PRODUCTION, MODELING, AND EDUCATION Economic impact of decreasing stocking densities in broiler production in Belgium A. Verspecht,*1 F. Vanhonacker,* W. Verbeke,* J. Zoons,† and G. Van Huylenbroeck* *Ghent University, Department of Agricultural Economics, Coupure Links 653, 9000 Ghent, Belgium; and †Provincial Centre for Applied Poultry Research (Province of Antwerp), Poiel 77, 2440 Geel, Belgium ABSTRACT Stocking density is a prominent topic in public debates on animal welfare and was one of the reasons for the European Commission to set limits to the stocking density on broiler farms. The objective of this paper was to calculate in detail the financial impact of changes in technical and management variables due to decreasing stocking densities in line with new European Union (EU) regulations. Therefore, the productive performance indicators such as BW, mortality, or feed conversion and farm technical data such as water consumption and heating of 3 independent experiments conducted at a poultry research station in Flanders (Belgium) were combined. Using the par-
tial budget technique only those elements that change with stocking density have been taken into account. Reducing stocking density implies a recalculation of all costs on a reduced number of birds. This yields an economic situation that leaves hardly any profit margin for most of the broiler producers under the present market conditions. It was found that the critical threshold of stocking density for maintaining profitability under the present market and technical conditions is around 46 kg/m2, thus well above the EU maximum of 42 kg/m2. It is shown, however, that with changing broiler feed and meat prices, the impact might be less negative in economic terms.
Key words: stocking density, broiler performance, economic analysis 2011 Poultry Science 90:1844–1851 doi:10.3382/ps.2010-01277
INTRODUCTION Animal welfare is an increasingly relevant issue that is a source of lively debate in today’s industrialized societies. Citizens ask for more animal-friendly livestock production practices (Vanhonacker et al., 2010), and the market segment that takes animal welfare into account during food purchasing is steadily growing. The focus on maximizing productivity, production efficiency, and profitability led to the development of intensive animal rearing conditions (e.g., indoor housing to maximize production within the minimum space). This intensive animal production is now increasingly criticized for its trade-off in terms of farm animal welfare. The urbanized and lay public strongly perceives animal welfare from a human-centered perspective, in which attention to “natural living” occupies a primary role (Vanhonacker et al., 2008). From this perspective, high stocking densities applied to maximize profit per unit area result in negative general welfare perceptions and evaluations of animal production systems among ©2011 Poultry Science Association Inc. Received December 3, 2010. Accepted April 22, 2011. 1 Corresponding author:
[email protected]
the public, in particular within the poultry production sector (European Commission, 2005; Castellini et al., 2008; Vanhonacker and Verbeke, 2009). However, from a scientific point of view no consensus has been reached on the range in which stocking density affects poultry welfare (Buijs et al., 2009). Negative associations of high stocking densities on broiler health (Estevez, 2007) and on broiler behavior (Spinu et al., 2003), as well as societal demand for more welfare-friendly production practices, were direct reasons for the establishment of guidelines or limits on stocking density in animal production through legislation (Estevez, 2007). The European Union (EU) Council Directive 2007/43/EC specifies a maximum stocking density for broilers of 33 kg/ m2 from June 2010 on. If a broiler producer complies with certain good farming practices, such as technical requirements and the registration of data on feeding, heating, ventilation, and disinfection, the maximum stocking density can be increased to 39 kg/m2 and up to a maximum of 42 kg/m2 if additional conditions regarding monitoring and mortality can be guaranteed. Small changes in stocking density can strongly influence the zootechnical performance of broilers (Dijkhuizen and Morris, 1997). Growth rate and food conversion ratio of broilers were shown to be lower at higher stocking densities (Cravener et al., 1992; Al-Homidan
1844
ECONOMIC IMPACT OF DECREASING STOCKING DENSITIES
et al., 2003; Dozier et al., 2005), although the negative performance impact from high stocking densities can be counterbalanced by improved and adjusted housing conditions and management practices (Bessei, 1993; Feddes et al., 2002; Dawkins et al., 2004; Jones et al., 2005). From an economic perspective, higher welfare standards are in general inversely related to zootechnical performance or productivity (McInerney, 2004). Economic benefits per square meter are under most conditions still higher if the broilers are stocked more densely (Cravener et al., 1992; Feddes et al., 2002). Broiler production systems delivering higher standards of farm animal welfare induce production costs that are significantly higher than those of conventional systems, which is the result of higher factor input costs (labor and feed in particular) and lower overall productivity due to the reduction in stocking density (SCAHAW, 2000). Farm-level decisions about stocking densities have been primarily driven by cost-effectiveness. However, economic profit may come at the cost of reduced bird performance, health, and welfare if densities are excessive (Estevez, 2007). Yet stocking densities for broilers vary greatly between countries and husbandry systems (SCAHAW, 2000). In addition, lower stocking densities are usually not supported by primary producers, given the tight profit margins per bird they currently face in the open and global poultry market. Questions about the eventual economic implications of the EU Council Directive 2007/43/EC have given rise to various calculations in different EU member states [e.g., Sheppard and Edge (2005) for the United Kingdom; van Horne (2005) for the Netherlands; Gabriels and Van Gijseghem (2006) for Belgium]. Some of these studies included an estimation for the variation in BW and feed intake as a function of stocking density (van Horne, 2005; Gabriels and Van Gijseghem, 2006). However, other costs such as water and ventilation, for example, were kept constant. The UK study calculated the change in costs and output at an aggregate farm level, whereas the Dutch and Belgian studies performed the calculations per broiler. This affects the accounting of the building and equipment costs because such costs will be higher per broiler in case of lower stocking densities. van Horne (2005) reported a 50% reduction of farm income for Dutch broiler producers when decreasing the standard stocking densities (around 45 kg/m2) to 38 kg/m2. A stocking density of 30 kg/m2 resulted even in a negative farm income (i.e., economic loss instead of profit). In Belgium, farm income was reported to decrease with 46 and 74% with reducing densities from 46 kg/m2 to 38 and 30 kg/m2, respectively. Calculations for the United Kingdom yielded a reduction in total production with 21% and an increase in cost of 5.6% when changing stocking densities from 38 to 30 kg/m2 (Sheppard and Edge, 2005). The objective of this study was to calculate in detail the economic impact of reduced stocking densities. To the authors’ knowledge, empirical data about the
1845
consequences of stocking densities on different technical parameters have not been reported previously. The present study applies a unique data set to investigate the tradeoff between stocking density and farm economic profitability in the region of Flanders, the northern Dutch-speaking part of Belgium. Broiler production in Flanders in 2009 amounted to 17 million units produced on about 650 broiler farms (FPS Economy, 2010b), which is more than 80% of the total Belgian broiler production. This study uses secondary data, complemented with primary data from 3 independent on-farm experiments in a partial budget analysis. The approach is original and different from other available studies because it takes into account a more comprehensive set of technical parameters influenced by stocking density. The results are relevant to support debates on farm management, legislation, and policy implementations related to broiler production.
MATERIALS AND METHODS Data Sources Data from Stocking Density Experiment. Primary data originate from 3 independent experiments conducted at a poultry research station under on-farm conditions (Provincial Center for Applied Poultry Research, Province of Antwerp, Geel, Belgium). A first experiment in 1996 tested the impact of 4 different stocking densities on the technical performance of broilers. A second experiment was conducted in 2001 to investigate the role of ventilation on litter quality for 2 different stocking densities. The objective of the third experiment, which started in 2008, was to measure the effect of stocking density and light intensity on the performance of broilers. The 3 experiments differed in stocking density ranges, duration, number of rounds, number of birds per round, breed, and overall objective of the study (Table 1). The experimental setting from 2008 differed from the previous experiments because the feeding and watering system was not adapted to stocking density, whereas in 1996 and 2001 more drinking cups and feeding pans were provided at higher densities. The research station has 2 poultry stables that each have 2 compartments of about 300 m2. In the different rounds, stocking densities changed in the compartments to rule out technical differences between compartments. During the experiments, data were collected on individual weight, feed intake, feed conversion, and mortality (which will be used as productive performance indicators in the present economic analysis) and on energy use for ventilation and lighting, water consumption, and heating (which will be used as farm technical data in the present economic analysis). Data from the different experiments could be compared because each experiment included data for a stocking density of 20 birds per square meter (which was considered as the reference situation). Because of
1846
Verspecht et al.
Table 1. Characteristics of the 3 experiments Year
No. of birds/m2
Duration (d)
No. of rounds
Surface (m2)
No. of initial birds/round
Breed
Density (kg/m2)
1996 2001 2008
12 18 20 22 16 20 13 20
42 42 42 42 41 41 39 39
6 6 6 6 6 6 4 4
300 300 300 300 600 600 600 600
3,600 5,400 6,000 6,600 9,600 12,000 7,800 12,000
Ross Ross Ross Ross Ross Ross Ross Ross
27.0 39.1 43.2 47.1 36.9 45.3 32.7 48.4
the differences in year, duration, and breed between the different experiments, the 20 birds per square meter do not match exactly with the same number expressed as kilogram per square meter. For the years 1996, 2001, and 2008, corresponding stocking densities for 20 birds per m2 were 44.4, 47.0, and 48.4 kg/m2, respectively. Depreciation Costs on Building and Equipment. Additional primary data on costs or depreciation for buildings, equipment, and machinery were collected through telephone interviews (data from 2009) with some major construction companies active in the Flemish poultry business. These interviews revealed that costs for construction are independent of stocking density. Yet for some equipment such as ventilation or watering systems, less capacity is needed for lower stocking densities. Differences in costs for stables with 12, 16, or 20 birds per square meter were estimated by the construction companies to be about 10 to 14% lower for 12 vs. 20 birds per square meter, and about 3 to 7% lower for 16 vs. 20 birds per square meter. Installers mentioned that the geometry of the stable is of greater importance regarding the number of ventilators, drinking lines, and cups or feed pans than the stocking density. Most basic elements of the construction and equipment, including labor costs, are independent of the stocking density that will be applied. Replacement values for buildings are set at 195€/ m2 with a duration of 30 yr. Equipment has a replacement value of 95€/m2 with a duration of 15 yr (KWINV, 2007). Taking into account the reduction in cost at
208 208 208 208 308 308 308 + Cobb 500 FF 308 + Cobb 500 FF
smaller densities, this results in a range of 83 to 97€/ m2 replacement value for the equipment. Other Technical and Financial Data. Complementary secondary data are used in the calculations of the performance indicators and costs (see Table 2). These include zootechnical information, stable dimensions, energy and water need, and market prices used in the calculation of the performance indicators and the partial budget analysis. These numbers are based on an average broiler stable in Flanders with a common stocking density of 48 kg/m2.
Regression Analysis on Technical Performance Indicators Mortality rate, final weight of the broiler, feed intake, feed conversion, water consumption, energy use, and heating are expected to vary with different stocking densities (Shanawany, 1988; Cravener et al., 1992; Pettit-Riley and Estevez, 2001; Dozier et al., 2005; Estevez, 2007). Regression analyses will be performed to express these parameters as a function of stocking density (kg/m2) including a year effect by introducing 2 dummies. Because parameters (e.g., productivity per bird) changed over time, absolute parameters from the different years cannot simply be used. It is necessary to find a regression function that enables an estimation for different stocking densities compared with a certain parameter for year 2011 on Belgian poultry farms.
Table 2. Input data for calculation of technical indicators and costs1 Technical data
Amount
Price
Stable surface (m2) Production time (d) No. of rounds/yr Average weight (kg) Mortality (%) Feed consumption per bird (kg) Water consumption per bird2 (L) Water for cleaning2 (L/m2) Heating4 (kWh/bird) Energy4 (kWh/bird) Litter (kg/m2)
1,400 40 7.4 2.4 3 4.0 6 20 0.17 0.11 1.2
Price of fodder (c€/kg) Meat price (€/kg) Price of water (c€/L) Price of litter (c€/kg) Cost disinfection stable (c€/bird) Manual catching cost (c€/bird) Manure disposal cost3 (€/T) Day-old chick cost (c€/bird) Health costs (c€/bird)
1Source:
De Baere and Zoons (2007) except where otherwise noted. Vermeij et al. (2007). 3Source: Lemmens et al. (2007). 4Source: Snoek et al. (2000). 2Source:
Amount
29.9 0.8 0.15 17.5 1.2 3 23 31.5 4.99
ECONOMIC IMPACT OF DECREASING STOCKING DENSITIES
Therefore, results of each round are expressed as the ratio of a variable (e.g., feed intake) for a certain stocking density compared with that variable for the reference stocking density (i.e., 20 birds per m2) of that round. The regression analysis of the data from these 3 different experiments is not straightforward, and some assumptions had to be made. The proposed method assumes that i) the effect itself of stocking density is constant over time (there is only one slope possible for the different experiments); ii) the ratios show the same functional form as the absolute variables; iii) stocking density expressed in kilograms per square meter is more relevant than the number of birds per square meter; and iv) differences in breeds, health of breeder hens, chick quality, breeding techniques, fodder composition or experimental design, and objectives are sufficiently counteracted by introducing a year effect in the regression analysis. The regression function is a linear regression model using ordinary least squares. The significance level was 0.05. Data were analyzed using SPSS 17.0 (SPSS Inc., Chicago, IL). The resulting equations will be the basis for the estimation of impact of the technical performance indicators at the different stocking densities in the partial budget analysis. The boundaries with respect to stocking density are defined by the experiments and set on 12 birds per square meter (about 27 kg/m2) and 22 birds per square meter (about 48 kg/ m2). Absolute values are obtained by multiplying the ratios with a normal on-farm value of the 20 birds/m2 reference situation.
Partial Budget Analysis The impact of stocking density on farm profitability is calculated with the use of the partial budget technique. A partial budget analysis contains only those elements or parameters of a complete budget (or farm budget) that will change if a proposed modification (in this case stocking density) is implemented (Dijkhuizen and Morris, 1997). The aim is to estimate the change that will occur in terms of farm profit or farm loss from some change in the farm management system (Boehlje and Eidman, 1984). Partial budgeting is based on the principle that a small change in the organization of a farm or ranch business will have one or more of the following effects: eliminate or reduce some costs, eliminate or reduce some returns, cause additional costs to be incurred, or cause additional returns to be received (Dalsted and Gutierrez, 2010). The net effect will be the sum of positive economic effects minus the sum of negative economic effects. It is important to mention that the final outcome of this technique is a net effect (i.e., it gives a measure for the possible change in farm profit). In the literature, this method is often used to calculate changes in farm management, change of crops, or the introduction of a new technology (Ehui and Rey, 1992; Engle and Brown, 1999; Overton, 2005; Allen,
1847
2006; Nahamya et al., 2006; Cox et al., 2009). The unit of analysis in these studies is the farm or a farm unit (e.g., stable). To analyze the effect of stocking densities, however, it is important to include investment costs. The investment costs are equal per stable but are different per animal (i.e., investment costs per animal increase with lower stocking densities). Therefore, the partial budget analysis will be performed per unit of broiler. As a result, also all other variables will be calculated per unit of bird. This enables us to investigate in detail the effect of stocking density on the productive performance indicators and on-farm technical data. In the final stage of analysis, the profit per bird will be reconverted to a profit per square meter. This will result in a more comprehensive, realistic, and detailed description of the impact on farm profit. The difference between the farm profit found with the partial budget analysis and the total farm profit is small because only a few costs do not vary per bird (e.g., day-old chick cost), so the total farm profit will also be calculated.
Calculation of Partial Budget Profit and Total Farm Profit The partial budget profit is calculated as the partial farm output per unit of bird minus the partial costs per unit of bird that change when changing stocking densities. Because it is assumed that final weight is dependent on stocking densities, partial output equals total output: partial output = total farm output = final weight × meat price. The partial costs only include the costs that differ per bird at the different stocking densities. Those are costs for feed, water, litter, energy, heating, catch costs, disinfection, manure disposal, depreciation, and maintenance of buildings and equipment. The total farm profit is total farm output minus total costs. Total costs include, apart from the partial costs, other costs that are constant per bird, such as chick cost, health cost, insurance, and rent on animals. Profits, outputs, and costs are mentioned per bird and converted afterward to farm profit per square meter.
Economic Impact of Price Variations To illustrate the impact of possible market price variations, different price change scenarios are taken into account. Broiler meat prices and broiler feed prices are interlinked, which can be illustrated by historical figures of Belgian market prices. The exact relation between broiler feed and broiler meat prices is calculated based on a time series (IRIMA) and ordinary least squares regression analysis similar as for the technical performance indicators. Broiler feed and meat are indices with base year 2000 where feed or meat price of 2000 are set as index 100. The interdependency of these 2 prices is taken into account in the impact analysis of
1848
Verspecht et al.
Table 3. Descriptive statistics of the data set from the stocking density experiments Item1 % Mortality Water per bird (L/bird) Feed intake per bird (kg/bird) Final BW (g) Energyratio Heatingratio
No. of rounds
Minimum
Maximum
Mean
SD
52 52 52 52 52 46
1.31 5.81 3.47 2,104.40 0.80 0.83
7.12 7.35 4.24 2,662.30 1.13 1.50
2.94 6.53 3.79 2,349.91 0.97 1.05
1.29 0.43 0.20 147.84 0.08 0.14
1Subscript “ratio” indicate parameter is expressed as the ratio of a variable for a certain stocking density compared with that variable for the reference stocking density (i.e., 20 birds per m2) of that round.
market price variations. Three different price changes are simulated.
RESULTS Equations on Performance Indicators and Farm-Technical Data Table 3 gives an overview of the descriptive statistics of the independent variables of the regression analysis. The number of rounds in the 3 different years is 52. For heating, there are only 46 observations due to the occurrence of technical problems during registration. Table 4 presents the equations for the productive performance indicators and farm-technical variables as a function of stocking density, relative to the reference situation of 20 birds/m2. With the exception of heating and mortality, all parameters are significantly related to stocking density. Heating costs and mortality rates will be kept constant over the considered range of stocking densities. For mortality it is possible that other elements such as genetic differences, problem of diseases with mother animals, and differences in chick quality play a more important role. Final weight, feed intake, water intake, and energy use were explained by the differences in stocking densities and by the differences in the different experiments. Between 55 and 63% of the variation in the variables is explained by the regression equations. Final weight and feed intake decrease with higher stocking densities. As productivity of broilers increased during the years, final weight and feed intake is higher in 2001 than in 1996 and higher in 2008 than in 2001. Water intake increases with stock-
ing density and seems to be higher for the earlier experiments. The equation for the consumption of water includes the number of cups to control for the different settings in the 3 experiments regarding the amount of cups. Energy consumption increases with higher stocking densities due to more ventilation.
Partial Budget Analysis Results In Table 5 the results of the partial budget analysis are given for stocking densities in the range of investigation (i.e., from 33 through 48 kg/m2). These densities include the densities specified in the EU Council Directive and are within the boundaries of the 3 on-farm experiments. The highest density of 48 kg/m2 is presently a common density among Flemish broiler farms. The total farm profit contains costs that are independent of stocking densities (e.g., chick cost and health care costs). Results are mentioned per bird and converted afterward to farm profit per square meter. The partial budget impact of different stocking densities is within the range of 4.0 to 7.5 €/m2 per production round. When taking all costs into account, the total farm profit per year decreases with 7.3 €/m2 from the highest to the lowest densities covered in the study. Farm profit is only positive from densities above 46 kg/ m2.
Economic Impact of Price Variations Historical figures of Belgian market prices show an interaction between chicken feed and meat prices. Both prices follow more or less similar evolutions (Figure 1).
Table 4. Performance and technical data as a function of stocking density (StD) and year Production parameter1
Equation
% Mortalityratio Final weightratio Feed intakeratio Water per birdratio Energyratio Heatingratio
0.727(**) 1.161(**) 1.118(**) 1.096(**) 0.602(**) 1.127(**)
+ − − − + −
0.006(NS) × StD − 0.042(NS) × A + 0.057(NS) × B 0.003(**) × StD − 0.031(**) × A − 0.021(**) × B 0.002(**) × StD − 0.025(**) × A − 0.019(**) × B 0.009(**) × cup + 0.002(**) × StD + 0.075(**) × A + 0.052(**) × B 0.008(**) × StD + 0.049(**) × A + 0.038(**) × B 0.002(NS) × StD + 0.014(NS) × A − 0.003(NS) × B
Significance (F-statistic)
R2 (adjusted)
0.052
