Research and Reviews: Poultry and Swine

Special Circular 171-00

The Effect of Selection for Increased Body Weight, Egg Production, and Shank Width on Developmental Stability in Turkeys

K. E. Nestor ¹,
J. W. Anderson,
and R. A. Patterson
The Ohio State University
Department of Animal Sciences

¹ For more information, contact at: The Ohio State University,
Ohio Agricultural Research and Development Center, 125 Gerlaugh Hall, 1680 Madison Ave.,
Wooster, OH 44691; 330-263-3757; Fax: 330-263-3949; e-mail: nestor.1@osu.edu.

Abstract

Bilateral asymmetry is a measure of developmental stability. At 20 weeks of age, the right and left shank length, shank width (width laterally at the dew claw), shank depth (width perpendicular to the dew claw), and face length (between auditory canal opening and the posterior junction of the upper and lower mandible) were measured in three randombred control and three selected lines of turkeys. The lines were grown intermingled, with the sexes being grown in different houses. The selected lines had been selected for increased egg production (38 generations), increased 16-week body weight (32 generations), or increased shank width (19 generations), and had a higher level of inbreeding (average = 36.9%) than the randombred controls (average = 11.6%). The bilateral differences (right minus left) were analyzed for the presence of asymmetry. In order to adjust for possible scaling effects, relative asymmetry (RA), in which the mean of the absolute differences between sides was divided by the mean of the two sides and the resulting value multiplied by 100, was used as a measure of bilateral asymmetry. The randombred control and selected lines were contrasted to study the effect of homozygosity on RA. Likewise, the large-bodied lines (F, FL, and RBC3) were contrasted to the small-bodied lines (RBC1, E, and RBC2) to study the effect of body weight on RA.

The level of asymmetry for the traits was ranked: face length > shank width = shank depth > shank length. The individual lines differed in RA for shank length and shank width for both sexes and for shank depth and face length in females. In general, the influence of body weight, as measured in the contrast of large-bodied and small-bodied lines, on RA was larger than that of homozygosity, as measured by the contrast of the selected and randombred control lines.

Introduction

Developmental homeostasis may be reduced by genetic selection within populations (Lerner, 1954). Individual animals have multiple copies of morphological structures, and if development of such structures on the two sides of a bilaterally symmetrical animal are under genetic control, then both sides are expected to be identical because they are products of the same genome (Leary and Allendorf, 1989). There are three possible types of asymmetry - fluctuating (FA), directional (DA), and antisymmetry (AS) (Van Valen, 1962). Fluctuating asymmetry is defined as the differences of the right and left sides having a mean of zero with normal variation. For DA, the differences are not zero but the variation is normal. If the differences between sides have a mean of zero with non-normal distribution (usually bimodal), the differences are AS.

Fluctuating asymmetry is generally thought to be a good measure of environmental and genetic (inbreeding and founder effects) stresses in laboratory and natural populations (Van Valen, 1962). However, DA and AS are not as useful for measurement of developmental stability, since these forms of asymmetry might be controlled genetically (Palmer and Strobeck, 1992).

Generally, FA and homozygosity of genes are positively correlated (Leary and Allendorf, 1989; Parsons, 1990), but exceptions do occur (Thoday, 1958; Palmer and Strobeck, 1986). A positive correlation was observed between the coefficient of variation of a trait and level of FA (Clarke, 1998).

Bilateral asymmetry has been studied in commercial (Møller et al ., 1995) and experimental (Yang et al ., 1997, 1998; Yang and Siegel, 1998) populations of chickens. Møller et al . (1995) reported that FA was greater in fast-growing breeds than in slower-growing ones. When fast-growing chickens were kept at a density of 20, 24, or 28 chickens per square meter, FA was positively associated with population density. Yang et al . (1997) studied the relative asymmetry [RA; ((*left side-right side*)) (left side + right side ) 2)) X 100)] of shank length, shank width (perpendicular to the spur), face length (distance between the auditory canal and posterior junction of the upper and lower mandibles), and length and weight of the first primary flight feather in lines of chickens selected 23 generations for high or low antibody response to sheep red blood cells, sublines of the high and low lines that had been relaxed selected for eight generations, and reciprocal F1 crosses between the selected lines. Rankings among the various genetic stocks for RA, based on the means for the five traits, showed that the selected lines exhibited greater RA than the crosses between them. The RA of the two lines in which selection had been relaxed was similar to the RA of the selected lines. In another study, Yang and Siegel (1998) studied asymmetry in the length and weight of the shank, the length and weight of the first primary flight feather, length of the ceca, and weight of the lungs in the high and low antibody response lines and reciprocal F1 crosses among them. All forms of asymmetry were observed in the various subgroups (20 FA, 12 AS, and 8 DA). The heterosis of RA in the crosses was negative for all traits and significant for shank length, length and weight of the first primary feather, and ceca lengths in some groups.

Yang et al . (1998) studied the developmental stability at 240 days of age in a line of chickens selected 39 generations for low 56-day BW and a subline of the low line that had been maintained with relaxed selection. The selected line was divided into layers and non-layers. Relative asymmetry was similar among the three groups for shank length, shank depth, and face length. Values for first primary feather weight and length were higher for both low-weight selected groups than the relaxed selected subline. The overall mean of asymmetries was higher for both selected groups than the relaxed selected group.

The results of Yang et al . (1997, 1998) and Yang and Siegel (1998) suggest that asymmetry in bilateral traits may be a good measure of genetic stress and developmental stability. Turkeys have been selected long-term for increased body weight, increased egg production, and increased shank width. The purpose of the present study was to study developmental stability in the selected lines and three randombred control lines by measuring asymmetry of bilateral traits.

Materials and Methods

Lines of Turkeys

A line (E) of turkeys was selected long-term (38 generations) for increased egg production for various periods of production (McCartney et al ., 1968; Nestor et al ., 1996). The E line was started from a randombred control population (RBC1; McCartney, 1964) and selection was based on the offspring from the best dams. A line (F) mass selected long-term (32 generations) for increased 16-week body weight (Nestor, 1977; Nestor et al ., 1996) was developed from a different randombred control population (RBC2; Nestor et al ., 1969). A subline (FL; Nestor et al ., 1985) of the F line was initiated by mass selecting only for increased shank width with 19 generations of selection completed at the time of the present study. A third randombred control line (RBC3; Noble et al ., 1995) was developed from reciprocal crosses of the F line and a commercial sire line. The RBC3 line had been maintained for 12 generations.

The randombred control lines were maintained without conscious selection by a paired mating system using 36 parental pairs. The number of parental pairs used to reproduce the E line varied from 36 to 72 over the course of selection. For the F line, 36 parental pairs were used to reproduce the line through Generation 21, after which time 36 males were mated to 72 females with two females being assigned to each male. The FL line was maintained with 36 parental pairs from Generation 1 through 12; thereafter, the line was maintained by mating 36 males to 54 females.

Management and Measurement of Birds

The birds were produced in a single hatch based on a two-week collection of eggs from parents of similar age. The birds were grown, sexes separate, in confinement in separate houses. All birds were provided with a declining protein five-ration system (Naber and Touchburn, 1970) based on the schedule for males. Continuous lighting was provided from hatching to eight weeks of age at which time the length of the light day was reduced to 12 hours. At 16 weeks of age, the amount of light per day was reduced to 10 hours and remained at this level.

Body weight was recorded at 8, 16, and 20 weeks of age. At 20 weeks of age, measurements of shank length, shank width (laterally at the dew claw), shank depth (perpendicular to the dew claw), and face length (between auditory canal opening and the posterior junction of the upper and lower mandible) were made to the nearest tenth of a unit on both sides of the body by the same person. The number of turkeys in each line and sex subgroup was 50 or 51.

Estimation of Inbreeding

The average increase in inbreeding per generation was calculated from one-half the reciprocal of the effective population size. Effective population size was based on the total number of parents and variation in family size among the parents. The amount of inbreeding in the F line at the time of formation of the FL was added to the values for the FL line.

Statistical Analysis

The sexes were analyzed separately. Data on asymmetry were expressed for the right side minus the left side as signed and absolute values. The RA was obtained by dividing the absolute differences between sides by the average value of both sides and multiplying by 100. The signed bilateral asymmetry values were tested for normality with mean zero by the Shapiro-Wilk statistic and one sample t test (SAS Institute, 1988). Traits within line and sex subgroups were classified into type of asymmetry (FA, DA, or AS) based on the mean and normality of variation of the signed differences (Van Valen, 1962).

The influence of line on the various traits was tested by a one-way ANOVA. Means were separated by repeated t tests. In order to study the effect of body weight, contrasts were made of the small-bodied (RBC1, E, and RBC2) and large-bodied (F, FL, and RBC3) lines. Contrasts were also used to compare randombred (RBC1, RBC2, and RBC3) and selected (E, F, and FL) lines.

Results

The total accumulated inbreeding was 16.4, 13.4, and 5.1%, respectively, for the RBC1, RBC2, and RBC3 lines. The respective average increase in inbreeding per generation was 0.43, 0.42, and 0.43%. For the selected lines, the total accumulated inbreeding was 52.4, 28.1, and 30.2%, respectively, in the E, F, and FL lines. The increase in inbreeding per generation was higher in the E line (1.38%) than in the F (0.88%) and FL (0.95%) lines. Thus, the selected lines not only had higher accumulated inbreeding but a faster rate of inbreeding than the randombred control lines.

All individual turkey lines were significantly different in body weight at 8, 16, and 20 weeks of age within sexes (Table 1). Selection for increased egg production decreased body weight, whereas body weight was increased in the F and FL lines relative to the RBC2 line. The body weight of the randombred control lines was ranked in the order: RBC3 > RBC2 > RBC1. In general, the large-bodied lines were more than twice as heavy as the small-bodied lines.

With the exception of the FL line, line differences in average values for bilateral measurements of shank length, shank width, shank depth, and face length generally were similar to those in body weight (Table 2). The FL line had greater shank width (the selected trait) and shank depth than the F line even though body weight of the F line was more than 25% larger than the body weight of the FL line (Table 1). The RBC3 line had greater body weight than the FL line (Table 1), but the shank measurements were greater in the FL line than in the RBC3 line, and face length was similar in the FL and RBC3 lines (Table 2).

The signed and absolute differences and RA between the right and left shank lengths differed significantly among individual turkey lines for both males and females and were highest in the FL line (Table 3). The RA ranged from 0.61 to 1.30% in males and from 0.71 to 1.65% in females for the various lines. Fluctuating asymmetry was observed for shank length in all line and sex subgroups except for FL males and RBC1 females in which DA was found. When randombred and selected lines were compared as a group, only the absolute difference in male shank length was significantly different with the selected lines exhibiting a larger difference. The signed and absolute differences and RA were larger in the large-bodied lines than in the small-bodied lines for both sexes.

The FL line also had the greatest RA for male (range = 1.53 to 3.37%) and female (range = 1.67 to 5.02%) shank width (Table 4), and line differences were evident for signed and absolute differences between sides of the body and RA in both sexes. In the comparison of the randombred and selected lines, only the absolute differences between sides were significantly higher for the selected lines in males while in females the selected lines had larger signed and absolute differences and RA than the randombred lines. The large-bodied lines had greater signed and absolute differences and RA than the small-bodied lines in both males and females. Fluctuating asymmetry was observed in all line and sex subgroups except for the RBC1 males for which DA was observed.

Even though shank depth and shank width were measured at the same location (at the dew claw) but in a different direction (anterior-posterior for depth and laterally for width), the results were quite different between the two measurements. For males, the RBC1 line had a significantly lower signed difference between sides for shank depth than the other lines, but no line difference in absolute difference or RA was observed. Males of the selected lines had a greater signed difference than the randombred control lines. Males of the large-bodied lines had greater signed and absolute differences between sides than the small-bodied lines, but there was no difference in RA. For males, the asymmetry was FA in three lines and DA in the other three lines. For females, line differences were evident in all three measurements. The RA ranged from 2.21 to 3.91%. No significant differences were observed between the randombred and selected lines. The large-bodied lines had greater signed and absolute differences and RA than the small-bodied lines. All lines exhibited FA except for the F line which exhibited DA.

The line comparisons were different for males and females for lateral asymmetry in face length (Table 6). A line difference was observed for the signed differences between sides in males but there was no significant line variation in absolute differences between sides or in RA. Lines differed for all three measures in females. There was no significant difference between the randombred control and selected lines or between the small-bodied and large-bodied lines in any measure for either males or females. Directional asymmetry was noted in E line males and females and in RBC2 females and the remaining line and sex subgroups exhibited FA.

No individual line difference was evident in the total or average RA of shank length, shank width, shank depth, and face length for males (Table 7). Likewise, there was no difference in total or average RA for males between the randombred control and selected lines and small-bodied and large-bodied lines. The total RA for all traits and average RA were 12.4 and 3.1%, respectively. Lines differed in total and average RA for females. The large-bodied lines had greater values than the small-bodied lines. There was no difference between the randombred and selected lines. The total and average RA for females was 12.5 and 3.1%, respectively.

The RA was largest for face length and least for shank length in both sexes (Table 8). Shank width and shank depth had RA values in males and females intermediate between face length and shank length but significantly different from both. Shank width and shank depth did not differ in RA.

Discussion

Line differences in body weight, shank measurements, and face length were similar in all comparisons except those involving the FL line. Body weight and body measurements were positively correlated genetically in several lines (Johnson and Asmundson, 1957; Nestor et al ., 1967; Havenstein et al ., 1988). Selection for increased shank width in the FL line increased body weight and shank length (Nestor et al ., 1985), shank depth (unpublished data), and the relative amount of leg bones (Nestor et al ., 1988). The shank of the FL line is becoming more round as measured by the ratio of shank width to shank depth (unpublished data). Face length of the FL line was greater than expected based on BW in the present experiment so perhaps selection in the FL line is increasing the relative amount of bone throughout the body. Previous results have shown that selection for increased egg production in the E line was associated with decreases in body weight and in shank length (Nestor, 1971). Changes in BW and shank measurements in the F line have been previously reported (Nestor et al ., 1985).

The large line differences in BW and body measurements may have resulted in scaling effects (Palmer and Strobeck, 1986) in the signed and absolute differences between the two sides of the body. Therefore, to remove possible scaling effects, asymmetry was expressed as RA that adjusts for the trait mean (Thoday, 1958).

Selection and the resulting inbreeding increase homozygosity of genes. The comparison of the selected and randombred control lines in the present experiment should provide a reasonable model for studying the influence of homozygosity on bilateral asymmetry in turkeys. Fluctuating asymmetry and homozygosity of genes are positively correlated in natural populations and some laboratory populations (Leary and Allendorf, 1989; Parsons, 1990; Palmer, 1996) but exceptions do occur (Thoday, 1958; Palmer and Strobeck, 1986). In chickens, Yang et al . (1997) and Yang and Siegel (1998) observed that, based on mean RA for five traits, lines selected for high or low antibody response to SRBC had greater RA than crosses among them. In the comparison of the selected and randombred control lines in the present study, RA was not different for shank length, shank depth, and face length in males and females, and shank width in males. For shank width in females, RA was larger in the selected lines than in the randombred control lines. When the total or average RA for all four traits was considered, there was no significant difference between the randombred control and selected lines. Among the individual selected lines, the E line had been selected for the longest period of time and had the highest level of inbreeding but exhibited among the lowest RA values for some traits (shank length, shank width, and female shank depth). Egg production is a component of fitness and the effect of the increased egg production in the E line could have masked the effect of homozygosity on developmental stability. However, the RBC3 line had the least inbreeding among the randombred control lines and had higher RA values for some traits than the other two randombred lines. Thus, it appears that homozygosity of genes has little influence on bilateral asymmetry in turkeys.

The comparison of the selected and randombred control lines was complicated by a difference in body weight between the two groups of lines. The selected lines as a group had heavier body weight than the average of the three randombred controls. Therefore, the large-bodied (F, FL, and RBC3) and small-bodied lines (E, RBC1, and RBC2) were contrasted. The RA of shank length and shank depth in both sexes and shank width in females was larger in the large-bodied lines than in the small-bodied lines. There was no difference between the two groups of lines for RA of shank width of males and face length of both sexes. When the total or average RA for the four traits was analyzed, RA for males did not differ between the large-bodied and small-bodied lines but in females, RA was greater for the largebodied lines. The males and females were reared in different houses and all lines were reared intermingled. The RA was higher for small-bodied males than for small-bodied females, while the reverse was true for the large-bodied lines. It is possible that the social competition was a disadvantage for small-bodied males and large-bodied females when the lines were reared intermingled. Overall, the results of the present study using turkeys indicated that body weight had a greater influence on developmental stability than homozygosity of genes but that not all bilateral traits are similarity affected. In chickens, Møller et al . (1995) reported that FA of several traits was greater in fast-growing breeds than in slower-growing ones.

Both FA and DA was observed in the present study. No AS was observed. In the individual line and sex subgroups, the bilateral differences exhibited FA in 38 of 48 comparisons. In some cases, FA was observed in one sex while DA was found in the other sex of the same line. In chickens, Yang et al . (1997) and Yang and Siegel (1998) observed that the type of asymmetry could vary within lines and AS was observed in their studies.

The traits were ranked face length > shank width = shank depth > shank length for RA in the present study. A similar ranking of shank length, shank depth, and face length was observed in chickens by Yang et al . (1998).

In summary, the RA of shank length, shank width, shank depth, and face length was measured in three randombred control and three long-term selected lines. The selected lines were selected for increased egg production, increased 16-week body weight, or increased shank width and had a higher level of inbreeding than the randombred control lines. Homozygosity as measured by a comparison of the randombred control and selected line had little influence on RA. A comparison of large-bodied and small-bodied lines indicated that body weight had a greater influence than homozygosity on RA. Fluctuating asymmetry and DA, but not AS, were observed.

References

Clarke, G. M. 1998. The genetic basis of developmental stability. V. Inter-and intra-individual character variation. Heredity 80:562-567.

Havenstein, G. B., K. E. Nestor, V. D. Toelle, and W. L. Bacon. 1988. Estimates of genetic parameters in turkeys. 1. Body weight and skeletal characteristics. Poultry Sci. 67:1378-1387.

Johnson, A. S. and V. S. Asmundson. 1957. Genetic and environmental factors affecting size of body and body parts of turkeys. 1. The heritability and interrelationships of body weight and live body measurements. Poultry Sci. 36:296-301.

Leary, R. F. and F. W. Allendorf. 1989. Fluctuating asymmetry as an indicator of stress: Implications for conservation biology. Trends Ecol. Evol. 4:214217.

Lerner, I. M. 1954. Genetic Homeostasis. John Wiley and Sons, New York, N.Y. McCartney, M. G. 1964. A randombred control population of turkeys. Poultry Sci. 43:763-765.

McCartney, M. G., K. E. Nestor, and W. H. Harvey. 1968. Genetics of growth and reproduction in the turkey. 2. Selection for increased body weight and egg production. Poultry Sci. 47:981-990.

Møller, A. P., G. S. Sanotra, and K. S. Vestergaard. 1995. Developmental stability in relation to population density and breed of chickens Gallus gallus. Poultry Sci. 74:1761-1771.

Naber, E. C. and S. P. Touchburn. 1970. Ohio Poultry Rations. Ohio Cooperative Extension Service Bulletin 343. The Ohio State University, Columbus, Ohio.

Nestor, K. E. 1971. Genetics of growth and reproduction in the turkey. 3. Further selection for increased egg production. Poultry Sci. 50:1672-1682.

Nestor, K. E. 1977. Genetics of growth and reproduction in the turkey. 5. Selection for increased body weight alone and in combination with increased egg production. Poultry Sci. 56:337-347.

Nestor, K. E., W. L. Bacon, G. B. Havenstein, Y. M. Saif, and P. A. Renner. 1988. Carcass traits of turkeys from lines selected for increased growth rate or increased shank width. Poultry Sci. 67:1660-1667.

Nestor, K. E., W. L. Bacon, Y. M. Saif, and P. A. Renner, 1985. The influence of genetic increases in shank width on body weight, walking ability, and reproduction of turkeys. Poultry Sci. 64:2248-2255.

Nestor, K. E., M. G. McCartney, and N. Bachev. 1969. Relative contributions of genetics and environment to turkey improvement. Poultry Sci. 48:1944-1949.

Nestor, K. E., M. G. McCartney, and W. R. Harvey. 1967. Genetics of growth and reproduction in the turkey. 1. Genetic and non-genetic variation in body weight and body measurements. Poultry Sci. 46:1374-1384.

Nestor, K. E., D. O. Noble, J. Zhu, and Y. Moritsu. 1996. Direct and correlated responses to long-term selection for increased body weight and egg production in turkeys. Poultry Sci. 75:1180-1191.

Noble, D. O., D. A. Emmerson, and K. E. Nestor. 1995. The stability of three randombred control lines of turkeys. Poultry Sci. 74:1074-1078.

Palmer, A. R. 1996. Waltzing with asymmetry. BioScience 46:518-532.

Palmer, A. R. and C. Strobeck. 1986. Fluctuating asymmetry: Measurement, analysis, patterns. Annu. Rev. Ecol. Syst. 17:391-421.

Palmer, A. R. and C. Strobeck, 1992. Fluctuating asymmetry as a measure of developmental stability: Implications of non-normal distributions and power of statistical tests. Acta Zool. Fenn. 191:57-72.

Parsons, P. A. 1990. Fluctuating asymmetry: An epigenetic measure of stress. Biol. Rev. 65:131-145.

SAS Institute. 1988. SAS/STAT User's Guide. 6.03 Edition. SAS Institute, Inc., Cary, N.C.

Thoday, J. M. 1958. Homeostasis in a selection experiment. Heredity 12:401-415.

Van Valen, L. 1962. A study of fluctuating asymmetry. Evolution 16:125-142.

Yang, A., E. A. Dunnington, and P. B. Siegel. 1997. Developmental stability in stocks of White Leghorn chickens. Poultry Sci. 76:1632-1636.

Yang, A., E. A. Dunnington, and P. B. Siegel. 1998. Developmental stability in different genetic stocks of white rock chickens. J. Hered. 89:260-264.