Evaluating the benefits of standing cows off pasture to avoid soil pugging damage in two dairy farming regions of New Zealand

Transcription

1 New Zealand Journal of Agricultural Research ISSN: (Print) (Online) Journal homepage: Evaluating the benefits of standing cows off pasture to avoid soil pugging damage in two dairy farming regions of New Zealand PC Beukes, AJ Romera, DA Clark, DE Dalley, MJ Hedley, DJ Horne, RM Monaghan & S Laurenson To cite this article: PC Beukes, AJ Romera, DA Clark, DE Dalley, MJ Hedley, DJ Horne, RM Monaghan & S Laurenson (2013) Evaluating the benefits of standing cows off pasture to avoid soil pugging damage in two dairy farming regions of New Zealand, New Zealand Journal of Agricultural Research, 56:3, , DOI: / To link to this article: Published online: 24 Jul Submit your article to this journal Article views: 716 View related articles Citing articles: 9 View citing articles Full Terms & Conditions of access and use can be found at

2 New Zealand Journal of Agricultural Research, 2013 Vol. 56, No. 3, , RESEARCH ARTICLE Evaluating the benefits of standing cows off pasture to avoid soil pugging damage in two dairy farming regions of New Zealand PC Beukes a *, AJ Romera a, DA Clark a, DE Dalley b, MJ Hedley c, DJ Horne c, RM Monaghan d and S Laurenson d a DairyNZ, Hamilton, New Zealand; b DairyNZ, Lincoln University, New Zealand; c Massey University, Palmerston North, New Zealand; d AgResearch, Invermay Agricultural Centre, Mosgiel, New Zealand (Received 23 October 2012; accepted 1 July 2013) Many dairy farms in the Manawatu and Southland regions of New Zealand are situated on poorly drained soils that are prone to treading damage an undesirable occurrence on grazed pastures during the wetter months of the year. Standing cows on a loafing pad (stand-off) during wet conditions can reduce damage, but will incur costs. The objective of this study was to evaluate the trade-offs of grazing duration (from 0 to 10 hours grazing/day for lactating cows) versus pasture damage (wastage), feed and stand-off expenses, and farm operating profit. A simulated farm from each region was used in a farm systems model. This model simulates pasture-cow-management interactions, with site-specific climate data as input for the soilpasture sub-models. Days to recover previous yield potential for damaged paddocks can vary widely. A sensitivity analysis (40 to 200 days to recover) was conducted to evaluate the effect of this parameter on results. Full protection when there is risk of damage (0 grazing hours/day) appeared to be less profitable compared with some level of grazing, because advantages of reduced damage were outweighed by disadvantages of managing infrequently grazed pastures. The difference in operating profit between full protection and some level of grazing became less as the recovery time increased, but for both regions, grazing durations of 68 hours/day when risk of damage is present appeared to be a sensible option irrespective of recovery time. Keywords: simulation modelling; treading damage; loafing area; pasture utilization; profitability Introduction An important challenge facing dairy production in New Zealand is to develop production systems that can achieve increased production and profitability, with lower environmental impact, especially nitrogen (N) leaching. To address this objective, a large national set of farm systems trials (Pastoral 21 Phase 2) has been initiated with the aim of designing future farm systems for four dairy regions. Designs for future systems were developed for each region using strategic changes to the current system, one of which included installation of stand-off facilities (loafing pad for standing cows off pasture) for the Manawatu and Southland regions. Modelling was used to help fine-tune proposed strategies within the new farm design. The main objective of the stand-off strategy was to reduce treading damage, but also to reduce the deposition of urinary nitrogen onto nearsaturated soils, which in turn will reduce nitrogen losses via nitrous oxide emissions and nitrate leaching (de Klein et al. 2010). Decisions on when to implement stand-off were addressed in terms of hours to stand-off and, therefore, hours to graze per day. This study focuses on using the Manawatu and Southland future scenarios for evaluating the trade-offs between treading *Corresponding author. pierre.beukes@dairynz.co.nz # 2013 The Royal Society of New Zealand

3 Evaluating the benefits of standing cows off pasture to avoid soil damage 225 damage, pasture production and financial implications of different stand-off policies, using the current (baseline) scenarios as reference points. The potential effects of stand-off on nitrogen losses were not evaluated. Treading under animal grazing can have detrimental effects on soil physical quality by causing compaction or pugging (Houlbrooke et al. 2008). The extent and nature of soil structural damage is largely dependent on the soil moisture content at the time of grazing (Houlbrooke et al. 2011). Pugging occurs at high soil moisture content when air-pores are filled with water and soil has reached a point of plasticity i.e. plastic limit (Houlbrooke et al. 2011). In this state, pressure from the cow s hoof causes a deformation and remoulding of the soil resulting in an undulation of the soil surface. In the process, vegetation is buried and roots exposed and damaged by the impact of the hooves, which in turn reduces pasture utilization (Johnson et al. 1993). In comparison, soil compaction results in a reduction or compression of soil pore space, and occurs at lower soil moisture contents than pugging (Piwowarczyk et al. 2011). Decline in pasture growth potential is caused by a reduction in soil aeration and increased soil density, however this tends to be a gradual process that occurs over a period of years (Drewry et al. 2008). Both processes of soil damage can occur concurrently, however the potential for soil compaction generally decreases with greater water content, whereas the risk of soil pugging increases (Drewry et al. 2008; Houlbrooke et al. 2011). Treading damage is prevalent on many dairy farms in the Manawatu and Southland regions of New Zealand (Drewry et al. 2000) and is regarded by many farmers as inevitable. In a review describing the natural recovery of soils affected by both compaction and pugging damage, Drewry (2006) states recovery periods ranging between a period of weeks to, in some cases, years. This variation largely reflects the extent to which subsequent grazing events coincide with high soil moisture. In practice, the frequency with which stock is restricted from paddocks with the intention of protecting soil structure needs to be balanced against other concerns. Standing cows on a loafing area (standoff) during wet conditions, for instance, raises capital and maintenance costs, may require feed supplementation to ensure adequate intakes, and may result in increases in average herbage mass with a consequence of depressed net pasture growth rates because of greater losses through senescence (Chapman & Lemaire 1993). Because compaction occurs at lower soil moisture and is most prevalent in the subsurface soil (510 cm), it is difficult to observe. On the other hand, wet soils (close to or at saturation) are readily identified and can be used as decision support to reduce pugging damage. The objective of this study was to focus specifically on pugging damage, and to use a farm systems model, representing pasture-cow-management interactions, to evaluate the impacts of different levels of protection (from 0 to 10 hours grazing time/day for lactating cows) on pasture yield, damage and wastage, feed and stand-off expenses, and farm operating profit. Methods The DairyNZ Whole Farm Model and modifications The DairyNZ Whole Farm Model (WFM) (Beukes et al. 2008) was developed to assist with analysis and design of dairy farm system experiments through scenario testing under various system interactions that occur over multiple years. The model framework written in VisualAge Smalltalk (IBM) represents a pasture-based dairy farm with individual paddocks and cows simulated on a daily time step. Animals and paddocks are represented by a copy (or instance) of the relevant sub-model initialized for each. For example, the age, breed and other characteristics that are unique to an individual are used to initialize each cow instance, while for each paddock, the herbage mass and soil characteristics are specified. Molly is a mechanistic and dynamic model in the WFM that simulates critical elements of cow digestion and metabolism (Hanigan et al. 2009). The cow s

4 226 PC Beukes et al. production is influenced by the quantity and quality of feed, and capacity to absorb and convert nutrients into milk (i.e. genetic merit). Molly s feed intake is driven by metabolic demand. The metabolic energy content of the feed is a product of digestion and absorption of feed fractions, which are specified by the model user. Molly predicts enteric methane production, urinary and faecal N excretion, milksolids (MS fatprotein) production and milk urea N concentration. The pasture-soil model in WFM (Romera et al. 2009) is climate-driven using weather data provided by the National Institute of Water and Atmospheric Research (NIWA) from the nearest weather station, or using interpolations from the nearest weather stations (virtual climate) for a particular location (Tait et al. 2006). A standard soil water balance is used to determine actual soil water content. The water balance is modelled for two soil horizons. Water enters the top (A) horizon following rain until it is fully recharged. It then drains into the lower (B) horizon. Surplus rainwater is considered to drain through the profile or run off the soil surface. Water is also lost from the soil through evapotranspiration, which is a function of potential evapotranspiration and available soil water. Available soil water holding capacity, daily rainfall and daily potential evapotranspiration are all inputs to the model. Irrigation water is treated in the same way as rainfall (Romera et al. 2010). Some or all paddocks can be irrigated according to a user-defined irrigation policy. Pasture response to irrigation water is determined by soil moisture levels at the time of irrigation. Pasture growth responds to N applied as either mineral fertilizer or irrigated effluent. Paddocks are grazed rotationally and a particular herd of cows may take several days to graze all the breaks in a particular paddock depending on rotation length at the time. Post-grazing herbage mass (residual) is determined by the model as a function of the feed demand of the herd, grazing hours and the herbage allowance on that day. The WFM is coded to allow an average pasture intake of 2 kg dry matter (DM)/cow/hour following the results of Gregorini et al. (2009). The residual herbage mass of a paddock influences pasture regrowth of the paddock. Paddocks can be eliminated from the grazing rotation for all or part of the year as part of a cropping regime e.g. maize, cereal or brassica crops. Supplements (home grown or imported) can be fed to cows according to policies created by the user. Other user-defined policies related to cow management include breeding, grazing off the farm, stand-off, drying off, culling and replacement. Two modifications had to be made to the WFM code for this study. The first was to implement a pugging damage module representing the loss in pasture regrowth potential when cows are allowed to graze a paddock when soil moisture exceeds a certain threshold. In the model, this threshold is a user-defined moisture percentage below which there is no damage (assuming the focus in this study is pugging and not compaction damage), but at or above this threshold the loss of pasture growth potential occurred and was a function of stocking density (animals/ha for a particular break) and grazing duration (hours/ day). This pasture loss function was coded following the work by Betteridge et al. (2003) who produced a series of curves showing a small effect on percentage loss when grazing duration was short (23 hours/day) and stocking density changed from 100 to 400 animals/ha. Also, at a low stocking density (25 animals/ha) changes in grazing duration showed a small effect on percentage loss. However, at a stocking density of 270 animals/ha (equivalent to a stocking rate of 3 cows/ha on a 90- day rotation length in the winter months, a typical situation on many New Zealand farms) changing grazing duration from 2 to 12 hours/day can result in a large change in percentage loss (from 15% to 60%; Fig. 1). In the model, stocking density was known from the number of lactating or nonlactating cows in any particular herd (the milking herd or the dry herd) and the rotation length giving the break size on any particular day (break size total pasture area in ha/rotation length). Time on pasture or grazing duration is determined by user-defined settings for time in the dairy parlour (default 4 hours for twice-a-day milking),timeontheracesorlaneways(default

5 Evaluating the benefits of standing cows off pasture to avoid soil damage 227 Figure 1 Percentage loss in pasture regrowth potential in relation to stocking intensity and grazing duration. Each curve represents the percentage of potential pasture growth that will be lost during the month following grazing, at any given combination of stocking density and grazing duration when soils are liable to be damaged (taken with permission from Betteridge et al. 2003). 1 hour for twice-a-day milking), time on the feed pad (concrete surface with feed bins; user-defined in the supplementary feeding policy and dependent on type and amountsof supplements fed) and time on the stand-off pad or loafing area (userdefined in the stand-off policy; this can vary from 19 hours stand-off for no pasture grazing to 0 hours stand-off for 19 hours on pasture for lactating cows). In the WFM, each paddock is represented by one instance of the pasture-soil model giving a single pasture growth rate per day for the whole paddock. This demanded a simplification in applying the percentage loss, because only part of a paddock may be damaged during times when break sizes are smaller than paddock sizes. The simplification was to average the percentage loss over the whole paddock even though some parts may not be damaged. Recovery (the percentage loss diminishes over time) of pugged pastures was shown in a Waikato case study by Betteridge et al. (2003) where a dairy pasture fully recovered after 55 days following severe damage that reduced initial pasture regrowth potential by 50%. The WFM was coded to accept a user-defined recovery time (days), during which the initial percentage loss diminished linearly to zero. A further simplification was required for when pugging damage compounds on the same paddock due to two consecutive pugging events. The solution was to integrate the total percentage loss left from the first pugging event (area under the curve: percentage loss versus days left to recover) when the second event occurs, and also integrate the total percentage loss from the second pugging event, find the total and apply the appropriate losses to the paddock. The second modification was a wastage module that represents the increasing loss of herbage trodden to or below ground level with increasing soil moisture levels (Sheath & Boom 1997). Data from a historical report by D. Causley (unpublished, 1975) indicated that herbage wastage increases linearly up to field capacity at which point losses averaged 16%. Average wastage on saturated soils (exceeding field capacity) was 40%.

6 228 PC Beukes et al. This was implemented in the WFM by allowing no wastage up to 50% field capacity, then a linear increase in wastage up to 16% at field capacity. The pasture-soil model in WFM does not simulate soil moisture above field capacity but shows drainage when it rains on a soil already at field capacity. The drainage factor (5 mm increments) in the model was used to simulate increased wastage from 16% to 40% on rainy days when the soil was at field capacity. The wastage loss (%) was deducted in a ratio of 1:1 from the potential herbage intake of a herd of cows (kg DM), and from the potential post-grazing herbage mass under dry conditions (residual, kg DM/ha). Calculating the loss from potential intake is important because it makes wastage dependent on stocking density, in that more stock increases potential intake resulting in more wastage in absolute terms. The loss from potential residual is also important because it puts the pasture at a disadvantage when regrowth is compromised depending on how low the post-wastage residual is. In the model, wasted pasture (kg DM/ ha) was added to litter and disappears as a result of decay. The economics component in the WFM consists of a simplified profit and loss statement, balance sheet and return on assets. Revenue is primarily generated through milk solid (MS) sales, where MS production is multiplied by the price per kg MS. Additional revenue is earned from the sale of culled stock. A large proportion of costs weredefinedinanactivity-basedcostingframework with default values generated through the use of economic survey data. For this study, economic input data were updated from the DairyNZ Economics Group with the 2010/11 season s costs. The milk price for the 2010/11 season was fairly high at $7.36 kg/ms. Operating profit calculations were also performed with a more long-term average milk price of $6.00 kg/ms. Pugging threshold and stand-off days In the WFM, the number of days per year that a paddock is at risk of pugging was determined by the user-defined soil moisture level (% of field capacity), rainfall and evapotranspiration. Since the soil module in the WFM is a much simplified representation of the actual soils from the case study farms, it was necessary to determine how many days on average per year cows were removed from paddocks because of the pugging risk. Total days off paddock were used to calibrate the most appropriate pugging threshold settings in the model for each region. A daily soil water balance was used to explore the annual number of risk days for each region. In Southland, at a soil moisture deficit of 7 mm (7 mm below field capacity which corresponds to the plasticity limit of the Pallic soil, which is the dominant soil of the Southland farm), cows will be stood off pastures for 104 (920 SD) days per year (based on a 27-year average), including the 70 days of winter (1 June to planned start of calving on 10 August). During winter, all non-lactating cows are removed from pastures of the milking platform as part of the wintering strategy using brassica forage crops. In the WFM, the pugging threshold for the Southland scenario was set to a soil moisture level of 95% (equivalent to 9 mm below field capacity), which resulted in the model predicting an average number of risk days of 100 (923 SD). In Manawatu, the number of days that cows would be removed from paddocks was approximately 50 days per year. This was imitated in the WFM by setting the pugging risk threshold to 99.72% (:0.5 mm below field capacity), giving a 10-year mean of days with pugging risk per year. In Manawatu, cow removal during winter is not as complete as in Southland and cows may still be on the milking platform through the winter, but supplemented when necessary. Development of the farm scenarios The potential future farm scenario for the Southland region (Southland Alternative, Table 1) differed from the baseline by having a lower stocking rate, a larger support block to provide the pasture silage required to feed the wintered cows, a different crop on the milking platform (Italian

7 Evaluating the benefits of standing cows off pasture to avoid soil damage 229 ryegrass instead of turnips) and by having standoff for non-lactating cows for 24 hours/day from 1 June to 9 August while feeding them pasture silage (instead of wintering on a swede crop as in the baseline scenario). For the rest of the year, stand-off for lactating and non-lactating cows was for 12 hours/day (which equates to approximately 7 hours on pasture for lactating cows if milking takes 4 hours for twice-a-day milking and time on the lane ways takes 1 hour) when pugging risk was present. Outside the winter period all cows were fed as much pasture as was available or as much as grazing time allowed. When there were pasture deficits, Italian ryegrass from the crop area was fed until it was depleted, then pasture silage followed. The same rotation length/break size policy was used for baseline and alternative scenarios. The same amount of N fertilizer was also used on non-effluent paddocks in the baseline and alternative scenarios (140 kg N/ha/y), to avoid any confounding effects of N fertilization on pasture yield results for different stand-off treatments. The same effluent application rules (when there is effluent in the pond, irrigate over the whole effluent block when soil moisture levels are at or below 75% of field capacity) were used in baseline and alternative scenarios, although the effluent block was larger in the alternative scenario (33% of the milking platform) compared with the baseline (27% of the milking platform), because the stand-off pad collects effluent that gets added Table 1 Physical input parameters describing the baseline and alternative scenarios in the Southland and Manawatu regions of New Zealand. Southland baseline Southland alternative Manawatu baseline Manawatu alternative Milking platform area (ha) Support block area (ha) Stocking rate (cows/ha) Planned start of calving 10 Aug 10 Aug 15 Jul 1 Jul N fertilisation (kg/ha/y) Crops grown on platform Turnips Italian ryegrass Turnips Turnips Crop proportion of platform (%) Supplements imported Pasture silage Pasture silage Pasture silage, maize silage, palm kernel expeller Wintering of non-lactating cows On a brassica crop on the support block On the stand-off pad on the platform Stand-off (hours day 1 ) Never 12 hours when pugging risk; 24 hours during winter (1 June to 9 Aug) Effluent irrigated proportion of milking platform (%) 1 Applied only to the non-effluent area. 50% of the herd grazed off from 1 May to 1 July, the rest grazed on the platform Never Pasture silage, maize silage, palm kernel expeller Grazed on the platform 12 hours every day even with no pugging risk; 24 hours when pugging risk

8 230 PC Beukes et al. to the effluent coming from the dairy parlour. The Southland baseline scenario was modelled with a pasture conservation/cutting policy that allowed for silage making when pasture surpluses occurred from 1 October to 1 April. In the model, conservation paddocks were cut when pasture herbage mass exceeded 4000 kg DM/ha and when soil moisture levels were below the pugging threshold, and when it was a rainfree day (meaning heavy machines could enter the paddock without causing damage). The same silage-making policy was applied in the alternative scenario, but a further cutting rule was implemented. Whenever the average farm herbage mass exceeded 3000 kg DM/ha, then all paddocks (not only conservation paddocks) over 4000 kg DM/ha were cut, if soil moisture and weather conditions allowed. This extra silagemaking rule was introduced in the alternative scenarios to avoid excessive average farm herbage mass, senescence and pasture wastage because of limited grazing time with stand-off. The potential future scenario for Manawatu (Manawatu Alternative, Table 1) differed from the baseline by having a higher stocking rate, a larger crop area, a larger effluent irrigated area, and having stand-off for all cows (lactating and non-lactating) for 12 hours/day irrespective of pugging risk, and 24 hours/day when pugging risk was present. In the baseline scenario, 50% of the non-lactating cows were sent off farm to a grazier from 1 May to 1 July, whereas in the alternative scenario, all the animals were grazed on the platform for the winter period. In both systems, turnips were grown for summer feeding and supplements (pasture silage, maize silage or palm kernel expeller) were available all year round to feed the cows to demand when pasture intake was insufficient. In the alternative systems, the restricted grazing time when cows were stood-off meant that the supplement feeding rules were triggered more often. For the Manawatu region, the rotation length/break size policy was the same in the baseline and alternative scenarios. The amount of N fertilizer on the non-effluent paddocks was 150 kg N/ha/y in both scenarios. The same effluent irrigation rules as in Southland were applied in both Manawatu scenarios, but effluent was applied over 100% of the farm in the alternative scenario compared with 25% in the baseline. For Manawatu, the conservation/cutting policy in the baseline scenario allowed for silage-making when there was surplus pasture from 1 September to 1 April. Conservation paddocks were cut when paddock herbage mass exceeded 4000 kg DM/ha and when the soil moisture level was below the pugging threshold and when it was a rain-free day. To avoid excessive herbage mass, senescence and pasture wastage in the stand-off scenarios, a more aggressive conservation policy was implemented; a light cut where as soon as a paddock was closed for conservation, it was cut. Also, a further cutting rule was implemented. Cutting occurred whenever a paddock s herbage mass exceeded 5000 kg DM/ha, regardless of the average farm herbage mass, and whenever soil moisture and weather conditions allowed. The same economic input (2010/11) was repeated for all simulated climate years to avoid the potentially confounding effects of variable economic inputs on treatment effects. Stand-off costs for both regions were calculated and factored into the operating profit results after the simulations. This was done because stand-off costs were divided into fixed (depreciation and interest depending on capital costs) and variable (maintenance, labour, extra insurance, effluent disposal, etc.) costs, of which the latter were calculated based on model predictions of usage (cowhours on stand-off pad). It was assumed that capital required to construct the standoff facilities was borrowed (at 8.5% interest rate) and depreciation of the facilities was calculated at 5% per annum (J. McPherson, pers. comm., DairyNZ Economics Group). Different stand-off facilities were anticipated for the two regions, resulting in quite different fixed and variable costs. For the Southland alternative scenario, the construction costs for the standoff pad and effluent application facilities were taken from Beukes et al. (2011) who obtained figures from actual Southland farms where these structures were erected, for example, a total of

9 Evaluating the benefits of standing cows off pasture to avoid soil damage 231 $150,000 for a stand-off facility plus effluent application for a 373-cow herd (average herd size for NZ), which comes to approximately $400/cow. This equates to an annual fixed cost for interest and depreciation of approximately $54/cow. For the Manawatu alternative scenario, a more expensive indoor facility was proposed that served as a stand-off pad with an estimated construction cost of $1500/cow, the average of a range of facilities ranging from $1000 to $2000/cow (Hadley 2009), which gave an annual fixed cost of $203/cow. Variable costs consisting of woodchip/bark/sand/ straw replacement, labour, pumping/scraping of effluent and repairs and maintenance will differ between the different types of stand-off facilities. For the Southland stand-off, a variable cost of $0.04/cow/hour for an average usage of 2200 hours/cow/y was assumed. For the more expensive indoor facility in Manawatu, a variable cost of $0.07/cow/hour for an average usage of 4000 hours/cow/y was assumed. Simulations and measurements The WFM was initialized for the baseline and alternative scenarios for the 2010/11 farming season (1 June 2010 to 31 May 2011) for the Southland farm (Telford Agricultural College dairy farm: 46817?50.21ƒS, ?43.02ƒE) and the Manawatu farm (Massey University No. 4 dairy farm: 40839?73.94ƒS, ?31.06ƒE). A series of factorial experiments were set up in the WFM for the alternative scenarios for both regions by stepwise altering the stand-off hours to achieve grazing times of 0, 2, 4, 6, 8 or 10 hours on pasture/day whenever there was a pugging risk on paddocks prior to grazing. Days to recover for pugged paddocks can vary widely (Singleton et al. 2000; Betteridge et al. 2003; Drewry 2006). A sensitivity analysis was conducted to evaluate the effect of recovery time on results. Time for a pugged paddock to recover fully (days) was altered from 40 to 200 with increments of 40 days, creating 65 factorial combinations. In an attempt to capture the effects of climate variability, each combination was run over three climate blocks (using historical climate data from NIWA) for 3 consecutive years each, , and , giving a total of 180 simulations. Similarly, runs were also conducted for the baseline scenarios by altering time for a pugged paddock to recover and the same climate blocks, giving a total of 30 simulations. Results from the first year of each 3-year simulation were discarded because it was regarded as a run-in year allowing soil moisture and pasture covers to stabilize. Model outputs were averaged over years 2 and 3 of the 3-year climate blocks and are presented as graphs with days for pasture growth rates to recover from pugging damage as the independent variable. Outputs included pasture yield, pasture damage and wastage, pasture silage made, milk production (MS/cow, MS/ha), feeding expenses, total stand-off hours, and operating profit. Operating profit was expressed as $/ha with income from milk sales, culls and calves, and other cash income considered (default for 2010/11 of $46/ ha, DairyNZ Economics Group). The expenses included in calculating operating profit are listed in Table 2. Assumptions Apart from the inherent assumptions in the WFM, some further simplifications were made. In the sensitivity analysis of recovery time, regrassing costs (Table 2) were kept the same, irrespective of changes in recovery time. This is a simplification because it can be expected that with more days to recover, regrassing costs will increase, assuming farmers will action a rapid recovery of a damaged paddock. However, it was difficult to justify any changes to regrassing costs due to the interaction between grazing time and recovery time, e.g. with low grazing time (long stand-off time) and long recovery time, regrassing costs may be lower compared with the same situation but with longer grazing time. Health costs were higher in the alternative scenarios for both regions because it can be expected that the incidence of mastitis and lameness will be higher when cows spend time on

10 232 PC Beukes et al. Table 2 Economic input data for the financial year 2010/11 used for the simulations. A price of $7.36/kg milksolids was used. Per cow costs ($/cow) Cost Supplements ($/t DM) Cost Wages 210 Grass silage 290 Unpaid labour 135 Making silage 140 Farm dairy 21 Transport silage 40 Electricity 36 Feeding out silage 40 Animal health 69 Palm kernel expeller 270 Breeding and herd improvement 51 Maize silage 340 Grazing-off adult cow($/cow/week) 20 Grazing-off yearling ($/cow/week) 10 Costs per ha ($/ha) (i.e. total cost is fixed) Overheads ($/ha) (i.e. total cost is fixed) Weed and pest control 35 Administration 113 Regrassing 64 Insurance 52 Vehicle fuel 186 ACC 31 Repair and maintenance 338 Rates 91 Freight expenses 49 Spray out pasture 76 Pre-emergence weed control 130 Cultivation 348 Sowing 145 Seeds 450 Seed insecticide treatment 147 Fertilisation and irrigation costs Urea ($/t) 711 Potash Super 20% ($/t) - Driven by outputs Fertiliser spreading ($/ha) 12 Irrigation ($/ML) 67 1 Maintenance fertiliser: 0.8 kg Potash Super 20%/kg MS. unnatural surfaces such as stand-off pads or indoor facilities (Stewart et al. 2002). This was implemented in the alternative scenarios by increasing health costs by 10% to $76/cow. In modelling the different stand-off policies for the alternative scenarios, neither the size of the effluent block nor the effluent irrigation policy were altered to better capture effluent from the stand-off pad with respect to time. Different hours on the pad will result in different amounts of N collected in the effluent pond and, therefore, a different amount of N deposited on the effluent block under a similar irrigation rule. Pasture yield predictions from the effluent block may have been affected by this simplification, but this was assumed to be negligible considering the relatively small amounts of N concerned. The variable cost of effluent treatment with different stand-off policies (e.g. electricity and maintenance) were accounted for by expressing the stand-off cost in $/cow/hour, resulting in higher variable costs with more usage of the stand-off facilities. Results For Southland, all the stand-off treatments resulted in higher annual pasture yields compared

11 Evaluating the benefits of standing cows off pasture to avoid soil damage 233 with the baseline scenario, irrespective of recovery time (Fig. 2A). As recovery time increased, higher pasture yield was positively correlated with less grazing time (more stand-off time). In Manawatu, stand-off also benefited pasture yield, but full protection (0 grazing hours/day) resulted in lowerpastureyieldcomparedwiththebaseline when recovery time was less than 100 days, and also resulted in lower pasture yield compared with some level of grazing, irrespective of recovery time (Fig. 3A). Average annual reduction in potential pasture growth for the baseline scenarios from both regions was 2%9% and 3%14% for Southland and Manawatu farms, respectively, depending on recovery time. The percentage reduction in potential growth decreased with fewer grazing hours (Figs. 2B, 3B). In the baseline scenarios, pasture lost during grazing of wet paddocks (wastage) varied from approximately 480 kg DM/ha/y for Southland, which was lower than the 1000 kg DM/ha/y for Manawatu. This occurred due to the 24 hour/day stand-off for the winter months on the Southland farm. On the Manawatu farm, some grazing was allowed in the winter when the pugging threshold was not exceeded, but paddocks may have been wet enough to result in some wastage. In general, wastage decreased with less grazing time and also decreased with lower annual pasture yield (Figs. 2C, 3C). The model predicted higher MS production for the alternative scenarios compared with the baseline for both regions. This was mainly driven by more lactation days in Southland (240 days in baseline versus 270 days in alternative) and more cows/ha in Manawatu. However, the higher production came at the expense of much higher supplement demand and feeding costs (Figs. 2D, 3D) to compensate for the times cows were restricted from grazing. Supplements also provided a feed buffer so that neither stand-off treatment nor recovery time had much effect on MS production (Figs. 2E, 3E). In both regions, the baseline operating profit decreased with an increase in recovery time, mainly due to decreased annual pasture yield, a decrease in MS production, and an increase in feeding expenses with longer recovery times. Also, in both regions the alternative scenario with full protection (0 grazing hours/ day) showed the lowest operating profit of the stand-off treatments, mainly because of high feeding expenses (Figs. 2F, 3F, 2G, 3G). With a fairly high milk price ($7.36 kg/ms) the results showed that for Southland, some form of limited grazing (48 hours/day), but not full protection, appears to become more profitable compared with the baseline when recovery time exceeded 180 days (Fig. 2F). Similarly, for the Manawatu farm, some form of limited grazing (e.g. 6 hours/day) appears to become more profitable compared with the baseline when recovery time exceeded 100 days (Fig. 3F). With an average milk price ($6.00 kg/ms), none of the stand-off treatments were profitable compared with the baselines, irrespective of recovery time (Fig. 2G, 3G). Discussion When implementing a stand-off strategy there are interactions between pugging recovery time, pasture yield, pasture wastage, feed expenses and stand-off costs that ultimately determine the profitability of the strategy. This modelling study indicates that full protection (0 grazing hours/day when pugging risk is present) is not the most profitable option to deal with pugging risk for several reasons. The zero pugging damage (zero percentage reduction in pasture yield) associated with full protection does not translate into substantially higher pasture yields. In fact, in the Manawatu region, the full protection scenario resulted in the lowest pasture yield of all the stand-off treatments. This can be explained by pasture herbage mass increasing in paddocks as a result of cows being stood off for considerable lengths of time. When soil moisture levels decrease and cows are allowed to enter the next paddock in the round, herbage mass is higher resulting in higher post-grazing residuals. In the pasture model, the higher residuals result in slower regrowth because of shading (McCall & Bishop- Hurley 2003). The accumulation of herbage mass prior to grazing has a further negative effect in that more herbage is lost through senescence (Chapman & Lemaire 1993). Although silage

12 234 PC Beukes et al. Figure 2 Predicted results for baseline and alternative scenarios with different stand-off treatments (grazing time varying from 0 to 10 hours/day) and pugging recovery time for the Southland region of New Zealand. A, Annual pasture yields; B, pugging damage; C, pasture wastage; D, feeding expenses; E, MS production; F, operating profit at $7.36/kg MS; G, operating profit at $6.00/kg MS.

13 Evaluating the benefits of standing cows off pasture to avoid soil damage 235 Figure 3 Predicted results for baseline and alternative scenarios with different stand-off treatments (grazing time varying from 0 to 10 hours/day) and pugging recovery time for the Manawatu region of New Zealand. A, Annual pasture yields; B, pugging damage; C, pasture wastage; D, feeding expenses; E, MS production; F, operating profit at $7.36/kg MS; G, operating profit at $6.00/kg MS.

14 236 PC Beukes et al. cutting rules were implemented to prevent the increase in average farm herbage mass in the stand-off scenarios, these rules were only activated when soil moisture conditions were favourable to prevent damage from machinery. The combination of slower regrowth and higher senescence rates depressed pasture yield in the full-protection scenarios, especially in the Manawatu region. In the Southland region, this depression of pasture yield in the full protection scenario was not as obvious because much more silage was made in this scenario compared with the other standoff scenarios and compared with Manawatu. The silage-making rules in Southland allowed for more pasture to be harvested in an attempt to control excessive average farm herbage mass, therefore negating the negative impacts of high residuals on pasture growth. This meant that the benefits of full protection, i.e. lower wastage and less damage, were realized to a larger extent. Although full protection resulted in reduced wastage of kg DM/ha/y, which translates into a saving of $3060/ha/y (at an assumed cost of production of pasture of $0.15/kg DM), this saving must be compared with the higher feed expenses associated with full protection of approximately $200300/ha/y. The variable costs of using stand-off pads are higher in the full protection scenario due to the higher number of cow-hours on the stand-off pad. However, this is a relatively small contributing factor to the relatively low profitability of the full-protection scenario. The sensitivity analysis reported here covers a limited recovery time range of approximately 17 months. However, it demonstrates the expected trend of interactions between recovery time and grazing time, with differences in annual pasture yield and silage production becoming more pronounced for the different stand-off treatments as recovery time increased. Except for the full protection treatment (0 grazing hours/day), the baselines and all treatments showed a downward trend for pasture yield and silage made with increased recovery time. A consequence of this was an upward trend for feeding expenses, which made an important contribution to the downward trend in operating profit as recovery time increased. Stand-off variable costs and MS production did not vary much with treatment and recovery time and, therefore, had small effects on the differences in operating profit. With a fairly high milk price of $7.36/kg MS and with an approximate stand-off cost of $500/ha/y, some stand-off treatments (e.g. 8 hours grazing/day) for the Southland region were more profitable compared with the baseline scenario when recovery time exceeded 180 days. If recovery time is shorter than 180 days, then stand-off costs should be lower (e.g. $250/ha/y), or milk production and price should be higher to justify any stand-off treatment. With a milk price of $7.36/kg MSandwithanexpensivestand-offcostofapproximately $1200/ha/y, all stand-off treatments were more profitable compared with the baseline in the Manawatu region when recovery time exceeded 160 days. With the same milk production and price, and with a stand-off cost of approximately $1000/ha/y, the stand-off treatments with 6 to 10 hours grazing/day for Manawatu should be more profitable compared with the baseline, irrespective of recovery time. However, it is important to note that with a more conservative milk price of $6.00/kg MS, no stand-off treatment in any of the regions compared favourably with the baseline scenarios in terms of profitability. Milk production and price are key elements in the costbenefit analysis of alternative farm systems that include a stand-off treatment. The more expensive the envisaged stand-off facilities, the higher milk production should be above the baseline to make stand-off treatments profitable, although, in part, this defeats the purpose of designing low-footprint future dairy systems. This modelling study did not address the potential negative impacts of no stand-off (19 hours grazing/day by lactating cows) when soil moisture levels were below the pugging threshold on pasture yield and, therefore, silage dynamics, milk productions and operating profit. This can be misleading since Houlbrooke et al. (2009) found such restricted grazing thresholds do not protect against soil compaction by dairy cows. They concluded that the large number of grazing events

15 Evaluating the benefits of standing cows off pasture to avoid soil damage 237 when soil moisture conditions were high, yet considered safe according to the never-pugged treatment protocol, probably contributed to both soil compaction and reduced pasture growth during the traditionally wet early spring period. Conclusions Full protection (0 grazing hours/day when pugging risk) appeared to be less profitable compared with some level of grazing (610 hours/ day), because the advantages of reduced damage were outweighed by the disadvantages of increased senescence and the costs of making and feeding silage and/or importing supplementary feeds. When wet pastures are fully protected in a rotational grazing situation, covers build up in front of the cows resulting in slower daily growth rates (as a result of shading), higher senescence rates and therefore lower net herbage accumulation. Silage cutting machines cannot access the wet pastures resulting in permanent herbage losses that outweigh the benefits of protection. The difference in operating profit between full protection and some level of grazing was less as the recovery time increased, but for both Southland and Manawatu regions, grazing 68 hours/day when pugging risk was present appeared to be a sensible option, irrespective of recovery time. The profitability of alternative farm systems that include stand-off treatments will largely be determined by milk price and if a large proportion of the stand-off facility cost can be compensated for by higher milk production from the alternative system. It is quite likely that stand-off pads will deliver environmental benefits as some urinary nitrogen can be removed from the pastures at times when risks of N loss are greatest. Acknowledgements This work was funded by Pastoral 21, a collaborative research venture between DairyNZ, Fonterra, Dairy Companies Association of New Zealand, Beef Lamb NZ and the Ministry of Business, Innovation and Employment, and New Zealand dairy farmers through DairyNZ Inc. Thanks to Gil Levy for software development enabling the WFM to represent pasture wastage and pugging damage. References Betteridge K, Drewry JJ, Mackay AD, Singleton PL Managing treading damage on dairy and beef farms in New Zealand. Hamilton, New Zealand, Land and Environmental Management, AgResearch. 35 p. Beukes PC, Gregorini P, Romera AJ, Dalley DE The profitability and risk of dairy cow wintering strategies in the Southland region of New Zealand. Agricultural Systems 104: Beukes PC, Palliser CC, Macdonald KA, Lancaster JAS, Levy G, Thorrold BS, Wastney ME Evaluation of a Whole-Farm Model for pasturebased dairy systems. Journal of Dairy Science 91: Chapman DF, Lemaire G Morphogenetic and structural determinants of plant regrowth after defoliation. Proceedings of the XVII International Grassland Congress, 821 February 1993, Palmerston North, New Zealand. Pp de Klein CAM, Monaghan RM, Ledgard SF, Shepherd M A system s perspective on the effectiveness of measures to mitigate the environmental impacts of nitrogen losses from pastoral dairy farming. In: Edwards GR, Bryant RH eds. Meeting the challenges for pasture-based dairying. Proceedings of the 4th Australasian Dairy Science Symposium, Lincoln University, Christchurch, New Zealand. Pp Drewry JJ Natural recovery of soil physical properties from treading damage of pastoral soils in New Zealand and Australia: a review. Agriculture, Ecosystems and Environment 114: Drewry JJ, Cameron KC, Buchan GD Pasture yields and soil physical property responses to soil compaction from treading and grazing: a review. Australian Journal of Soil Research 46: Drewry JJ, Littlejohn RP, Paton RJ A survey of soil physical properties of sheep and dairy farms in southern New Zealand. New Zealand Journal of Agricultural Research 43: Gregorini P, Clark CEF, Jago JG, Glassey CB, McLeod KLM, Romera AJ Restricting time at pasture: effects on dairy cow herbage intake, foraging behaviour, hunger-related hormones, and metabolite concentration during the first grazing session. Journal of Dairy Science 92:

16 238 PC Beukes et al. Hadley C Plan a wintering system that works. Dairy Exporter. December. Pp Hanigan MD, Palliser CC, Gregorini P Altering the representation of hormones and adding consideration of gestational metabolism in a metabolic cow model reduced prediction errors. Journal of Dairy Science 92: Houlbrooke DJ, Drewry JJ, Monaghan RM, Paton RJ, Smith LC, Littlejohn RP Grazing strategies to protect soil physical properties and maximize pasture yield on a Southland dairy farm. New Zealand Journal of Agricultural Research 52: Houlbrooke DJ, Littlejohn RP, Morton JD, Paton RJ Effect of irrigation and grazing animals on soil quality measurements in the North Otago rolling downlands of New Zealand. Soil Use and Management 24: Houlbrooke DJ, Paton RJ, Styles T, Laurenson S Refining the proctor compaction test to guide off-grazing management decisions. In: Currie LD, Christensen CL eds. Adding to the knowledge base for the nutrient manager. ac.nz/publications.html. Occasional Report No. 24. Fertilizer and Lime Research Centre, Massey University, Palmerston North, New Zealand. 5 p. Johnson JR, McCallum DA, Thompson NA Pasture renovation after winter pugging damage. Proceedings of the New Zealand Grassland Association 55: McCall DG, Bishop-Hurley GJ A pasture growth model for use in a whole-farm dairy production model. Agricultural Systems 76: Piwowarczyk A, Giuliani G, Holden NM Can soil moisture deficit be used to forecast when soils are at high risk of damage owing to grazing animals? Soil Use and Management 27: Romera AJ, Beukes PC, Clark C, Clark DA, Levy H, Tait A Use of a pasture growth model to estimate herbage mass at a paddock scale on dairy farms. Computers and Electronics in Agriculture 74: Romera AJ, McCall DG, Lee JM, Agnusdei MG Improving the McCall herbage growth model. New Zealand Journal of Agricultural Research 52: Sheath GW, Boom CJ Impact of beef cattle grazing systems on treading damage and forage supply. Proceedings of the New Zealand Grassland Association 59: Singleton PL, Boyes M, Addison B Effect of treading by dairy cattle on topsoil physical conditions for six contrasting soil types in Waikato and Northland, New Zealand, with implications for monitoring. New Zealand Journal of Agricultural Research 43: Stewart M, Fisher AD, Verkerk GA, Matthews LR Winter grazing systems: management practices and cow comfort. Proceedings of the New Zealand Society of Animal Production 62: Tait A, Henderson R, Turner R, Zheng X Thin-plate smoothing spline interpolation of daily rainfall for New Zealand using a climatological rainfall surface. International Journal of Climatology 26: