al. 2010; Li et al. 2015; Luo et al. 2017). In order for the dairy sheep industry to similarly manage and therefore limit its environmental impact while maximising the value of sheep FDE, quantitative information is required on the volumes and nutrient contents of sheep effluent generated at the milking parlour. Similar information is also required for effluent produced from other farm facilities, such as wintering barns.

In this paper, we document summary results from an effluent monitoring programme undertaken over two lactation seasons (2014/2015 and 2015/2016) from three sheep milking farms in New Zealand.

Methods

Details of monitored farms

The first of the three monitoring farms (Farm A) is located in Southland. This farm is 203 ha in size and has the milking parlour and a wintering barn combined under one roof. The floor of the wintering barn was plastic grating that allowed dung and urine excreta to drop below to a shallow bunker, where it was collected via a scraper moving on a continuous chain. The ewes usually enter the barn in mid-May and remain there until late-August. However, in drier periods when pasture production is low, sheep are also housed and fed in the barn to maintain milk production. The numbers of milking ewes reach a peak of c. 3500 in December with milking occurring for 6–8 months of the year. Effluent generated from the milking parlour is washed into a stone trap, which drains into a sump. This effluent is then pumped from the sump to the end of the wintering barn, where it is used to liquefy any excreta from the wintering barn, allowing it to be pumped into an effluent pond. Extra water is used occasionally to dilute the high dung content of the barn excreta so that it can be easily spread to pasture. In addition, sprinklers are required in the wintering barn so that the grating can be regularly washed and kept clean. Of the three farms monitored in this study, this one was the only one with a wintering barn.

The second farm (Farm B) is 500 ha in size and is located on the south-western side of Lake Taupo. No milking took place during the 2014/2015 season. Milking did re-commence for the 2015/2016 season, with c. 2900 ewes being milked over a 130-day lactation from October to February. Initially, effluent from the milking parlour and holding yard went to a 30 m³ sump before being land-applied via a slurry tanker. For the 2015/2016 season a change to application via travelling irrigator was made, with a tractor power take-off being used to pump effluent from the sump to the irrigator.

The third farm monitored (Farm C) was an 11.5 ha property running 200 sheep on the outskirts of Masterton, Wairarapa. In both years that were monitored, milking commenced in mid-October and continued for 210 days. Effluent from the dairy shed was washed through a sand trap to a small 2.7 m³ capacity sump, with the maximum effluent storage capacity being equivalent to the volume of FDE generated from two days of milking. A submersible pump was used to transfer effluent via a 50 mm pipe to a stationary pot spreader which was moved periodically to cover 25% of the farm in any given year.

Monitoring effluent volumes

The volume of water used at Farm A was monitored monthly by reading a flow meter installed in the pump room of the milking parlour. Water used in the wintering barn originated from a different, unmetered source so was unable to be monitored. The volumes of effluent produced were monitored by means of two Euromag flow meters placed in the pipes from both the milking parlour effluent sump and the adjacent wintering barn effluent sump. The excreta volumes recorded by these meters were read at approximately monthly intervals over a 17-month period from May 2014 to September 2015.

At Farm B, estimates of effluent depth and hence total volumes from a given application event were derived from measurements of effluent collected in trays laid out on pasture at right angles to the direction of the travelling irrigator on four occasions from November 2015 until February 2016. In addition, the time taken and distance travelled by the travelling irrigator for its daily run were also recorded. As all effluent generated was applied each day, these measurements allowed for the calculation of daily effluent volumes. Similarly, at Farm C, estimates of effluent volumes were determined from effluent collected in trays laid out on pasture at 0.5 m intervals from the pot spreader in February 2014, 2015 and 2016. Once again the time for each application was recorded, allowing for volumes to be calculated.

Chemical composition

Effluent samples for determination of chemical composition were collected monthly from both sumps at Farm A. As the wintering barn sump contained effluent from both the milking parlour and the wintering barn, the chemical concentration of effluent from the wintering barn alone was calculated as the difference in nutrient load between the two sumps divided by the difference in effluent volume. On several occasions the wintering barn sump was either empty, or had a solid surface layer of dung that prevented sample collection. At these times, a sample was instead collected from the effluent pond adjacent to the wintering barn sump while the effluent was being agitated by a pond stirrer. Samples were collected by immersing a 1 L bottle in the sump or pond and allowing it to fill with effluent. A sub-sample for Escherichia coli (E. coli) analysis (20 mL) was taken before the bottle was sealed.

At Farms B and C, representative 1 L sub-samples of the effluent collected in the measurement trays were retained for chemical analysis. A total of six, monthly samples were collected at Farm B during 2015/2016; no milking occurred during 2014/2015 hence sampling was not possible. Additionally, effluent samples were collected at monthly intervals during the lactation period at Farm C (five in 2014/2015 and six during 2015/2016) and frozen for later analysis; because these samples had been frozen, E. coli concentrations could not be determined.

All samples were sent to commercial laboratories for chemical analysis. For Farm A, E. coli concentrations were measured at AgResearch Invermay using the Colilert^® media and the Quanti-Tray^® system (IDEXX laboratories, Maine, USA; Bucklaw et al. 2006). For Farm B, E. coli was determined as colony forming units (cfu; APHA 1998, method 9222 D).

Results and discussion

Volumes produced

There was greater variation in daily water use for the milking parlour at Farm A than in the effluent originating from the milking parlour (Table 1). This reflects the fact that the water is also used for cleaning the vats, etc. and as such not all the incoming water passes through the effluent system. Mean daily average water use in the milking parlour was 9.9 L ewe⁻¹ day⁻¹, while effluent generated amounted to 9.6 L ewe⁻¹ day⁻¹, with the difference being attributed to the water used for vat cleaning, etc. Additional effluent was generated from the wintering barn when it was in use, producing on average an extra 6.5 L ewe⁻¹ day⁻¹. This additional effluent from the wintering barn includes the water used for washing down the gratings, as well as the excreta from the housed animals. The volumes of effluent for Farms B and C, as calculated from the effluent application measurements, ranged between 3–5 and 6–7 L ewe⁻¹ day⁻¹, respectively. When converted to an equivalent stock unit (SU) basis it would appear that the sheep dairy industry effluent production of 6.9 L effluent SU⁻¹ (range 4.2–9.6 L SU⁻¹) is similar to that for the dairy cow industry of 5.9 L effluent SU⁻¹ (Heubeck et al. 2014) (ewe = 1 SU, Friesian cow = 8.5 SU; Beef and Lamb NZ 2016) and certainly within the range of effluent volumes often reported for that industry (2.4–11.9 L SU⁻¹).

Table 1. Recorded volumes of water used in the milking parlour at Farm A plus effluent volumes generated by the milking parlours for the three farms monitored and that generated by the wintering barn for Farm A (L ewe⁻¹ day⁻¹). Note that for Farm A the measurements recorded were cumulative flows.

	No of occasions data recorded	Mean	Median	Range
Farm A
Milking parlour water use	14	9.9	7.5	4–28
Milking parlour effluent	12	9.6	9.7	5–17
Wintering barn effluent	6	6.5	5.0	2–13
Farm B
Milking parlour effluent	4	4.2	4.0	3–5
Farm C
Milking parlour effluent	3	6.9	6.8	6–7

Chemical and microbial constituents

Nutrient and E. coli concentrations in effluent from the milking parlours for the three case study farms, and the wintering barn at Farm A, are presented in Table 2. Nitrogen concentrations in effluent from the milking parlours ranged from 130 to 608 g m⁻³, P concentrations from 24 to 49 g m⁻³, K concentrations from 127 to 201 g m⁻³, S from 16 to 37 g m⁻³ and organic carbon (C) from 698 to 4254 g m⁻³. The dry matter (DM) and nutrient concentrations in effluent from the milking parlour at Farm A were lower when the sheep were housed and fed grass as well as supplements such as grain, dairy pellets, lucerne hay or silage, than when they were solely grass fed. It is suspected that this is a result of more of the excreta being deposited in the wintering barn and less in the milking parlour. This hypothesis is supported by the DM content and nutrient concentrations from the wintering barn sump being an order of magnitude higher than those measured at the milking parlour sump for those occasions when sheep were being housed. While total P concentrations in effluent were higher from the wintering barn (434 g m⁻³) than from the milking parlour (49 g m⁻³), dissolved reactive P (DRP) concentrations were similar for both the milking parlour and wintering barn effluents (Table 2). The mean P concentration in effluent from the effluent pond (94 g m⁻³) was a reflection of the combined effluent sources. Mineral-N analysis indicates that c. 40% (range = 13–56%) of the total N content was in a plant-available form, mainly ammonium-N. The dissolved reactive P analysis indicated that c. 48% (range = 31–71%) of the total P in effluent generated from the milking parlour was in a plant-available form. The E. coli concentrations in effluent were broadly similar for all farms, despite the different methods used, ranging from 2.0 × 10⁷ cfu 100 mL⁻¹ to 9.67 × 10⁷ MPN 100 mL⁻¹. Organic C, N and E. coli concentrations in effluent from the wintering barn were higher than those in effluent from the milking parlour, reflecting the greater faecal content of the barn effluent. One very high E. coli sample collected from the effluent pond at Farm A skewed the results so that the pond E. coli concentrations were an order of magnitude higher than those in the effluent from the milking parlour.

Table 2. Average (median values in brackets) DM contents and, nutrient and E. coli concentrations in effluents derived from the milking parlour, wintering barn and effluents pond at Farm A, and for land-applied effluent at Farms B and C.

Nutrient analysed	Farm A				Farm B	Farm C
	Milking parlour		Wintering barn	Effluent pond
	Pasture fed	Pasture + Supplement^a
Number of sampling occasions	6	12	9	4	6	13
Sample location	Parlour sump	Parlour sump	Barn sump	Effluent pond	Paddock	Paddock
DM%	1.48 (1.23)	0.22 (0.22)	6.9 (5.9)	0.68 (0.63)	0.43 (0.34)	0.31 (0.29)
Dissolved reactive P (g m^−3)	23 (25)	10 (10)	27 (31)	30 (34)	9 (7)	23 (24)
Total P (g m^−3)	49 (54)	24 (24)	434 (559)	94 (98)	29 (31)	32 (30)
K (g m^−3)	201 (212)	127 (109)	2582 (3296)	694 (683)	166 (175)	137 (140)
NH4-N (g m^−3)^b	77 (87)	51 (36)	675 (721)	228 (222)	59 (54)	78 (80)
NO3-N (g m^−3)^c	0.40 (0.29)	0.58 (0.40)	1.53 (0.38)	0.36 (0.31)	No data	<0.1 (<0.1)
Total N (g m^−3)	608 (437)	130 (134)	1604 (1279)	428 (427)	200 (191)	139 (137)
S (g m^−3)	37 (23)	23 (19)	300 (433)	40 (44)	17 (19)	16 (13)
Ca (g m^−3)	117 (95)	60 (56)	997 (887)	174 (174)	169 (147)	64 (57)
Mg (g m^−3)	45 (36)	24 (22)	381 (559)	95 (96)	20 (22)	23 (20)
Na (g m^−3)	55 (47)	53 (48)	465 (547)	135 (139)	23 (22)	53 (43)
Organic C (g m^−3)	4254 (3100)	994 (686)	20135 (25297)	2232 (2117)	No data	698 (690)
E. coli (MPN^d 100 mL^−1)	9.67 × 10^7 (9.23 × 10^7)	3.04 × 10^7 (2.14 × 10^7)	1.40 × 10^8 (1.13 × 10^8)	3.08 × 10^8 (8.9 × 10^6)	2.01 × 10^7 (1.28 × 10^7)	No data

Note: Note that units for E. coli from Farm B are cfu 100 mL⁻¹.

^aSupplements included grain, dairy pellets, lucerne hay and silage;

^bNH₄-N = ammonium-N;

^cNO₃-N = nitrate-N;

^dMPN = Most probable number of viable cells.

Monitoring showed that effluent from dairy sheep farms has a high content of N and K, although for the three farms sampled in this study concentrations were only half those typically found in dairy cow effluent (Table 3). The concentrations of P and S found in dairy sheep effluent were also considerably lower than reported for either dairy goat or dairy cow effluents (Carlson et al. 2010; Longhurst et al. 2017). With the concentrations of nutrients in sheep dairy effluent being lower than for other dairy milking animals, the estimated value of the nutrients (May 2016 values) on a volume basis is also lower at $0.58 m⁻³, compared to $0.76 and $1.15 m⁻³ for dairy goats and dairy cows, respectively.

Table 3. Nutrient concentrations, DM contents and equivalent fertiliser nutrient dollar values of milking parlour effluents from various dairy sheep, dairy goat and dairy cow systems in New Zealand.

	Sheep	Goat^a	Cow^b
Sample size	37^d	11
Solids (DM%)	0.54	1.40	1.10
Nitrogen (g m^−3)	220	210	350
Phosphorus (g m^−3)	32	60	60
Potassium (g m^−3)	150	150	380
Sulphur (g m^−3)	22	140	50
Nutrient value^c ($ m^−3)	0.58	0.76	1.15

^aFrom: Carlson et al. (2010).

^bFrom: Hawke and Summers (2003); Longhurst et al. (2000, 2014); Monaghan and Smith (2004); Wang et al. (2004); Houlbrooke, Horne, Hedley, Hanly, et al. (2004); Longhurst and Nicholson (2011).

^cEquivalent fertiliser value assuming $1.10 kg⁻¹ N; $3.67 kg⁻¹ P; $1.33 kg⁻¹ K; $0.77 kg⁻¹ S. Taken from Ballance Agri-Nutrients price list May 2016 (http://www.ballance.co.nz/Our-Products/PriceListing).

^d37 samples collected from three farms.

Our results indicate a relatively low DM content of dairy sheep effluent from the milking parlour (Table 2). We suggest this is a result of the dung of the sheep being in the form of small hard pellets, which tended to float on top of the effluent in the sumps, and ponds. The sample collection method used at Farm A (i.e. grab sampling) is likely to under represent the solids layer at the surface of the pond/sump. There is the possibility that this could have resulted in an underestimation of some nutrient concentrations (e.g. P). Previously, it has noted that there was a highly significant relationship between N and DM in dairy effluent, with most of the N being in organic form associated with the DM (Kaufler and Corban 2009; Longhurst et al. 2017). Our results followed a similar pattern (not shown) with organic N accounting for between 44% and 88% of total N (Table 2). While this may explain some of the differences in total N at Farm A, it must be noted that the samples collected at Farms B and C were collected from the irrigators and as such represented the effluent that was being applied to land.

A physical point of difference between sheep dairy effluent and cow dairy effluent was the presence of wool contaminating the former effluent stream. While this was not noticeable on observation, it was apparent when viewed under a microscope. It was not possible to quantify the level of this contamination except to note that it was more prevalent when the sheep were being housed in the wintering barn. While this is not a problem with the effluent itself, it is a problem with the pumps, pipes and sprinklers that convey the effluent, potentially causing blockages of pipes and nozzles and increasing wear and tear on pumps. Similar problems have been experienced in the dairy goat industry.

Conclusions

While dairy sheep farms can produce similar volumes of FDE to dairy cow farms on an SU basis, monitoring of three case study farms has shown that mean nutrient concentrations are lower in sheep FDE than reported for N, P, K and S concentrations in effluent from cow dairy farms and P and S concentrations in effluent from goat dairy farms. Despite this, sheep FDE is still a valuable source of nutrients if applied to land, representing a current equivalent fertiliser N, P, K and S value of $0.58 m⁻³. Dairy sheep effluent is different to dairy cow effluent in that it contains wool fibres, which can cause problems with the pumps, pipes and sprinklers that transport the effluent.

Acknowledgements

We would also like to thank the owners of the three farms monitored and their staff for allowing us on their properties and occasional assistance.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors would like to thank the Ministry of Business, Innovation and Employment (MBIE) for funding this study (Contract No. C10X1305), together with three major New Zealand sheep milk producers (Antara Ag Farms, Maui Milk and Kingsmeade Cheese Limited).