Description, evaluation, and validation of the Teagasc Pig Production Model

Abstract

The Teagasc Pig Production Model (TPPM), a stochastic simulation model of a farrow-to-finish pig farm, was developed to investigate effects of changes in production systems on farm profitability. The model simulates, on a weekly basis, the annual production of a farm. Biological [e.g., herd size, number of litters/sow/year, and mortality rates (%)], physical (e.g., infrastructure), and technical (e.g., feeding practices) variables and their associated costs are included as components of the model. These inputs are used to calculate physical (e.g., feed usage and number of pigs slaughtered) and financial (e.g., annual cash flow, profit and loss account, and balance sheet) outputs. The model was validated using the Delphi method and by comparing the TPPM outputs to data recorded on 20 Irish pig farms through the Teagasc e-Profit monitor system and with complete receipts for the year 2016. Results showed that the TPPM closely simulates physical and financial performance of pig farms indicating that the TPPM can be used with confidence to study pig production systems under Irish conditions. Model applicability was demonstrated by investigating the impact of 2 changes in technical performance: 1) building of extra accommodation to increase body weight (BW) at sale by 15 kg (EXTRA ROOM) and 2) a change in feeding practices by providing finisher feed from 28 kg of BW (EARLY FINISHER) compared with over 38 kg of BW. In both scenarios, the same biological parameters were used. Mortality rates, feed ingredients costs, and price per kg of meat produced were included as stochastic variables with the input distributions derived based on historical data simulated using Monte Carlo sampling using the Microsoft Excel add-in @Risk. Annual mean net profit was €198,101 (90% confidence interval [CI]: €119,606–€275,539) for the TPPM base farm, €337,078 (90% CI: €246,320–€426,809) for the EXTRA ROOM, and €225,598 (90% CI: €146,685–€303,590) for the EARLY FINISHER. EXTRA ROOM was associated with higher costs and required higher income to cover the additional costs. The 90% CI of the EARLY FINISHER was similar to the TPPM base farm while the EXTRA ROOM scenario resulted in a wider confidence interval, suggesting that a change in feeding practices could be a better option for farmers looking to improve profit with minimum investment. Thus, the TPPM could be used to facilitate decision making in farrow-to-finish pig farms.

INTRODUCTION

Farm system modeling plays a vital role in estimating the performance and financial effects of different changes to production systems and allows realistic scenarios to be tested prior to implementation in a research or within a farm environment. Bio-economic models describe the links between the components of economic and biological processes (Kragt, 2012). They are used as tools to predict and understand system behavior by investigating such links (Sànchez-Marrè, 2014). Several models have been developed for pig production such as Auspig (Back et al., 1988), InraPorc (Dourmad et al., 2008; van Milgen et al., 2008), a sow replacement model (Niemi et al., 2017), and Pig2Win (Meensel et al., 2012). However, due to different structural and procedural practices between pig producing countries, as well as different purposes of developed models, it is important to develop a bio-economic model that is capable of simulating the conditions of a particular production system and market place. “Generic” models provide some basic analysis but a tailor-made model with specific country-based assumptions and practices is essential to interpret and mine the subtle and elusive effect of scenario changes in one particular region or another. To date, no bio-economic model that adjusts to the Irish pig production situation has been developed.

The pig industry is the third most important Irish agri-food sector, after dairy and beef, accounting for 8% of gross agricultural output (Pig Industry Stakeholder Group, 2016). There is an estimated 290 commercial farms in Ireland; pig population is estimated at ca. 1.6 million pigs including 149,900 breeding sows (Central Statistics Office, 2017). Irish pig farms are among the largest integrated herd sizes in Europe (InterPIG, 2017) and they are mainly family own and operated. Also, Irish pig farms have high feed costs and a low live weight at sale in the European context (InterPIG, 2017). Ireland is a net importer of cereals, which is a major challenge that the industry faces to stay competitive when compared with other pig-producing countries. Indeed, the Irish pig sector has traditionally suffered from considerable volatility due to market fluctuations in input and output prices, arising from changes in the global market.

As influencing the market prices is beyond the control of individual pig farmers, they can try to improve farm productivity by implementing changes in their production systems to increase output for the same level of inputs or reduce inputs for the same level of output. A first approach to increase farm productivity of the Irish pig industry centers on the possibility to increase weights at sale in an effort to reduce overhead costs by increasing carcass yields and ultimately increasing income per pig produced. Besides the need to acquire more debt to build additional facilities and the uncertainty around future pig prices, possible negative impacts on animal welfare and meat quality that could be associated with keeping pigs longer in the farm represent concerns for pig producers. Indeed, increasing weight at sale to 124 kg or more is associated with decreased performance and carcass leanness (Latorre et al., 2004).

A second option to increase productivity of Irish pig farms would be to adopt feeding practices similar to those used in other European pig producing countries like Spain and Germany such as phase feeding. At the present, most Irish pig farms use a sole finishing diet from around 38 kg to slaughter. This approach implies a waste of nitrogen because the requirements of the pig decrease drastically along the finishing phase. Using phase feeding, more than one diet is fed for shorter periods of time in order to closely meet pigs’ nutrient requirements according to their physiological state (Han et al., 2000). In the Irish scenario, providing pigs with finisher feed at approximately 28 kg of body weight, about 10 kg lighter than what is currently done in Irish farms (InterPIG, 2017), and introducing a second finishing diet would reduce over-feeding nitrogen and thus, reduce feeding costs (Han et al., 2000).

Over the years, a series of bio-economic models describing Irish production systems including dairy (Shalloo et al., 2004), beef (Crosson et al., 2006), and sheep (Bohan et al., 2016) have been developed. Such models have had an important impact on the industry and are continuously used in various aspects of production such as investment decisions (Shalloo et al., 2004), genetics (Mccarthy et al., 2007), GHG emissions (O’Brien et al., 2010; O’Brien et al., 2012), or nitrogen efficiency (Ryan et al., 2011), among others. It is expected that a bio-economic model for the Irish pig sector will be also used as a decision tool in different areas of pig production. Thus, this study contributes to the literature by illustrating how a model tool suited to support decision-making on real pig farms can be developed and validated.

The objectives in this study were 1) to develop and document a bio-economic pig farm model which is suited to analyze potential technology and management options which are available to pig farmers; 2) to validate the model against real farm data; and 3) to demonstrate the applicability of the bio-economic model by comparing the profitability of two changes in technical performance (i.e., increased live weight at sale by 15 kg and a change in feeding practices).

MATERIALS AND METHODS

Bio-Economic Model Development

The Teagasc Pig Production Model (TPPM) is a bio-economic simulation model describing, on a weekly basis, a farrow-to-finish pig farm. It was developed in a Microsoft Excel spread sheet using a similar approach to the Moorepark Dairy Systems Model (Shalloo et al., 2004) and the Teagasc Lamb Production Model (Bohan et al., 2016). Biological (e.g., herd size, number of litters/sow/year, and mortality rate for each production stage), physical (e.g., infrastructure), and technical (e.g., feeding practices and healthcare) variables and their associated costs are included as components of the model. The base farm scenario was a farrow-to-finish system with weekly farrowing batches with a mean of 2.38 litters per sow per year and 26.25 pigs produced per sow per year following data from the National Pig Herd Performance Report for 2016 (Teagasc, 2017). Farm performance was simulated for an entire year. The farm included 7 animal categories: 1) piglets (0 to 4 wk of age); 2) weaner stage 1 (5 to 9 wk of age); 3) weaner stage 2 (10 to 13 wk of age); 4) finishers (14 to 24 wk of age); 5) maiden gilts (24 to 32 wk of age); 6) gestating sows (≥32 wk of age); and 7) lactating sows (≥48 wk of age) which was based on the classification that is general on pig farms in Ireland. Boars (>10 mo of age) used for heat detection were included in the breeding female group as they receive similar feeding and are housed in similar accommodation.

For the TPPM, financial and economic outputs included variable and fixed costs, gross income, net profit, cash flow, and a balance sheet. Stochastic features were included into the budget by performing stochastic simulation by a process of Monte Carlo sampling to determine the influence of variation in biological inputs, feed ingredient costs, and carcass prices on farm profitability. Net profit was defined as income minus variable and fixed costs (including depreciation) and interest cost of capital. A schematic diagram of the TPPM is provided in Figure 1 and additional details of the development of the model are described subsequently.

Figure 1.

Schematic representation of the major component of the Teagasc Pig Production Model.

Animal Growth

A Gompertz growth function (Wellock et al., 2004) was used to simulate pig growth from birth to slaughter. The Gompertz growth function was fitted to longitudinal growth performance data for 677 Large White × Landrace Irish pigs (323 females and 354 males) with 4,158 available records in PROC NLIN of SAS v9.4 (SAS Inst. Inc., Cary, NC). Pigs originated from 3 different batches and bi-weekly body weight records were available for each pig during the weaner–finisher period. The following formula was used:

$${\displaystyle \begin{array}{l}{\displaystyle BW=W_0\mathrm{e}\mathrm{x}\mathrm{p}\left[{\mu }_0(1-e^{-Dt})/D\right],}\end{array}}$$ (1)

where BW is the body weight; $W_0$ is the value of the growth function at age 0; ${\mu }_0$ is the logarithm of the relative growth rate at age 0; and $D$ is the slope of the logarithm of the relative growth rate. Corresponding values were ${\mathrm{W}}_0$= 1.0189; ${\mathrm{\mu }}_0$ = 0.4067; and D = 0.0712.

Estimated body weight for the weaner–finisher period is presented in Table 1. The growth curve is representative of Irish pigs where animals are sent to slaughter at approximately 109 kg of BW at 24 wk of age (Teagasc, 2017)

Table 1.

Estimated body weight (BW), estimated approximate energy (kcal ME/d), and SID lysine (g/d) requirements and estimated daily feed (kg) used in the model for the growth–finisher period

Stage	Age, weeks	Estimated BW1, kg	Approximate ME requirements, kcal/day2	Estimated feed intake (including 5% wastage) kg/day3
Weaner 1	5	6.7	1,592	0.52
Weaner 1	6	7.4	1,592	0.52
Weaner 1	7	9.6	1,592	0.52
Weaner 1	8	12.2	2,092	0.95
Weaner 1	9	15.2	2,509	0.95
Weaner 2	10	18.7	2,973	0.95
Weaner 2	11	22.7	3,464	1.11
Weaner 2	12	27.1	3,974	1.27
Weaner 2	13	32.0	4,494	1.44
Finisher	14	37.4	5,012	1.62
Finisher	15	43.3	5,522	1.78
Finisher	16	49.5	6,014	1.94
Finisher	17	56.2	6,483	2.09
Finisher	18	63.1	6,923	2.24
Finisher	19	70.4	7,331	2.37
Finisher	20	77.9	7,705	2.49
Finisher	21	85.6	8,045	2.60
Finisher	22	93.5	8,351	2.70
Finisher	23	101.5	8,625	2.79
Finisher	24	109.6	8,868	2.86

¹Calculated fitting the Gompertz growth curve to longitudinal data of 547 pigs originating from 3 batches with bi-weekly growth performance records. $\mathrm{B}\mathrm{W}={\mathrm{W}}_0\mathrm{e}\mathrm{x}\mathrm{p}\left[{\mathrm{\mu }}_0(1-{\mathrm{e}}^{-\mathrm{D}\mathrm{t}})/\mathrm{D}\right]$, where BW = body weight; ${\mathrm{W}}_0$= the value of the growth function at age 0; ${\mathrm{\mu }}_0$ = logarithm of the relative growth rate at age 0; and $\mathrm{D}$ = slope of the logarithm of the relative growth rate. Corresponding values were ${\mathrm{W}}_0$= 1.0189; ${\mathrm{\mu }}_0$ = 0.4067, and D = 0.0712.

²Calculated following the National Research Council (NRC) Nutrient Requirements of Swine (NRC, 2012) equations for estimating nutrient requirements for pigs.

³Calculated as daily feed intake = $\frac{dailyMErequirements}{estimatedMEcontedofthediet}$.

Nutritional Management

Nutritional requirements (i.e., energy, amino acids, and minerals) were the central components of the model. Requirements varied for each animal category and were estimated following the recommendations from the National Research Council (NRC) Nutrient Requirements of Swine (NRC, 2012).

Feeding Practice and Diet Formulation

For each animal category, wheat–barley–soya–based diets were formulated to meet or exceed NRC (2012) requirements. All diets (Table 2) were formulated on a net energy, digestible amino acids, and ideal protein basis (i.e., the profile of amino acids the animal needs to meet maintenance and protein accretion requirements [Fuller, 2004]) with a least cost approach. The diets used in the different production stages were home-milled on-farm which is particularly relevant for Ireland, with 42.9% of farms home-milling (R. da Costa et al., in preparation). This allows future model applications to investigate the effect of changes in diet composition in Irish pig farms. The formula to estimate milling costs and values used were obtained from Lynch et al. (2002):

Table 2.

Diet composition used for the different production stages simulated in the Teagasc Pig Production Model¹

	Gestation	Lactation	Weaner 12	Weaner 23	Finisher 14	Finisher 25
Ingredient, kg per ton
Barley	488.16	392.52	418.76	458.27	436.69	520.80
Soya bean meal 48% crude protein (CP)	64.97	150.53	176.69	146.36	141.40	86.27
Wheat	400.00	400.00	350.00	350.00	394.07	364.64
Soya oil	20.86	25.81	12.67	8.76	0.00	0.00
l-Lysine	0.87	3.09	7.67	7.78	4.66	4.51
dl-Methionine	0.00	0.02	1.19	1.27	0.51	0.35
l-Threonine	0.25	0.85	3.08	3.20	1.70	1.53
Di-calcium phosphate	11.75	14.43	13.74	13.10	8.16	9.70
Limestone	7.39	7.14	7.35	6.92	8.44	7.76
Sodium chloride	3.25	3.11	6.35	1.83	1.87	1.94
Min/Vits	2.50	2.50	2.50	2.50	2.50	2.50
NRC6 requirements
Net energy (NE), MJ	10.5	10.5	10.1	10.1	10.0	10.0
Metabolizable Energy (ME), MJ	13.8	13.8	14.0	14.0	13.8	13.8
CP, %	13	16	17	16	16	14
Standard Ileal Digestible (SID) Lysine, %	0.50	0.87	1.29	1.23	0.98	0.84
SID methionine, %	0.14	0.25	0.36	0.36	0.28	0.24
SID threonine, %	0.32	0.56	0.80	0.77	0.63	0.54
SID tryptophan, %	0.08	0.14	0.22	0.20	0.16	0.13
Ca, %	0.62	0.7	0.7	0.66	0.59	0.59
Standardized total tract digestible (STTD) P, %	0.27	0.34	0.33	0.31	0.24	0.24
Na, %	0.15	0.15	0.28	0.10	0.10	0.10
Cl, %	0.12	0.12	0.32	0.08	0.08	0.08
Calculated values
NE, MJ	10.5	10.5	10.1	10.1	10.0	10.0
ME, MJ	13.8	14.1	13.9	13.7	13.6	13.4
CP, %	13	16	17	16	16	14
SID lysine, %	0.50	0.87	1.29	1.23	0.98	0.84
SID methionine, %	0.20	0.24	0.36	0.36	0.28	0.24
SID threonine, %	0.38	0.52	0.80	0.77	0.63	0.54
SID tryptophan, %	0.30	0.40	0.43	0.40	0.40	0.33
Ca, %	0.62	0.7	0.7	0.66	0.59	0.59
STTD P, %	0.27	0.34	0.33	0.31	0.24	0.24
Na, %	0.15	0.15	0.28	0.10	0.10	0.10
Cl, %	0.34	0.4	0.69	0.42	0.36	0.34

¹All diets were formulated on a net energy and ideal protein basis with a least cost approximation.

²Five to 9 wk of age; approximately 7 to 19 kg of body weight.

³Ten to 13 wk of age; approximately 19 to 38 kg of body weight.

⁴Fourteen to 24 wk of age; approximately 38 to 109 kg of body weight.

⁵Used for an scenario investigated the economic implications of phase feeding in the finisher stage where a second finisher diet was provided from 19 to 24 wk of age; approximately 70 to 110 kg of body weight.

⁶ National Research Council, 2012. Nutrient requirements of swine. National Academies Press.

$${\displaystyle \begin{array}{l}{\displaystyle \begin{array}{ll}& {\displaystyle Millingcostsperton=Feedingredienthaulageperton}\\ & {\displaystyle +\left({\displaystyle \frac{\begin{array}{c}Powerconsumptionperton,KWH\\ \times PriceperKWH\times total\\ feedmanufacuredperweek\end{array}}{totalfeedmanufacturedperweek}}\right)}\\ & +\left({\displaystyle \frac{laborcosts}{totalfeedmanufacturedperweek}}\right)\\ & +maintenanceperton,\end{array}}\end{array}}$$ (2)

Feed ingredient haulage per ton was set at €12, power consumption was 16 KWH at a price of €0.16 per KWH, and maintenance cost per ton was set at €1.34.

Maiden gilts and gestating sows were feed restricted based on energy demands (see Energy Demand), with a common gestation sow diet during 8 and 16 wk, respectively. On average, gilts received 3 kg of feed per day and gestating sows received 2.50 kg of feed per day. Lactating sows were fed ad libitum a common lactating sow diet during the 4-wk lactation period. From weaning to slaughter, pigs were fed ad libitum. Weaner pigs stage 1 were fed a creep diet during the first week postweaning, link diet during weeks 2 and 3 postweaning, and a weaner diet for 2 wk. Weaner pigs stage 2 were fed a second weaner diet for 4 wk and finisher pigs received a finisher diet for 11 wk. In all stages, feed was provided as dry meal to simulate common practice in Irish pig farms (R. da Costa et al., in preparation).

Energy Demand

For maiden gilts, gestating and lactating sows, energy demand was calculated as an average from the NRC (2012) and different parities and/or litter sizes were not modeled separately. Approximate energy requirements were 9,910 kcal of metabolizable energy (ME)/d, 8,182 kcal ME/d, and 20,700 kcal ME/d for replacement gilts, gestating and lactating sows, respectively. Feed intake was calculated by dividing approximate ME requirements by the estimated dietary energy of the feed provided. Then, shadow formulation was used to calculate net energy intake where average daily feed intake was multiplied by the estimated net content of the diet.

For weaner pigs stage 1 and stage 2 and finisher pigs, required daily energy intake was calculated on a ME basis following the NRC (2012) equations for estimating nutrient requirements for weaner–finisher pigs according to their estimated BW obtained from the Gompertz growth curve. Required daily energy intake was calculated for entire males and females separately. The results were then averaged as the majority of Irish pigs are housed in mixed sex pens (van Staaveren et al., 2018) using formulas in equations 2 and 3:

$${\displaystyle \begin{array}{ll}& RequiredMEintake,\\ & females\left(kcal/day\right)=10,967\times \{1-exp[-exp(-3.803)\\ & \times B{W}^{0.9072}\left]\right\}\end{array}}$$ (3)

$${\displaystyle \begin{array}{ll}& RequiredMEintake,\\ & entiremales\left(kcal/day\right)=10,638\times \{1-exp[-exp(-3.803)\\ & \times B{W}^{0.9072}\left]\right\}\end{array}}$$ (4)

Required kcal/day of ME were transformed to MJ/ day as

$${\displaystyle \begin{array}{l}{\displaystyle \frac{kcal\times 4.184}{1000}}\end{array}}$$

For all animal categories, average daily feed intake was calculated as follows:

$${\displaystyle \begin{array}{l}{\displaystyle AverageDailyFeedIntake=\frac{\begin{array}{c}RequireddailyME\\ intake(MJ/kg)\end{array}}{\begin{array}{c}EstimatedMEcontent\\ ofdiet(MJ/kg).\end{array}}}\end{array}}$$ (6)

A 5% feed wastage was assumed (NRC, 2012) and added to the above average daily feed intake calculation. Estimated daily net energy intake was calculated by multiplying the estimated average daily feed intake by net energy content of the diet. Approximate energy requirements (kcal ME/day) and estimated daily feed (kg) for the growth–finisher period are presented in Table 1.

Reproductive Management

All reproductive parameters used for the TPPM were obtained from the Teagasc Pig e-Profit Monitor, an online financial analysis tool for assessing farm profitability which contains biological and economic records for over 65% of the Irish pig herds. It was assumed that replacement gilts were home reared as per common practice in Irish pig farms (Rodrigues da Costa et al., 2019). Gilts were selected at 24 wk of age and remained in finisher pens until 32 wk of age. During this period, gilts were exposed daily to a rotation of two mature vasectomized boars using direct daily single boar contact and observed for signs of standing estrus. Gilts were artificially inseminated on their second estrus and moved to the gestation accommodation. A 90% gilt selection rate was assumed. All breeding females were artificially inseminated when standing estrus was observed and 24 h after the first service. A 92% conception rate and 86% farrowing rate were used. All breeding females were group housed in the gestation barn until 1 wk before their expected due date when they were transferred to the farrowing accommodation.

In the farrowing accommodation, sows were individually housed in farrowing pens fitted with a centrally positioned farrowing crate. Lactating sows remained in the farrowing accommodation for 4 wk after farrowing. At weaning, sows were moved to the gestation barn and artificially inseminated approximately 3 to 5 d after weaning based on detection of standing estrus and 24 h after the first service.

Livestock Movements and Valuations

Movements of animals from one category to another are represented in Figure 2. Number of gilts, gestating and lactating sows as well as number of piglets, weaners, and finisher pigs were calculated each week within the model based on the mortality rate for the different production stages. Additionally, numbers of culled, dead, and slaughtered pigs were also calculated. Number of replacement gilts was calculated based on sow mortality and culling (including both voluntary and involuntary) rate using data obtained from the Teagasc National Pig Herd Performance Report for 2016 (Teagasc, 2017). The modeled annual sow mortality and culling rates were 4.9% and 50.1%, annually, respectively. Piglet, weaner, and finisher mortality were also obtained from the Teagasc National Pig Herd Performance Report for 2016 (Teagasc, 2017) and were set at 10.9%, 2.85%, and 2.49%, respectively. Stock valuation was set for each animal category based on the cost of production (i.e., market conditions not taken into account). A total stock value was calculated for the start and end of each week.

Figure 2.

Timeline for the animal movements between the different animal categories in an Irish pig farm. The pink section represents animal movements in the breeding herd while the gray section represents animal movements in the weaner–finisher period.

Buildings and Capital

Farm buildings were depreciated at 5% per annum using the straight line method as per the industry norm. Input values were calculated based on the assumption that the buildings were in the 10th year of their useful life. A 15-yr bank loan at a nominal interest rate of 5% was used to fund the cost of the buildings (Thorne et al., 2015). The bank term loan was assumed to be in its 10th year and the interest was considered a finance expense.

Labor

Labor requirements for Irish pig farms have not been described previously. Therefore, a general categorization for management and farm activities was considered. A total of 44-h work per week per employee was used based upon consulting with the members of the Teagasc Pig Advisory team. Five farm operatives and one farm manager were employed at the farm. Number of farm operatives was calculated based on the information reported by Rodrigues da Costa et al. (2019) of one farm operative per 154 sows obtained from a cross-sectional survey carried out in 72 Irish pig farms. Additionally, 20 h of labor per week was assumed to operate the feed mill. Labor costs were extrapolated from data available in the Teagasc pig e-Profit monitor for the year 2016. Cost for owner/operator labor was set at €20 per hour for farm-related activities (e.g., perform artificial insemination, assist farrowing, vaccination, moving pigs from one stage to another, sorting pigs, maintenance of equipment, weighting pigs, among others) and €22 per hour for management activities.

Animal Health

The model considered a high health status farm in which health and veterinary costs varied by animal category. The farm was considered only positive for erysipelas, parvovirosis, and enzootic pneumonia. Thus, maiden gilts and lactating sows received a single vaccine against Erysipelothrix rhusiopathiae and Porcine Parvovirus costing €0.74 per dose. Weaner stage 1 pigs were vaccinated against Mycoplasma hyopneumonia costing €0.79 per dose. No in-feed antibiotics were used in the farm. Prices for the vaccines were obtained from a major veterinary distributor in Ireland. Two veterinarian visits per year at €300 each were considered for the model as per usual practice (personal communication Jesus Borobia, DVM).

Other Costs

Headings for other costs including electricity, annual subscription to the Environmental Protection Agency, monthly feed ingredient prices, manure handling cost per m³, and transport costs per pig to the abattoir were obtained from the Teagasc National Pig Herd Performance Report for 2016 (Teagasc, 2017). Electricity and manure handling costs were included on a per pig space basis. Electricity usage was set at 600 KWH sow/year (McCutcheon, 2016) with a price of €0.16 per KWH. It was assumed that 5% of electricity was used in the gestation barn, 55% in the farrowing rooms, 20% in the weaner stages, and 20% in the finisher stage. The assigned percentages were obtained by consulting the Teagasc Pig Advisory team members. Annual subscription fee paid to the Environmental Protection Agency was set at €10,000 per year. Monthly feed ingredient prices for the 4 cereals (barley, wheat, and soy bean) are included in Figure 3. Soya oil price ranged from €644 to €776 with a mean price of €732.6 ± 41.65. Price per ton for l-Lysine (€1,300), dl-Methionine (€3,050), l-Threonine (€1450), Di-Calcium phosphate (€700), limestone (€120), sodium chloride (€195), and vitamin and mineral mix for sows (€4,450), weaners (€5,800), and finisher (€4,450) pigs were also obtained from the Teagasc e-Profit monitor. Regarding manure handling costs, it was considered that a sow produced 21 m³ of manure per year costing €2 per m³ (Nolan et al., 2012). It was assumed 12% of manure was produced by gestating sows, 8% of manure was produced by lactating sows, 20% of manure was produced by weaner pigs, and 60% of manure was produced by finisher pigs. The assigned percentages were obtained by consulting the Teagasc Pig Advisory team members. Transportation costs per pig to the abattoir were set to €0.93 per pig (Teagasc, 2017).

Figure 3.

Monthly cereal prices (€/ton) in 2016 used to for feed formulation and pricing during the development of the Teagasc Pig Production Model, a bio-economic simulation model for farrow-to-finish pig farms.

Source of Income

The only source of income was livestock sales including culled sows and slaughtered finisher pigs. Finisher pigs were slaughter at 24 wk of age with an estimated average daily gain from weaning to slaughter of 740 g/d. Average daily gain was calculated based on the growth curve previously described which is similar to that reported in the National Pig Herd performance report for 2016 (Teagasc, 2017). In Irish farms, pigs are slaughtered around 109 kg of live weight (Teagasc, 2017); this is mostly done to avoid boar taint as males are not castrated in Ireland. A kill out percentage of 76.4% was assumed (Teagasc, 2017). Cold carcass weight was calculated by multiplying body weight at sale at 24 wk of age by kill out percentage. Average monthly price per kg/meat was obtained from the Teagasc pig e-Profit Monitor and ranged from €1.34 to €1.63 per kg of meat with a mean price of €1.49 ± 0.10. Premiums with the processing plants were not considered for this model. Culled sow value was set at €120 (Carroll, 2011).

Model Outputs

Outputs from the model include annual cash flow budget, annual profit and loss account, and annual balance sheet. Cash flows were summarized quarterly and indicate cash surpluses or deficits. The estimated annual farm profit was presented on a total farm basis, as well as per pig produced and per kg of meat sold. Net profit, return on total capital investment (%) and liquidity, and solvency indicators were also outputs built into the model.

Model Validation

The TPPM was validated using the Delphi method where a group of experts (i.e., pig advisors and researchers) evaluated the methodology and values used for the model. Once the experts agreed, a second evaluation was carried out by comparing TPPM outputs with actual farm data from the Teagasc pig e-Profit Monitor. The model was parameterized to simulate the biological performance of 20 Irish pig farms (ePM farms). Farms were selected based on the following criteria 1) farms must be Teagasc clients participating in the Teagasc e-Profit monitor, 2) farms must be home milling animal feed, and 3) farms must have complete records for the year 2016 including sales, variable costs, fixed costs, and net profit. A qualitative comparison was performed between the simulated physical and economic results and the average performance of the ePM farms.

Model Application

To demonstrate an application of the TPPM, the impact of two changes in technical performance scenarios in farm net profit were investigated. In both scenarios, a 775 sow farm was simulated. Biological parameters such as number of litters per sow per year, number of piglets born alive per litter, and mortality rates were the same used for the TPPM-based farm.

The first scenario aimed to increase live weight at sale by 15 kg per pigs in an effort to reduce overhead costs by increasing carcass yields. This involved the construction of extra finisher accommodation (EXTRA ROOM) and keeping pigs on farm for 2 more weeks. In total, 830 new finisher spaces (minimum space per pig = 1 m²) were required at €275 per space (including infrastructure and all necessary equipment such as lights, ventilation system, feeders, feed bins, and feed and water lines) for a total investment of €228,250. Newly constructed facilities were funded through a term loan (15 yr at a nominal interest rate of 5%) and the newly constructed facilities were depreciated over a 20-yr time frame. This scenario was decided based on the most common plans to improve profitability as expressed by farmers to the Teagasc pig advisory team in farmer discussion groups. Possible implications on meat quality were not taken into consideration as currently there is no paying scheme according to meat quality characteristics in Ireland.

The second scenario was to introduce phase feeding by providing finisher diets earlier, i.e., from 28 kg of BW weight instead of 38 kg of BW and the introduction of a second finisher diet from 70 kg of BW (EARLY FINISHER). The advantage of a phase feeding is that of allowing farmers to meet pigs’ nutrient requirements more closely, minimizing over-feeding nutrients, and reducing feeding costs (Han et al., 2000; Lee et al., 2000) without adversely affecting growth performance (Friesen et al., 1995; Canh et al., 1998; Menegat et al., 2017). This scenario was decided based on what is common practice in other European pig producing countries such as Spain and Germany (InterPIG, 2017). For this scenario, a new feed bin was installed in the farm at a total investment of €10,000 euros (including the bin, feed lines, and installation costs). This was financed through a term loan (15 yr at a nominal interest rate of 5%) and depreciated over a 20-yr time frame.

Risk Analysis

To account for uncertainty, stochastic features were included into the budget by performing stochastic simulation by a process of Monte Carlo sampling to determine the influence of variation in biological inputs, feed ingredient costs, and carcass prices on farm profitability using the Microsoft Excel add-in @Risk (Palisade, 2013). During Monte Carlo risk assessment, a specific probability distribution is assigned to each stochastic variable from where a set of values is drawn at each iteration (Phillips and Maldonado, 1999). Stochasticity was used in the TPPM base farm as well as in the EXTRA ROOM and the EARLY FINISHER scenarios. Stochastic variables included mortality for the different age groups, number of piglets born alive, number of litters per sow per year, monthly creep and link feed cost, monthly feed ingredient costs, and monthly price per kg of meat. Minimum, most likely, and maximum estimates were generated based on data recorded on the Teagasc pig e-Profit monitor between the years 2012 and 2016 (Table 3). To account for possible co-variation between stochastic variables, spearman correlations were estimated in PROC CORR of SAS v9.4 (SAS Institute Inc., Cary, NC) and they were included during the Monte Carlo simulation; however, correlations were low and were not significantly different from zero. A Program Evaluation and Review Technique (PERT) distribution was fitted for each of the stochastic variables. A PERT distribution uses the minimum, most likely, and maximum values similar to the triangular distribution; however, values around the most likely are more likely to occur as extremes are not emphasized (Palisade, 2013). During the Monte Carlo simulation, 10,000 iterations were done for each risk variable. Multiple regression analysis was performed on the simulated data to obtain the partial coefficients of determination to measure the relative contribution of each stochastic variable to mean net profit for each scenario.

Table 3.

Range (minimum and maximum values) for the stochastic values included in the Teagasc Pig Production Model

Stochastic variable	Minimum	Most likely	Maximum
Number of piglets born alive per litter	12.7	12.85	13.2
Litters per sow per year	2.27	2.35	2.39
Sow culling rate %	48.50	49.28	50.10
Sow mortality %	4.80	5.00	5.20
Piglet mortality %	10.60	11.00	11.40
Weaner mortality %	2.54	2.67	2.85
Finisher mortality %	2.38	2.44	2.49
Creep feed, €/ton	861.0	922.19	978.0
Link feed, €/ton	576.0	626.71	682.0
Barley, €/ton	142.0	175.69	178.0
Soya bean meal, €/ton	382.0	418.77	480.0
Soya oil, €/ton	669.0	809.67	1070.0
Wheat, €/ton	157.0	193.19	259.0

RESULTS

TPPM Physical and Economic Outputs

Herd size remained constant at 775 sows, including 600 gestating and 175 lactating sows, at all times throughout the year. During the year, a total of 388 sows were culled and 38 sows died. A total of 23,920 piglets were born alive; 2,600 piglets died during lactation, and 21,320 piglets were weaned. During the weaner stages, 624 weaner pigs died and 20,696 weaner pigs were transferred to the finisher stage. Five hundred twenty finisher pigs died and a total of 20,176 finisher pigs reached slaughter age. A total of 468 females were selected as replacement gilts (Table 4). Feed requirements varied according to each animal category. Gestating sows consumed 686.2 tons of feed and lactating sows consumed 402.3 tons of feed. Weaner pigs required 63.3 tons of creep feed, 149.2 tons of link feed, and 949.8 tons of weaner feed. Finisher pigs required 3,470.1 tons of feed (Table 4). In total, 14,768 h of labor were required to run the farm for a year. Labor requirements varied according to activity performed with 1,040 h for feed milling, 11,440 h for farm-related activities, and 2,288 h required for managing the farm (Table 4). A total of 19,708 finisher pigs (i.e., 20,176 finishers minus 468 replacement gilts) were sold producing 1,649.6 tons of meat. Average price per kg of meat produced from finisher pigs was €1.49 and average price per finisher pig sold was €122.6 for a total income of €2,416,014 for the entire year. Annual culled sow income was €46,593 (€2.36 per finisher pig sold and €0.03 per kg of meat produced from finisher pigs; Table 4).

Table 4.

Physical inputs and outputs from the Teagasc Pig Production Model1

	Weeks												TOTAL
	Quarter 1			Quarter 2			Quarter 3			Quarter 4
	1–4	5–8	9–13	14–17	18–21	22–26	27–30	31–34	35–38	39–43	44–47	48–52
Stock numbers
Sow herd size	775	775	775	775	775	775	775	775	775	775	775	775	–
No. of gestating sows	600	600	600	600	600	600	600	600	600	600	600	600	–
No. of lactating sows	175	175	175	175	175	175	175	175	175	175	175	175	–
No. of piglets born alive	1840	1840	2300	1840	1840	2300	1840	1840	2300	1840	1840	2300	23,920
No. of piglets weaned	1,640	1,640	2,050	1,640	1,640	2,050	1,640	1,640	2,050	1,640	1,640	2,050	21,320
No. of weaner pigs transfer to finisher	1,592	1,592	1,990	1,592	1,592	1,990	1,592	1,592	1,990	1,592	1,592	1,990	20,696
No. of finisher pigs	1,552	1,552	1,940	1,552	1,552	1,940	1,552	1,552	1,940	1,552	1,552	1,940	20,176
No. of finisher pigs selected as replacement gilts	36	36	45	36	36	45	36	36	45	36	36	45	468
Mortality and culling rates
Sow culling	30	30	37	30	30	37	30	30	37	30	30	37	388
Sow mortality	3	3	4	3	3	4	3	3	4	3	3	4	38
Piglet mortality	200	200	250	200	200	250	200	200	250	200	200	250	2,600
Weaner mortality	48	48	60	48	48	60	48	48	60	48	48	60	624
Finisher mortality	40	40	50	40	40	50	40	40	50	40	40	50	520
Feed usage, tons
Gestation feed	52.8	52.8	66.0	52.8	52.8	66.0	52.8	52.8	66.0	52.8	52.8	66.0	686.2
Lactation feed	30.9	30.9	38.7	30.9	30.9	38.7	30.9	30.9	38.7	30.9	30.9	38.7	402.3
Creep feed	4.9	4.9	6.1	4.9	4.9	6.1	4.9	4.9	6.1	4.9	4.9	6.1	63.3
Link feed	11.5	11.5	14.4	11.5	11.5	14.4	11.5	11.5	14.4	11.5	11.5	14.4	149.2
Weaner feed	73.1	73.1	91.3	73.1	73.1	91.3	73.1	73.1	91.3	73.1	73.1	91.3	949.8
Finisher feed	266.9	266.9	333.7	266.9	266.9	333.7	266.9	266.9	333.7	266.9	266.9	333.7	3,470.1
Sales
No. of finisher pigs sold	1516	1516	1895	1516	1516	1895	1516	1516	1895	1516	1516	1895	19,708
No. of kg of meat sold	126891	126891	158614	126891	126891	158614	126891	126891	158614	126891	126891	158614	1,649,585
Finisher pig carcass value, c/kg	138	138	134	136	141	149	154	156	160	163	157	157	–
Finisher pig carcass value, €/pig	115.5	115.5	112.2	113.8	118.2	124.7	128.9	130.6	133.9	136.4	131.4	131.4	–
Culled sow salvage value, €/sow	110	110	110	110	110	110	110	110	110	110	110	110	–
Culled sow income, €	3,285	3,285	4,107	3,285	3,285	4,107	3,285	3,285	4,107	3,285	3,285	4,107	42,710
Labor requirement, h
Feed milling	80	80	100	80	80	100	80	80	100	80	80	100	1,040
Farm-related activities	880	880	1100	880	880	1100	880	880	1100	880	880	1100	11,440
Farm management	176	176	220	176	176	220	176	176	220	176	176	220	2,288

1The Teagasc Pig Production Model (TPPM), a bio-economic stochastic simulation model for farrow-to-finish pig farms. The TPPM models on a weekly basis the annual production of a farm. For brevity, results are summarized on a quarterly basis here.

Total farm sales (i.e., finisher sales plus culled sows sales) for the year were €2,462,608 (€124.95 per pig and €1.52 per kg of meat produced), total variable cost were €1,592,215 (€80.79 per pig and €0.98 per kg of meat produced), and total fixed cost including depreciation charges were € 663,350 (€33.66 per pig and €0.41 per kg of meat produced). Farm net profit was €207,041 (€10.51 per pig and €0.13 per kg of meat produced; Table 5). The farm had a return on investment of 4.38% and a liquidity ratio (i.e., indicator of the ability of a company current assets to meet their short term financial obligations) of 0.40. Solvency indicators such as farm debt to assets ratio, equity to assets ratio, and debt to equity ratio were 0.40, 1.40, and 0.29, respectively.

Table 5.

Trading profit and loss accounts for the Teagasc Pig Production Model base farm

	€/year	€/pig produced	€/kg meat produced
Sales
Finisher pigs	€2,416,014.59	€122.59	€1.49
Culled sows	€ 46,593.00	€2.36	€0.03
Total sales	€2,462,607.59	€124.95	€1.52
Variable costs
Gestation feed	€126,879.56	€6.44	€0.08
Lactation feed	€99,685.03	€5.06	€0.06
Creep feed	€58,381.41	€2.96	€0.04
Link feed	€88,648.56	€4.50	€0.05
Weaner feed	€240,114.07	€12.18	€0.15
Finisher feed	€798,930.40	€40.54	€0.49
Replacement gilts	€76,585.41	€3.89	€0.05
Dead animal disposal	€12,645.04	€0.64	€0.01
Healthcare	€17,469.92	€0.89	€0.01
Reproduction	€37,288.24	€1.89	€0.02
Manure handling	€16,732.80	€0.85	€0.01
Transport	€18,763.68	€0.95	€0.01
Total variable costs	€1,592,124.12	€80.79	€0.98
Fixed costs
Admin and accounting	€2,500.00	€0.13	€0.00
Electricity, heating, and light	€79,281.43	€4.02	€0.05
Insurance	€20,533.14	€1.04	€0.01
Repairs	€20,533.14	€1.04	€0.01
Environment	€10,000.00	€0.51	€0.01
Labor	€279,136.00	€14.16	€0.17
Loan repayments–of which interest	€76,346.47	€3.87	€0.05
Depreciation charges	€175,020.63	€8.88	€0.11
Total fixed costs	€663,350.80	€33.66	€0.41
Total farm costs	€2,255,474.92	€114.44	€1.39
Net profit	€207,132.67	€10.51	€0.13

Model Validation

Results from the model validation indicate that the TPPM closely simulates the 20 ePM farms. Physical inputs and outputs are presented in Table 6. Physical inputs such as herd size, farrowing rate, litters produced per sow per year, number of piglets born alive per litter, pigs produced per sow per year, mortality rates in the different animal categories, BW at sale, and kill out % were similar between the TPPM and the ePM farms. Physical outputs such as number of pigs sold per year and number of kg of meat produced were similar between the TPPM and the ePM farms. Likewise, TPPM-simulated results for feed usage in the different animal categories were similar to the average feed usage for the ePM farms. Economic performance outputs are presented in Table 7. The TPPM had lower total sales, lower feed costs, and higher nonfeed variable costs than the ePM farms. Net profit was similar between the TPPM and the 20 ePM farms.

Table 6.

Comparison of physical inputs and outputs of the Teagasc Pig Production Model (TPPM) with data from 20 farms with records in the Teagasc pig e-Profit monitor (ePM) used to validate the TPPM

Performance variable	TPPM	ePM farms (n = 20)
		Mean ± SD	Minimum	Maximum
Inputs
Sow herd size	775	810 ± 495	175	2,479
Conception rate, %	95.0	93.5 ± 3.3	86.1	98.5
Farrowing rate, %	86.0	85.4 ± 5.5	76.4	92.1
Litters per sow per year	2.4	2.3 ± 0.12	2.0	2.4
Average number of piglets born alive per litter	13.2	13.3 ± 0.57	12.4	14.5
Culling rate, %	50.1	50.6 ± 8.10	36.0	70.1
Sow mortality rate, %	4.9	4.8 ± 2.51	2.1	12.2
Piglet mortality rate, %	10.8	10.5 ± 2.79	5.2	15.0
Weaner mortality rate, %	2.9	2.7 ± 1.24	1.0	6.6
Finisher mortality rate, %	2.5	2.0 ± 0.98	0.8	4.6
Average BW on sale, kg	109.6	108.5 ± 4.10	100.1	120.1
Kill out %	76.4	77.1 ± 7.00	74.6	82.7
Outputs
Average piglets/sow/year	26.3	26.1 ± 1.79	22.9	28.9
No. of pigs sold	20,748	19,594 ± 11,555	5,295	57,561
No. of kg/DW produced, ton	1,709.6	1,648.4 ± 1,023.9	398.9	5,051.2
Gestating sow feed, ton	714.7	675.0 ± 444.54	202.0	2,170.1
Lactating sow feed, ton	413.8	401.6 ± 241.58	98.0	1,091.9
Creep feed, ton	66.1	61.7 ± 44.36	12.0	180.5
Link feed, ton	155.8	160.5 ± 191.18	27.8	783.8
Weaner feed, ton	992.7	1,046.8 ± 721.11	282.5	3,127.3
Finisher feed, ton	3,649.0	3,707.0 ± 2,386.62	846.0	11,914.7

Table 7.

Comparison of trade profit and loss accounts of the Teagasc Pig Production Model (TPPM) with data from 20 farms with records in the Teagasc pig e-Profit monitor (ePM) used to validate the TPPM

€ per pig produced	TPPM	ePM Farms
		mean ± SD	Minimum	Maximum
Sales
Finisher pigs	119.8	127.0 ± 8.52	115.9	148.8
Culled sows	2.3	1.88 ± 0.62	0.1	2.9
Total sales	122.1	128.9 ± 8.65	118.0	151.6
Variable costs
Gestating sow feed	6.3	8.0 ± 1.17	6.4	10.5
Lactating sow feed	4.8	5.8 ± 1.32	3.9	9.7
Creep feed	2.9	2.8 ± 1.21	1.0	5.3
Link feed	4.4	4.1 ± 2.03	1.3	9.7
Weaner feed	11.8	12.5 ± 3.13	6.9	19.7
Finisher feed	39.6	42.7 ± 4.8	35.3	51.4
Replacement gilts	3.61	–	–	–
Dead animal disposal	0.53	0.4 ± 0.24	0.0	0.9
Healthcare	1.28	4.7 ± 2.25	2.2	11.9
Reproduction	1.76	1.7 ± 0.49	1.0	3.2
Manure handling	0.83	0.6 ± 0.66	0.0	2.1
Transport	0.93	0.5 ± 0.74	0.0	2.8
Total variable costs	78.7	83.8 ± 8.81	69.9	103.0
Fixed costs
Admin and accounting	0.12	0.4 ± 0.54	0.0	2.5
Electricity, heating, and light	3.68	2.5 ± 1.36	0.3	5.6
Insurance	0.52	0.6 ± 0.65	0.0	2.5
Repairs	0.52	1.6 ± 1.31	0.0	4.6
Environment	0.47	0.3 ± 0.29	0.0	1.0
Labor	6.69	7.6 ± 3.38	3.5	17.9
Loan repayments—of which interest	1.25	0.8 ± 0.97	0.0	3.2
Depreciation charges	2.77	3.8 ± 5.17	0.0	17.6
Total fixed costs	16.0	17.5 ± 8.75	5.7	35.5
Total costs	94.7	101.4 ± 10.77	78.8	117.7
Net profit	27.4	27.5 ± 12.45	4.7	51.8

Model Application

Differences in trading profit between the base TPPM simulated farm and two alternative scenarios used to demonstrate the TPPM applicability to real farming situations are presented in Table 8. The EXTRA ROOM scenario required 812 tons more of finisher feed amounting to €185,104 extra in feed costs. In this scenario, additional 242.7 tons of meat was sold per year compared with the TPPM base farm resulting in an increase of €355,532 in finisher pig sales. Dead animal disposal costs increased by 6.6% and manure handling costs increased by 3.2% in the EXTRA ROOM scenario compared with the TPPM base farm. Additionally, electricity, insurance, repairs, and interest loan repayments and depreciation charges increased by 8.3%, 10.9% 10.9% 14.6%, and 5.9%, respectively. Return on investment increased to 6.69% and current ratio increased to 0.45. Debt to asset ratio (0.42), equity to assets ratio (1.42), and debt to equity ratio (0.30) for the EXTRA ROOM scenario were similar to those for the TPPM base farm.

Table 8.

Difference in trading profit and loss accounts between Teagasc Pig Production Model base farm and 2 alternative farm changes scenarios: 1) construction of extra finisher accommodation to increase target body weight at sale up to 125 kg (EXTRA ROOM) and 2) installation of a new feed bin to provide finisher feed from 12 wk of age instead of 14 wk of age (EARLY FINISHER)

	€/year		€/pig produced		€/kg meat
	EXTRA ROOM	EARLY FINISHER	EXTRA ROOM	EARLY FINISHER	EXTRA ROOM	EARLY FINISHER
Sales
Finisher pigs	+355532.0	0.0	+18.04	0.00	0.00	0.00
Culled sows	0.0	0.0	0.00	0.00	0.00	0.00
Total income	+355532.0	0.0	+18.04	0.00	0.00	0.00
Variable costs
Gestation feed	−246.8	−12.9	−0.01	0.00	−0.01	0.00
Lactation feed	−173.9	−9.1	−0.01	0.00	−0.01	0.00
Creep feed	0.0	0.0	0.00	0.00	0.00	0.00
Link feed	0.0	0.0	0.00	0.00	−0.01	0.00
Weaner feed	−2763.8	−95655.9	−0.14	−4.85	−0.02	−0.06
Finisher feed	+185104.1	+70129.6	+9.39	+3.56	+0.04	+0.04
Replacement gilts	−36.2	−12.0	0.00	0.00	−0.01	0.00
Dead animal disposal	+838.3	0.0	+0.04	0.00	0.00	0.00
Healthcare	0.0	0.0	0.00	0.00	0.00	0.00
Reproduction	−2.6	−1.8	0.00	0.00	0.00	0.00
Manure handling	+531.8	0.0	+0.03	0.00	0.00	0.00
Transport	0.0	0.0	0.00	0.00	0.00	0.00
Total variable costs	+183250.9	−25562.1	+9.30	−1.30	−0.03	−0.02
Fixed costs
Admin and accounting	0.0	0.0	0.00	0.00	0.00	0.00
Electricity, heating and light	+6603.7	4.0	+0.34	0.00	0.00	0.00
Insurance	+2232.1	+98.0	+0.11	0.00	0.00	0.00
Repairs	+2232.1	+98.0	+0.11	0.00	0.00	0.00
Environment	0.0	0.0	0.00	0.00	0.00	0.00
Labor	0.0	0.0	0.00	0.00	−0.02	0.00
Loan repayments—of which interest	+11157.8	+499.6	+0.57	+0.03	0.00	0.00
Depreciation charges	+10271.3	+608.4	+0.52	+0.03	−0.01	0.00
Total fixed costs	+32497.0	+1308.1	+1.65	+0.07	−0.04	0.00
Total costs	+215747.9	−24254.0	+10.95	−1.23	−0.06	−0.02
Net profit	+139784.1	+24254.0	+7.09	+1.23	+0.06	+0.02

The EARLY FINISHER scenario used 382 tons of weaner feed less and 415 tons more of finisher feed than the TPPM base farm leading to a reduction of €25,526 in feed cost per year. There was no difference in the number of kg of meat sold, and therefore, there was no difference in total income between the EARLY FINISHER and the TPPM farm-based scenario. Additionally, return on investment (4.87%) increased while current ratio (0.35), debt to asset ratio (0.42), equity to assets ratio (1.42), and debt to equity ratio (0.29) were similar to those from the TPPM base farm.

Risk Analysis

Results from the risk analysis showed that when the farm applied the EXTRA ROOM scenario changes, annual mean net profit increases by 70.1 % while changes in the EARLY FINISHER scenario increased mean annual net profit by 13.9% (Figure 4). The 90% confidence interval (5–95%) was €119,606 to €275,539 for the TPPM base farm, €246,320 to €426,809 for the EXTRA ROOM scenario, and €146,685 to €303,590 for the EARLY FINISHER scenario.

Figure 4.

Box and whisker plots of the annual mean net profit, 5th and 95th percentile and the interquartile ranges of the TPPM base farm compared with to alternative changes in technical performance: 1) construction of extra accommodation to increase live weight at sale by 15 kg (EXTRA ROOM) and 2) a change in feeding practices by providing finisher feed from 28 kg of body weight compared to over 38 kg of body weight (EARLY FINISHER).

Price per kg of meat, number of piglets born alive per litter, and number of litters per sow per year were the three main variables contributing to variation in annual mean net profit in the TPPM (86.1%, 2.7%, and 2.7%, respectively), EXTRA ROOM (86.7%, 2.7%, and 2.7%, respectively), and the EARLY FINISHER (85.6%, 2.8%, and 2.8%, respectively) scenarios.

DISCUSSION

The TPPM was developed with the animal as a unit and then scaled to simulate the entire herd. Data used to develop the TPPM originated from results published elsewhere in the scientific literature, the Teagasc e-Profit monitor database, and information from the Teagasc pig advisory team. Different aspects of production were described in the model based on the available data; however, some limitations were encountered while trying to build certain sections of the model. For example, there are no available data regarding work organization and labor requirements in Irish pig farms. Therefore, due to the lack of available information, values for the labor section of the TPPM were agreed upon consulting the members Teagasc Pig Advisory team. Although this gave a good estimate of labor costs during the model validation, it is understood that Irish pig farms work organization and productivity are by far much more complicated than what has been incorporated into the model. Based on the receipts for 20 farms that were used to validate the TPPM, it is likely that some farmers under-report labor costs. More detailed information is required for a more optimal parameterization of the labor section of the TPPM and future studies should be designed to characterize work organization and labor in Irish pig farms. Similarly, it was decided to simulate a high-health status farm as there was no detailed information on average disease status for Ireland at the time the TPPM was developed.

TPPM Physical and Economic Outputs

The pig production system described in the TPPM is representative of the predominant intensive farrow-to-finish pig-producing farms in Ireland with large herd sizes, weekly farrowing batches, and several animal categories with different infrastructure and feeding practices. Physical outputs such as feed usage per pig and number of pigs produced per sow per year were similar to those reported in the Teagasc National Pig Herd Performance Report for 2016 (Teagasc, 2017). Feed usage was lower in the TPPM compared with the average feed usage for 13 European countries reporting to InterPIG (2017). Additionally, number of piglets sold per sow per year was 1.6 pigs more in the TPPM than the average number of pigs sold per sow per year for 13 European countries reporting to InterPIG (2017).

It is difficult to compare the TPPM financial indicators such as return on investment, current, and solvency ratios; however, a rate of return, which is higher than the opportunity cost of capital, is required before it is financially justified to carry out an investment (Power et al., 2009). The opportunity cost of capital is determined by the best alternative where the funds could be invested instead of the pig production, i.e., the return on pig farm investment should be at least as high as return expected from this alternative investment. When external funding is used, the return should also exceed the cost of external capital. Simulated return on investment figures did not exceed the minimum hurdle rate suggested by Meier and Tahran (2007) of 3% to 7.5% above the cost of funds.

Model Validation

The TPPM was validated using a similar approach to the one used by Shalloo et al. (2004) and Bohan et al. (2016) during the development of similar bio-economic models for the Irish Dairy and Sheep industries, respectively. Results from the validation using actual data showed the capabilities of the TPPM to realistically represent Irish pig farms. Physical outputs were similar between the TPPM and the average for the ePM farms; however, the TPPM produced 1,154 more pigs than the mean number of pigs sold from the 20 ePM farms. This discrepancy likely arose because the animals sold during the first 5 to 6 mo of the year from the 20 ePM farms were born in the previous year (i.e., 2015), where annual mean number of litters produced per sow per year and number of piglets born alive per litter were lower than in 2016 (Teagasc, 2017).

The TPPM simulated lower variable costs, mainly due to costs associated with gestating and lactating sow feed. Feed usage was similar but the price per ton for the sow diets formulated within the TPPM was cheaper than the prices reported by the ePM farms. Although there are no data available regarding ranges for nutrient specification in gestating and lactating diets in Ireland, it is possible that farmers provide nutrient values higher than those specified in the NRC (2012) increasing feed prices. Indeed, most commercial diets are formulated to exceed nutrient requirement recommendations in an effort to account for variation in nutrients concentrations and bioavailability (Flohr et al., 2016), and it is reported that in countries like the United States, pig producers consistently use higher SID lysine concentrations than the those recommended by the NRC (2012) in the breeding herd (Calderón Díaz et al., 2015).

Non-feed variable costs were higher for the TPPM than the average of the 20 ePM farms mostly due to the on-farm replacement gilts cost which is not recorded on the ePM system. Healthcare cost was 3.6 times lower for the TPPM compared with the average healthcare costs for the 20 ePM farms as it was considered that the TPPM had a high health status which might not have been the case for all the ePM farms. The TPPM predicted lower total sales income for the farm compared with the 20 ePM farms. In the TPPM, sales are calculated by multiplying kg of meat produced by meat price and does not consider any premiums or long-term contracts with processing plants that would lead to higher income as some pig farmers in Ireland do. It was not possible to discern if or how many of the ePM farms use for the TPPM validation have such agreements with the processing plants due to the confidential nature of the data used.

Model Application

The 2 scenarios used to demonstrate the TPPM applicability were chosen based on practices discussed by Irish farmers. This is important because approximation to reality is what makes bio-economic models applicable (Annetts and Audsley, 2002). Unsurprisingly, both the EXTRA ROOM and the EARLY FINISHER scenarios improved net profit. Total costs were higher in the EXTRA ROOM scenario compared with the TPPM base farm. Besides increased feed costs due to increased feed usage, other costs such as dead animal disposal and manure handling also increased in the EXTRA ROOM scenario. Increase in dead animal disposal costs is explained by the fact that farmers pay by kg of body weight and therefore, removing heavier dead carcasses from the farm would be more costly. Similarly, by staying 2 wk extra on the farm, finisher pigs produced more manure increased manure handling costs. Increase in electricity, insurance, and repairs costs and increased loan interest and depreciation payments are associated with the increase in pig spaces required for the EXTRA ROOM scenario. Other costs such as labor did not vary compared with the TPPM base farm as labor was calculated based as 1 farm employee per 154 sows and because keeping pigs for 2 wk more on farm at the end of the finisher stage would not greatly increase labor requirements as labor intensive activities such as vaccinations are not performed at that stage.

Nonetheless, in spite of the increased in variable and fixed costs, the EXTRA ROOM generated more kilograms of meat increasing income per pig resulting in an increased profit. This resulted in an increased return on investment even when the EXTRA ROOM scenario was associated with greater capital investment such as the new buildings and equipment. Also, this scenario had a greater spread in the net profit as shown by the 90% confidence interval of the annual mean profit during the risk analysis indicating that greater risk is associated with the EXTRA ROOM scenario. However, a limitation of this scenario is the fact that possible effects on meat quality were not taken into consideration as currently there is no paying scheme based on carcass quality characteristics in Ireland. Nonetheless, increasing body weight at sale is associated with a deterioration on quality by increasing fat content and decreasing pork tenderness (Ellis et al., 1996; Beattie et al., 1999; Latorre et al., 2004), and thus, financial gains associated with the EXTRA ROOM scenario could be over-estimated. Future studies are needed to investigate implications of increasing body weight at sale for carcass traits under Irish conditions and to explore the impact that such meat quality changes could have on pork prices and farm profitability.

The improved net profit in the EARLY FINISHER scenario was associated with a reduction in feed cost during the weaner stage 2 and finisher stage. Weaner feed is more expensive than finisher feed as it has higher nutrient concentration as the animal requires more protein (i.e., amino acids) per kilogram of body weight to deposit muscle. Although dietary amino acids concentrations for the weaner stage 2 diets are in line with animal requirements during the first weeks animals receive this feed, it is likely that by the end of the third week in the weaner stage 2, amino acids such as SID lysine are being over supplemented (Symeou et al., 2016). A similar situation arises when finisher pigs are provided with the same dietary specification during the long finisher period. Therefore, by providing weaner stage 2 pigs with finisher diets and by providing a second finisher diet, dietary nutrient concentrations are adjusted to the pigs’ nutritional requirements decreasing feed costs. The only extra capital investment in the EARLY FINISHER scenario compared with the TPPM base farm was the acquisition of a new €10,000 long-term bank loan to finance a new feed bin. Because there was a substantial increase in farm profitability associated with this small investment, the return on investment was very large. This indicates that farmers trying to improve profitability with minimum investment or with no access to large amounts of credit could implement changes similar to those in the EARLY FINISHER scenario without substantially increasing risk or affecting farm liquidity and solvency.

Risk Analysis

Risk analysis assesses the effect of changes in inputs taking into account probability distribution of the stochastic variables and the associations among them (Groenendaal, 1995). The wider 90% confidence interval associated with the EXTRA ROOM scenario could be partially attributed to its associated greater variable and fixed costs and the need of higher income to cover them. The 90% confidence interval for the EARLY FINISHER scenario was similar to the TPPM base farm and indicates similar risk between both scenarios as there was a low extra capital investment and animals were still provided with feed that met their nutritional requirements and thus production was not affected. The main variable contributing to mean net profit variance was pig prices (over 85% in all scenarios).

In conclusion, the TPPM simulates biological and economic performance of a farrow-to-finish pig farm with weekly farrowing batches. The model was parameterized using real Irish data from several sources. Some limitations were encountered while building the model mainly associated with the lack of data on work organization and labor requirements, and thus, availability of more detailed information would greatly improve model parameterization. Results from the validation showed that the TPPM is able to represent the performance of a farrow-to-finish Irish pig farm. Thus, the TPPM simulated results can be used with confidence to evaluate the impact of real life scenarios on farm productivity and profitability. Stochastic budgeting using Monte Carlo results suggests that a change in feeding practices could be a better option for farmers looking to improve profit without substantially increasing risk or affecting farm liquidity and solvency.

TPPM-simulated results could be used to facilitate decision making to address the challenges that Irish pig farmers face on a daily basis. A model extension which takes into account stochasticity related to disease outbreaks would be a natural continuation for the TPPM.