Commercial Approaches to Genetic Selection for Growth and Feed Conversion in Domestic Poultry D. A. EMMERSON1 Campbell Soup Company, P.O. Box 719, Farmington, Arkansas 72762 Received for publication July 8, 1996. Accepted for publication February 25, 1997. 1Present address: Department of Animal and Poultry Sciences, Litton Reaves Hall, Virginia Tech, Blacksburg, VA 24061. ABSTRACT Tremendous genetic progress has been observed historically for growth and feed conversion through the efforts of the primary breeding companies. However, significant between-strain variation still exists due to differences in selection emphasis and selection techniques practiced by these organizations. This paper provides an overview of methods currently employed in commercial poultry breeding with reference to factors complicating program design and future challenges facing the industry. Mass selection for body weight has resulted in a significant reduction in the number of days required to grow bird to market weight with indirect improvements in feed conversion. Direct selection for feed conversion is accomplished through part record testing of males that have been preselected for body weight, conformation, and defect traits. Data are commonly subjected to complex statistical analysis both to correct feed conversion for variation in body weight and to improve the accuracy of breeding value estimates. Feed conversion breeding values of male sibs are sometimes used for the selection of female candidates as well. Selection for growth rate and efficiency has resulted in negative complications, such as ascites, reduced reproductive performance, skeletal abnormalities, and increased carcass fatness. Some of these factors may be partially ameliorated through modified selection practices. If not addressed by the breeding industry, the disruption of physiological homeostasis might ultimately represent economic and genetic barriers to further progress in improving growth and efficiency. Modern techniques in molecular genetics, utilized in conjunction with traditional quantitative genetic approaches, will provide additional opportunities to circumvent these physiological complications associated with genetic selection for growth and feed efficiency. (Key words: genetics, growth, body weight, feed conversion, feed efficiency) HISTORICAL IMPROVEMENT OF GROWTH AND FEED CONVERSION Havenstein et al. (1994a) compared broiler performance of a 1957 randombred control population with that of a 1991 broiler strain cross in two different nutritional environments. Improvements between 250 to 300% were observed in weight for age depending on the point of measurement (Figure 1). Feed conversion was also significantly improved at both a constant age and weight; however, the magnitude of improvement was greater for comparisons at a constant body weight (Figure 2). Interestingly, feed conversion was improved for the 1991 broiler in spite of a higher level of 1997 Poultry Science 76:1121 1125 abdominal and carcass fat that was observed for this strain (Havenstein et al., 1994b). Despite the marked historical improvement that has been observed for growth rate and feed conversion, significant within- and between-strain variations for growth and feed conversion and related traits still exist in contemporary commercial broiler strains. Data presented for growth, feed conversion, and related traits presented in Table 1 represent a summary of approximately 45 contemporary broiler crosses. These results demonstrate greater than 10% variation between broiler crosses for body weight, growth rate, and feed conversion, and over 30% variation for abdominal fat pad weight. Viewed from another perspective, a 5-d reduction in the growout period would allow a broiler producer to raise an additional broiler flock annually in every broiler house, whereas a reduction of 0.17 in feed conversion amounts to 1.25 /lb or more than a 5% reduction in live cost. This economically important variation reflects differences in both breeding goals and evaluation and selection methods practiced by the primary breeding companies. 1121
1122 EMMERSON TABLE 1. Summary of performance for body weight, days to 2,850 g, feed conversion, and percentage leaf fat of 45 contemporary commercial broiler strain crosses Worst Average Best Trait performance performance performance Deviation (%) 52-d BW, g 2,701 2,838 3,001 11 Days to 2,850 g, d 54 52 49 10 Feed conversion, g:g 2.10 2.01 1.93 9 Leaf fat, % 2.5 2.2 1.9 32 BREEDING GOALS FOR COMMERCIAL POULTRY Market requirements are somewhat different for commercial broiler and commercial turkey production and, consequently, different selection strategies are required. Although requirements are changing rapidly, the broiler industry is an age-for-weight market with slaughter at a fixed target weight, as resulting chicken products generally must fall within relatively narrow weight ranges. Birds that fall outside these prescribed ranges result in processing inefficiencies and some associated economic loss. Thus, the age at which broilers are marketed may change significantly throughout the year due to seasonal and other production factors influencing growth rate. The turkey industry, in contrast, is primarily a weight-for-age market with marketing at a fixed age. Turkey further processing involves significant restructuring of carcasses or parts to form final products. Consequently, precise control of bird weight is not as critical and turkey producers generally market birds at fixed ages with significant weight variation observed throughout the year. Broilers are also commonly marketed at an earlier point in development (30 vs 60% of sexual maturity for broilers and turkeys, respectively) than turkeys. Thus, broiler selection occurs at an earlier point in the growth curve, which places greater emphasis on rapid growth and early carcass development. Slightly different selection strategies are appropriate as a result of these subtle differences in breeding goals for commercial broiler and turkey production. INDUSTRIAL SELECTION FOR GROWTH Growth rate was the first trait to receive attention in the breeding industry during the emergence of commercial broiler production due to its economic importance and the relative ease with which it can be improved. Growth is moderately to highly heritable and can be rapidly improved through mass selection. However, growth is a dynamic process that involves both an increase in mass and the synchronous differentiation and maturation of many tissues (Siegel and Dunnington, 1987). Consequently, selection results are highly dependent on the particular methods used, including the age of primary selection, the intensity of selection, selection emphasis placed on correlated traits, and the environment (including nutritional aspects) under which selection is exercised. Intense growth selection has produced a number of complications, such as reduced reproductive performance, increased carcass fat, skeletal abnormalities, and ascites, resulting from a disruption of a physiological equilibrium (Chambers, 1990). This disruption of physiological homeostasis may in the future represent significant physiological or economic barriers to continued improvements in growth performance. Some of these complications have been partially ameliorated through modified selection schemes but all will FIGURE 1. Growth curves for male and female broilers from a 1991 commercial broiler strain cross and a 1957 randombred control population (adapted from Havenstein et al., 1994a). FIGURE 2. Feed conversion for male and female broilers from a 1991 commercial broiler strain cross and a 1957 randombred control population (adapted from Havenstein et al., 1994a).
SYMPOSIUM: GENETIC SELECTION STRATEGIES FOR THE FUTURE 1123 require greater emphasis in the future. Thus, the complex interaction of these many physiologically related factors complicates selection for growth rate. There is significant variation in methods of growth selection currently being practiced by commercial breeding companies. All programs are based on pure line selection at a fixed age that corresponds with either a commercial age or a commercial body weight. Preselection of candidates at a relatively immature stage prior to final selection at a commercially relevant age is common practice, especially in lines under extreme selection for growth rate. Selection environments vary from a simulated commercial environment to an optimal environment that allows full expression of genetic potential for growth. Commercial environments exclude disease challenge but allow evaluation of response to density, climate, social stress, and other physiological stresses. Mass selection is common, although many organizations are currently utilizing more complex methods for breeding value estimation, including family selection, single and multi-trait indices, and Best Linear Unbiased Prediction (BLUP). There are many industry examples of genotype by environment interactions under field conditions; however, little emphasis has been placed on developing genetic solutions to these environment- or managementinduced production problems. Marks (1993) has demonstrated the potential for genetic adaptation to selection environment following long-term selection of Japanese quail for body weight under low- and high-protein selection environments. Although a greater selection response was observed in a line (P) selected in a highprotein environment, this line developed a higher protein requirement and did not express its full genetic potential for growth when fed lower protein diets. The line selected for growth in a low-protein (T) environment performed equally well under a broad range of dietary protein levels. However, when lines P and T were reciprocally crossed, heterosis and even overdominance were observed for 4-wk BW in all protein environments tested (Marks, 1995). This result suggests that a strategy involving selection and crossing of complementary lines that have been selected under diverse environmental conditions might optimize genetic progress and still confer adaptability to a variety of environmental circumstances. INDUSTRIAL SELECTION FOR IMPROVED FEED CONVERSION Feed conversion is a complex, highly aggregate trait that is the net result of the interaction of many different component traits. Although some of these factors are identified in Figure 3, additional component traits can be identified if the trait is dissected to a higher level of complexity. As such, feed conversion represents a rather crude measure of biological efficiency and, consequently, selection for feed conversion influences these FIGURE 3. Schematic representation of underlying component traits contributing to feed conversion ratio. component traits in a relatively undirected manner. The relationship between body weight or growth rate and feed conversion has long been recognized. Indeed, much of the historical improvement in feed conversion can be attributed to increases in growth potential. This indirect approach to improvement of feed conversion is associated with negative traits such as increased appetite and carcass fatness. Direct selection for feed conversion was adopted in the 1970s to select for components of efficiency that were lowly or negatively associated with growth rate. However, the association between body weight and feed conversion continues to complicate selection for improved efficiency in industrial breeding programs. The largest single difficulty in selection for feed conversion is the apparent conflict between practical selection criteria and the actual breeding goal. The commercial goal of feed conversion selection is to reduce the amount of feed required to grow birds to a constant market weight. This goal favors fast-growing individuals, as they will reach market weight at an earlier age and, consequently, energy requirements for maintenance will make up a smaller percentage of total energy intake. However, feed conversion can only be practically evaluated for a fixed age period, which penalizes heavier individuals because their maintenance requirements are higher and energy consumption for maintenance makes up a larger percentage of feed consumed during the test period. The concept of residual feed consumption, commonly applied in caged layer breeding, can appropriately be applied to selection for efficiency in broilers and turkeys. Residual feed consumption refers to the amount of feed consumed after statistical adjustment of feed intake for variation in body weight or other energy requiring outputs (Wing and Nordskog, 1982). This adjustment effectively makes it possible to discretely partition selection for growth rate and the efficiency of feed utilization. There is great variability in industrial approaches to feed conversion selection. The most common approach involves individual testing of male candidates that have been preselected for body weight and other physical traits. Feed conversion is evaluated during a 1- to
1124 3-wk period beginning between 3 and 8 wk of age. Feed conversion results are generally standardized for body size using statistical procedures or based on estimates of energy requirements for maintenance. Males with superior feed conversions are selected through mass selection, sib selection, or index methods and, in some cases, female are selected based on the performance of their male sibs. Indirect selection methods, such as serum very low density lipoprotein concentrations, have been attempted but have generally not been utilized as successfully in industrial settings. EMMERSON Shifts in market requirements influence not only relative selection emphasis placed on a trait but also can change the way we view particular traits. For example, the rapid growth of the deboning segment of the broiler market not only increases the importance of meat yield, but also might make meat conversion (units feed required/unit meat produced) a more appropriate measure of economic efficiency in future breeding schemes. Anticipation of changes in market priorities is particularly critical considering the 3- to 5-yr delay from pedigree selection to field evaluation. FUTURE CONSIDERATIONS Population genetics theory indicates that genetic variation for growth and feed efficiency will diminish with continued selection. However, results from longterm selection experiments provide little evidence of long-term genetic plateaus, and suggest that plateaus are only temporary when they do occur (Marks, 1991). In addition, commercial breeding companies utilize relatively large populations and have additional tools available to recapture variation, such as outcrossing and the development of synthetic lines, should such plateaus develop. Negative traits associated with growth and feed conversion, such as ascites, sudden death syndrome, reduced immune competence, tibial dyschondroplasia, reduced reproductive performance, and other metabolic disturbances, might someday represent physiological or economic limits to future progress. Some correlated responses, such as reduced appetite, might also be considered positive or negative depending on circumstances. For example, whereas appetite is generally viewed as a negative trait due to its association with fatness and inefficiency, it can also be seen as a desirable trait for very efficient strains grown under high environmental temperatures. Genetic and nongenetic solutions to these homeostatic concerns and a more holistic or integrated approach to genetic improvement will be required if historic genetic gains are to be sustained in the future. Genotype by environment interactions have been largely disregarded in commercial poultry breeding. However, poultry breeding is becoming increasingly centralized, whereas commercial poultry production is growing worldwide. Genetic products developed in a finite range of environments are expected to meet the requirements of diverse markets, environments, and management systems throughout the world. This development is particularly important when one considers that broiler production is currently growing most rapidly in Central and South America and the Pacific rim, areas with climates that differ significantly from traditional chicken-producing areas where the modern broiler has been developed. Changes in market requirements also will have a significant impact on genetic development strategies. MOLECULAR GENETIC APPROACHES TO SELECTION FOR GROWTH AND FEED CONVERSION Traditional quantitative genetic approaches have been very successful in improving both growth and feed conversion of commercial poultry, with little evidence of reduced response. Molecular genetic tools are rapidly being developed that should augment methods currently being used in the primary breeding industry. Direct marker assisted selection for growth and feed conversion is possible; however, this approach is not likely to be more effective than current methods considering the high heritabilities observed for these traits. Marker assisted selection for specific component traits associated with growth and feed conversion might allow selection to be directed at particular growth and efficiency characteristics that are commercially desirable and might also help to circumvent some of the negative complications associated with genetic improvement of these traits. Marker assisted selection against physiological barriers, such as ascites, reduced disease resistance, and other metabolic disorders, could ensure that sustained genetic improvement of growth and feed conversion is economically feasible. Although intra- and interspecies gene transfer are theoretically possible, many technical, ethical, and consumer acceptance issues must be resolved. In addition, specific genes with significant influence on growth and feed conversion have not yet been identified as candidates for gene transfer. Indirect contributions of such techniques through the development of improved vaccines, drugs, feed grains, and ingredients will also have a significant impact on meat bird growth performance and efficiency. Although molecular genetic developments continue to hold much promise for the improvement of commercial poultry, these techniques are likely to work in concert with traditional methods rather than replace them. REFERENCES Chambers, J. R., 1990. Genetics of growth and meat production in chickens. Pages 599 643 in: Poultry Breeding and Genetics. R. D. Crawford, ed. Elsevier Science Publishing Co., New York, NY.
SYMPOSIUM: GENETIC SELECTION STRATEGIES FOR THE FUTURE 1125 Havenstein, G. B., P. R. Ferket, S. E. Scheideler, and B. T. Larson, 1994a. Growth, livability, and feed conversion of 1957 vs 1991 broilers when fed typical 1957 and 1991 broiler diets. Poultry Sci. 73:1785 1794. Havenstein, G. B., P. R. Ferket, S. E. Scheideler, and D. V. Rives, 1994b. Carcass composition and yield of 1991 vs 1957 broilers when fed typical 1957 and 1991 broiler diets. Poultry Sci. 73:1795 1804. Marks, H. L., 1991. Eight-five generations of selection for high four-week body weight in Japanese quail. Pages 113 132 in: Proceedings of the Fortieth Annual National Poultry Breeders Roundtable, St. Louis, MO. Marks, H. L., 1993. The influence of dietary protein on body weight of Japanese quail lines selected under high- and low-protein diets. Poultry Sci. 72:1012 1017. Marks, H. L., 1995. Heterosis and overdominance following long-term selection for body weight in Japanese quail. Poultry Sci. 74:1730 1744. Siegel, P. B., and E. A. Dunnington, 1987. Selection for growth in chickens. Crit. Rev. Poult. Biol. 1:1 14. Wing, T. L., and A. W. Nordskog, 1982. Use of individual feed records in a selection program for egg production efficiency. I. Heritability of the residual component of feed efficiency. Poultry Sci. 61:226 230.