HATCHABILITY of eggs as a trait of

Hatching Time and Hatchability in the Fowl 1 B. B. BOHEEN, L. B. CRITTENDEN 2 AND R. T. KING 3 Purdue University Agricultural Experiment Station, Lafayette, Indiana (Received for publication July 6, 1960) HATCHABILITY of eggs as a trait of the fowl has been intensely studied. The scientific interest in this trait is a result of its great economic importance and its extreme biological complexity, which challenges the curiosity of investigators. Many factors, both genetic and environmental have been found to affect hatchability, although in many cases the results are not consistent, and in others the nature of the effect on hatchability is not clearly evident. The resulting voluminous literature from this work has been reviewed and interpreted by Landauer (1948), Taylor (1949) and Bernier, Taylor and Gunns (1951). A number of factors, both genetic and environmental, which have been shown to affect hatchability, have also been shown to affect the required hatching time of the embryo. Apparently, however, no workers have attempted to consider the relationship between hatching time itself and hatchability. The purpose of this paper is to present some preliminary data concerning the relationship of hatching time to hatchability and to consider the possible effect of hatching time on the relationship of other factors such as egg weight and preincubation storage to hatchability. REVIEW OF LITERATURE It has long been known that hatchability is reduced after extended storage of hatch- 1 Journal Paper No. 1635 of the Purdue University Agricultural Experiment Station. 2 Present address, National Institutes of Health, Bethesda, Md. 3 Present address, Creighton Brothers Poultry Farm, Warsaw, Ind. 620 ing eggs. Waite (1919), Scott (1933), Funk (1934) and others, have demonstrated that chicken eggs will decline in hatchability after various periods of storage. That the same holds true for turkey eggs has been shown by Asmundson (1947), Abplanalp and Kosin (1953) and Kosin (1954). The conclusions vary as to the age of the eggs when a decline occurs. Thus, Scott (1933) found that turkey eggs could be held as long as 28 days with little decline in hatchability, while the others found much shorter storage periods deleterious. Abplanalp and Kosin (1953) observed a higher estimate of heritability for hatchability when based on turkey eggs stored eight to 14 days than when based on eggs stored only one to seven days prior to setting. Kosin (1954) further observed differences between strains of turkeys in resistance to storage effects with the high hatchability strains, in general, being more resistant to the harmful effect of storage than those with lower hatchability. This result might in part explain the findings of Scott (1933) who reported a hatch of over 85 percent for the strain of turkeys studied. That chicken eggs held prior to setting require a longer period of incubation than fresh or non-stored eggs has been shown by Funk (1934) and Byerly (1933). The results of Kaufman (1938) suggest that the increased hatching time of stored eggs is not a result of a differential growth rate for embryos of different storage periods but is caused by a delay in initiation of growth after being placed in the incubator. On the contrary, Olsen (1942) found no difference in hatching time between turkey eggs stored for one to 16 days prior to in-

HATCHING TIME AND HATCHABILITY 621 cubation. It would seem unlikely that such a difference between chickens and turkeys really exists, especially in view of the similar results for the two species with regard to the effect of holding time on hatchability. Fortunately Kosin (1950) has further checked this relationship in turkeys and obtained results compatible with the data reported for chickens. No explanation for the deviation of Olsen's observations is evident. Another factor which has been reported to affect both hatchability and hatching time is egg size. While the consensus among investigators is that large eggs tend to hatch poorer than medium size eggs, considerable disagreement is found with regard to the relative hatchability of small eggs as compared to the medium size eggs (Landauer, 1948). The recent work of Coles (1956) suggests an inverse relationship between egg weight and hatchability, both within and between hens. That a similar relationship may hold for turkeys as well as chickens was shown by Byerly and Marsden (1938). It has been shown by Byerly (1933), McNally and Byerly (1936) and Williams, Godfrey and Thompson (1951) that in chickens, large eggs require a longer incubation time than smaller ones. A similar relationship was found in turkeys by Olsen (1942). Such a relationship between egg volume and hatching time was observed between avian species from humming birds to ostriches as reported by Worth (1940). Evolutionary significance is attached to the egg size-incubation time relationship by Worth (1940). He observed that those species of birds having shorter incubation periods than predicted on the basis of their egg volumes were "chiefly species subject to predation or some other type of environmental onslaught." On the other hand, those species with longer incubation periods than perdicted were largely birds of prey with adequate means of offense and defense, to which a shortened incubation period would offer no survival advantage. It is also pointed out that eggs of domesticated hens would be expected to require a longer incubation period than the jungle fowl in view of the differences in egg size. While no conscious selection for a shortened incubation period may have occurred in fowls, it is possible that unconscious or "natural" selection for this trait has occurred. This is of interest in connection with the reports by Lerner (1951) and Lerner and Gunns (1952), who showed that in flocks selected for large egg size, the egg size giving optimum hatchability is usually below the mean egg size of the population. He interprets this as genetic homeostasis resulting from artificial selection for egg size being opposed by "natural" selection. It is possible in this case that the opposing "natural" selection is for hatching time. The fact that hatching time is heritable, and therefore subject to either artificial or natural selection, is suggested by the report of Smyth and Howes (1949). These workers selected two lines which after only four generations were distinctly different in hatching times. Strain differences in hatching time have been observed by Byerly (1933) and Henderson (1950). Correlations between relatives as reported by Hays (1941) also support the heritable nature of hatching time. The effect of oxygen level on hatchability has been extensively studied (see Riddle, 1924; Cruz and Romanoff, 1944). This factor is of particular importance in high altitude incubation problems which have received attention from a number of workers including Ells and Morris (1947), Stephenson (1950) and Thompson (1952). It is clear that a deficiency of oxygen reduced the hatchability of both chicken and turkey eggs. Wilgus and Sadler (1954) demonstrated, in addition, that both the oxygen

622 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING level and the C0 2 concentration had an effect on hatching time. Buss (1956) found that the pattern of hemoglobin development was different at high and low oxygen levels. Apparently hemoglobin development was retarded at the lower oxygen level. It was also found that hatching time was increased at the low oxygen level. Furthermore, adjustment of hen mean hatchability for hatching time by multiple correlation analysis eliminated the relationship between hemoglobin level and hatchability. It is probable that the hemoglobin analysis was reflecting differences in developmental stage and the resulting differences in hatching time. The data of Wilgus and Sadler (1954) show a consistent decrease in "dead" embryos at 21 days of incubation resulting from oxygen supplementation. The differences amounted to seven to 12 percent. Thompson (1952) reports finding as many as 80 percent late dead embryos. Such late dead embryos were examined by Davis (1955) to determine if they were truly dead or merely unhatched, but the result of such examination was not reported. Thompson (1952) also states that eggs incubated under hens hatch equally well from sea level to 10,000 feet in altitude. The time required for hatching under hens at the various altitudes, however, is not reported. Strain differences in resistance to low oxygen are suggested by Stephenson (1950). Hereditary differences are also suggested by Wilgus and Sadler (1954), who further observed that the greatest hatchability response to added oxygen was obtained from individuals with the lowest hatchability without oxygen. Davis (1955) also observed individual hen differences and found that the heritability of hatchability was higher when estimated from air-hatched eggs than when oxygen supplementation was used. The interaction between hens and oxygen levels as reported by Wilgus and Sadler (1954) was not observed by Davis (1955). Other factors are also known to affect both hatchability and hatching time. For example, the effects of incubation temperature and humidity are well known (see Landauer, 1948). It was shown by Mc Cartney and Shaffner (1949) that eggs from hens fed either thiouracil or thyroprotein required a longer period to hatch than untreated controls. No data on the hatchabilities of the three lots were presented. In the case of eggs from hens on iodine deficient rations, Rogler, Parker, Andrews and Carrick (1959) obtained results showing a greatly increased hatching time for the iodine deficient eggs. It is evident that in many studies in which hatching time has been critically observed, the environmental factors known to affect hatchability, have also had an effect on hatching time. It is unfortunate that more studies have not included observations on hatching time. Neither have most workers indicated the length of the incubation period allowed when reporting data on hatchability. MATERIAL AND METHODS Hatching eggs were collected for a period of 12 days (June 22-July 3, 1955) from 100 mass-mated yearling White Leghorn hens. The hens were all hatched during March 1954 and were completing their first laying year after having been used in the pedigree breeding pens. Each evening all eggs laid that day were weighed to the nearest half gram on a shadowgraph balance. The egg weight, date of lay and hen number were recorded on the small end of each egg. The eggs were stored in an egg room held at temperatures from 45 to 55 F. and were turned daily until six hours before setting. A total of 741 eggs were obtained. For six hours prior to setting, the eggs were held at room temperature, during

HATCHING TIME AND HATCHABILITY 623 which time they were trayed at random in six incubator trays. At 7:00 p.m. on July 4, all eggs were placed in a Buckeye Clipper incubator automatically maintained at a temperature of 99.5 F. and a wet bulb thermometer reading of 85 F. After 18 days of incubation, the eggs were candled and all dead embryos and infertile eggs removed. Infertility was confirmed by macroscopic examination of broken-out eggs. The remaining eggs were placed in a Buckeye separate hatcher with automatic temperature and mist spray type humidity controls. The temperature was maintained at 97 F. and the wet bulb thermometer reading was controlled to 87 F. The hatcher was opened every two hours to remove any chicks hatched during the preceding two hours, along with the shells from which they emerged. The eggs were identified by the data on the shells and the hatching period recorded. A chick was considered "hatched" only if it was free of its shell and no consideration was given to whether the down was fluffed or still wet. The first chick hatched during the first two hours of the 21st day of incubation, so this was assigned period number one. Removal of chicks continued through 27 two hour periods so that the total time of hatching covered 54 hours. Temperature and humidity returned to the operating levels in no more than 15 minutes following any opening of the machine for removal of chicks. Fertility was unfortunately low. Of the 741 eggs set, 474 were fertile, or 64 percent. A large portion of infertility was due to the complete infertility of the eggs from a number of individual hens. A total of 366 chicks were hatched at the end of the test giving a 77.6 percent hatch of fertile eggs set. The hatch from these hens during the normal hatching season was 75.3 percent of fertile eggs set. For analysis, all percentages were transformed to angles by means of the arc sine transformation as shown in table 16.9 from Snedecor (1940). The variables to be analyzed then are egg weight (Xj) in grams, hatching time (X 2 ) in periods of hatch, holding time (X 3 ) in days, and hatchability (Y) as transformed percentages. In order to observe the effect of the allowed incubation time on the relationships between the variables, three arbitrary incubation times are definited. Thus hatching was considered complete at the end of hatch period 12, or at the end of 21 days of incubation (Yi). Similarly hatching was considered as complete at the end of hatch period 18, or 21J/2 days of incubation (Y 2 ) and at the end of hatch period 27, or 22*4 days of incubation (Y 3 ). ANALYSIS AND RESULTS The percent of the chicks hatched which emerged during each two hour period and the cumulative percent of chicks hatched to the end of each period are shown in Figure 1. The distribution of hatching times approaches normality with a mean of 13.4 periods or 21.125 days and a standard deviation of 4.48 periods or 8.96 hours. It was feared that the lowered humidity due to the frequent opening of the incubator would retard or prevent the hatching of the normally late emerging chicks, resulting in a skewed distribution. Severe skewness is not evident, indicating that the effect of the frequent opening of the machine was negligible. It may further be seen from Figure 1, that at the end of the 21st day of incubation (period 12) only 43.2 percent of the potential chicks had emerged. At the end of ZV/i days (period 18) 86.7 percent of the potential chicks had hatched, while 97.8 percent were out after 22 full days of incubation (period 27). It may be of some significance that this strain had been allowed a 22 day incubation period for a number of years. The effect of the holding time of the eggs

624 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING I 2 3 4 56 78 9 10!' 12 13 14 15 1617 18 19202122 2324252627 HATCHING PERIOD FIG. 1. Percent of potential chicks hatched in each two hour hatch period, and cumulatively to 22^4 days of incubation. prior to incubation on their hatchability at the three arbitrary incubation times is shown in Figure 2. The analysis of variance of the transformed percentages is shown in Table 1. The variance among the holding period means due to chance is estimated as 821 divided by the harmonic mean number of eggs set in each period, as shown by Bartlett (1947). A significant difference in hatchability was found between days of holding after 21 and 21% days of incubation but not after 22% days. However, the mean square due to linear regression in each case is highly significant, suggesting that an effect of holding time on hatchability still exists after 22% days of incubation. Failure to obtain significance between holding periods in this analysis could result from the small value of the mean square due to deviations from regression. The magnitude of the coefficients of regression of hatch time on holding time declined sharply as the length of the incubation period increased (see Table 1) but are significant at all three lengths of incubation. It is possible that some of the normally late hatching embryos are unable to hatch in any length of time because of altered incubation environment due to frequent opening of the incubator. This possibility seems remote in view of the near normal distribution of hatch times and the normal hatchability percentage at the end of incubation, so it is probable that some other factor affecting hatchability, such as early embryonic death, is also being af- 80 70 60 50 30 DAYS OF INCUBATION 21 1/2 OAYS OF INC \UBATION \1 21 OAYS OF INCUBATION I Z 3 4 5 6 7 8 9 10 II 12 DAYS OF HOLDING PRIOR TO INCUBATION FIG. 2. The effect of length of pre-incubation storage on hatchability at three arbitrary lengths of incubation.

HATCHING TIME AND HATCH ABILITY 625 TABLE 1. Analysis of variance of hatchability {arc sine \fp) defined at three periods of incubation, as affected by period of holding (all fertile eggs included) Source of variation Between holding periods Due to linear reg. Due to dev. from reg. Variance of means due to chance (821/fih) 11 * Significant (P<.05). ** Highly significant (P<.01). 00 1 10 fected by holding time, as observed by Scott (1933) and Kaufman (1938). A highly significant correlation of.736 between days of holding and percent dead and infertile in the present data suggests that the true correlation is greater than zero. The deviations from regression of hatchability on days of holding were a significant source of variation for the 21 day incubation length but not for the two longer incubation times. Since the mean squares for deviations were small and not significant in two of the three analyses, the one value reaching significance is considered to be a random event and it is concluded that the primary effect of storage of eggs up to 12 days prior to incubation on hatchability is linear and begins with the first day of storage. The large change in the regression coefficients as the allowed incubation period is changed suggests that any shortening of the incubation time, or conversely, any increase in the required incubation time of the egg, either by genetic or environmental influences, with a fixed period of incubation will increase the observed effect of holding time. The correlations between the holding time means for egg weight, hatch time and hatchability at the three arbitrary incubation periods are shown in Table 2. Egg 21 days 233.8** 2,087.55** 48.43* 21.04 b=-3.82** Duration of incubation 21 days 52.44** 442.68** 13.42 21.04 b=-1.76** 22J days 28.17 122.86** 18.70 21.04 b.927** weight is positively correlated with hatching time and negatively correlated with hatchability at all three periods of incubation, although not significantly at the S percent level in any case. Lack of significance here could be due to the small number of observations and/or the small variation of the egg weight means by holding periods (see Table 3). Hatching time is negatively and significantly correlated with hatchability at hatch periods 12 and 18, and would account for even more of the variance in hatchability at these hatching times than would holding time itself (see Table 1). This suggests that at these incubation times, the effect of holding time on hatchability is primarily due to the effect of holding time on hatch time. At hatch period 27, or after all chicks had hatched, the correlation between hatch time and hatchability had declined and was nonsignificant. It was also lower than the cor- TABLE 2. Correlations between the means by holding periods of egg weight, hatching time and the three measures of hatchability Egg wt. (Xi) Hatching time Hat. at per. no. 12 Hat. atper.no. 18 Hatching time (Xs).3157 ** Highly significant (P <.01). Hatchability at period no. 12 -.238 -.975** 18 -.301 -.916**.966** 27 -.131 -.434.570.713**

626 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING relation between holding periods and hatchability, again suggesting that at hatch period 27, holding time was affecting hatchability in some manner other than its effect on hatch time. The relationships of holding time to egg weight and hatching period, on an individual egg basis, are shown in Table 3. As would be expected, no significant differences between holding times in egg weight were found. The effect of holding time on hatch period, however, is highly significant, concurring with previous investigations (see Kaufman, 1938). The linear regression of hatch period on holding time accounts for a significant and large portion of the sum of squares of hatch time. Nevertheless the deviations from linear regression account for a significant portion of the variation of the holding period means. The total relationship between days of holding and hatch time in these data is thus shown to be nonlinear. It should be pointed out, however, that in this case significance of the deviations from regression does not necessarily indicate a non-linear relationship between the two variables of primary interest, or time of holding and hatch period. Some purely random environmental influence, such as daily fluctuation in temperature, or the time of being placed in the cooler, could bring about a deviation of the day mean from the regression line without affecting the magnitude of the variation between TABLE 3. Analysis of variance of egg weight (Xi) and hatching time (X 2 ) as affected by time of holding (X 3 ) {All eggs which hatched) Source of variation Between holding periods Due to linear reg. Dev. from linear reg. Within holding periods Total 11 354 365 ** Highly significant (P-C01). 1 10 Egg weight (Xi) Mean square 15.91 14.43 Hatching time (Xj) Mean square 190.09" 1,613.01** 47.80** 16.34 TABLE 4. Analysis of Hie regression of hatch-lime on holding period Source of variance Linear regression Quadratic Cubic Quartic Quintic Residual Total between holding periods Within holding periods 1 1 1 1 1 6 11 354 * Significant (P<.05). ** Highly significant (P<.01). Mean square 51.85** 1.39 7.39* 1.01.08 1.34* 6.34.54 eggs within the days. Keeping this fact in mind, the regression was analyzed by use of orthogonal polynomials with the result shown in Table 4. The quadratic, quartic and quintic terms are non-significant while the cubic and residual terms are significant. This suggests an erratic relationship probably brought about by some random environmental effect on the day means which does not appear within days of holding. It is concluded then that the relationship between hatch period and holding time is primarily linear in the range observed. The correlation between egg weight and hatching time was estimated as.109 ( = 363) which is barely significant (P <.05). The correlation between these two variables within holding times was.094 ( = 352) which is not significant but closely approaches the five percent level. The correlation between the holding period means for egg weight and hatching time was.325 ( = 9) which does not approach significance. Thus there is evidence for a correlation between egg weight and hatching time but it does not appear to be associated with the holding time means for these variables. The variables egg weight, hatching time and holding time, were next analyzed on an individual egg basis, with the hen producing each egg considered as the variable of

HATCHING TIME AND HATCHABILITY 627 classification. Observations on three eggs per dam were considered an adequate estimate of the dam's phenotype for these variables, and 54 dams having at least three observations on each variable were available. The sums of squares and cross products for the three variables are presented in Table 5. The ratios of the mean squares show significant differences between hens for egg weight (P <.01) and hatch time (P <.05). As expected no significant difference was observed between hens for holding time. The total correlation between egg weight and hatching period in this analysis was again significant (r =.122; P <.05). The correlation between these variables within dams was small and not significant (r =.004). Between dam means, however, the value obtained was significant (r =.305; P <.05). Thus it appears that the relationship between egg weight and hatching time can be accounted for through an association between dam means for these two variables and it is not a function of individual eggs within hens. The correlation between the holding time and hatching period of individual eggs was highly significant (r =.476; P <.01). Within hens this correlation was also significant (r =.541; P <.01) but between hens means the correlation was not significant (r =.175) due to the lack of variation among hen means in holding time. The coefficient for the regression of hatching time on holding time (days) was.64 hatching periods or 1.28 hours. Adjustment of the sums of squares in the analysis of variance of hatch period for egg weight reduced the F value only slightly and the mean square for dams was still significant (see Table 6). Adjustment of the sums of squares for holding time, however, resulted in reduction of the within hen mean square for hatch period so that the between hen mean square became highly significant (P <.01). Thus it is reasonable that differences between hen means in hatching time exist and that in raw data such differences may be partially masked by the effect of holding time on hatching time variation among eggs within hens. To study relationships involving hen mean hatchability, only hens having eight or more fertile eggs were considered. Only 31 hens having the required number of fertile eggs were available. The results of the analyses of variance of hatchability at the three incubation times are shown in Table 7. The sum of squares for hens is reduced as the length of incubation is increased. Adjusting the sum of squares between hens for the hatching period means of the hens resulted in reduced sums of squares at each hatching period. The reduction at the three hatch periods are, however, quite different. The reduction at hatch period 12 (21 days) was very large and reduced the mean square to non-significance. At hatch period 27 (22% days) only TABLE 5. Sums of squares and cross products for egg weight, hatching lime and ' time within and between hens Source of variation Egg weight Sums of squares Hatching time (X 2 ) Holding time (X 3 ) Sums of products X1X2 X2X3 Between hens Between eggs within hens 53 278 2,982.4 1,537.0 1,714.0 5,583.0 407.0 3,604.0 689.1 146.0 10.9 2,428.0 Total 331 4,519.4 7,297.0 4,011.0 700.0 3,574.0

628 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING TABLE 6.- Analysis of variance of hatch lime adjusted for holding time or egg weight Source of variation Unadjusted Adjusted For holding time For egg weight Between hens Eggs within hens 53 278 32.3* 20.1 53 277 32.04** 14.25 30.29* 20.15 Total 331 330 * Significant (P<.05). ** Highly significant (P<.01). a small but significant reduction occurred. The effect of adjusting the sum of squares for hen mean egg weights also declined as the incubation time was extended, but the difference between hatch periods was not so great as the changes resulting from the adjustment for hatch time. These results suggest that as traits of the hen, hatchability defined at 21 days of incubation (period 12) is a different trait than hatchability defined at 22^4 days (period 27). At period 12, hatchability appears to be reflecting primarily differences in hatching time of the hen, while at period 27, hatchability is only slightly affected by hatching time and is proportionately more affected by egg weight. For further study, the 31 dams were classified into three arbitrary groups of approximately equal size according to their hatchability, or high, medium and low hatchability groups. Three periods of preincubation holding were also defined. These were one to four days, five to eight days and nine to 12 days of holding inclusive. The 278 fertile eggs produced by the 31 hens were then classified as to the hatchability group of their dam and their length of holding time before incubation. This process was repeated for each of the three arbitrary incubation times. The analyses at each incubation time are shown in Table 8. For analysis, the variables hatchability group and holding period were considered as being fixed. The mean squares between hatchability groups were highly significant at each incubation time but decreased as the allowed incubation time was lengthened just as did the analysis of individual hen differ- TABLE 7. Analysis of hen mean hatchability and the result of adjusting for hen mean egg weight or hatching time Source of variation Between hens Between hens after adjusting for hatching time Between hens after adjusting for egg weight Theoretical chance variance of the hen means (821/fih) 30 29 29 CO 12 212.03** 59.62 Analysis at end of hatch period no. 154.75** 91.94 18 198.84** 137.20* 152.05* 91.94 27 151.66* 136.78* 116.62 91.94 Significant (P<.05). ** Highly significant (P<.01).

HATCHING TIME AND HATCHABILITY 629 TABLE 8. Analyses of variance of hatchability as affected by the hatchability grouping of the dams and the length of pre-incubation holding, at three lengths of incubation Source of variation Between hatchability groups Between holding periods Hatch group X holding period 2 2 4 Mean squares from analysis after hatch period no. 12 716.98** 884.84** 20.33 18 484.94** 131.12** 44.49 27 308.40** 6.54 26.14 Theoretical error (821/flh) 00 26.66 26.66 26.66 ** Highly significant (P<.01). ences (Table 7). Similarly the mean squares due to holding periods decreased as the incubation time was increased, and was non-significant at period 27 or 22*4 days. This result agrees with that obtained previously when holding periods were defined as single days (Table 1). The interaction of hatch groups by holding periods was not significant for any length of incubation, although approaching significance at hatch period 18. The few degrees of freedom involved in this analysis invite particular caution in interpreting non-significant results. A better picture of the relations between these variables for the three -HIBH HATCHABILITY 8R0UP -MEDIUM HATCHABILITY GROUP -LOW HATCHABILITY GROUP FIG. 3. The relationship between the hatchgroup of the dams and pre-incubation ability storage time on hatchability after 21 days of incu- bntion. incubation lengths may be observed in Figures 3, 4 and 5. From Figure 3, at 21 days of incubation, it is seen that the three hatchability groups are clearly separated and are all equally and apparently linearly affected by holding time. Figure 4 shows the result at 21% days of incubation and it is seen that the high hatching group is unaffected by holding time. The medium and low hatching groups show about the same effect of holding time as was evident at 21 days of incubation. After 22% days of incubation, as shown in Figure 5, only the low hatchability group of dams appears to be affected by the length of the holding period. These results tend -HIGH HATCHABILITY GROUP -MEDIUM HATCHABILTTY GROUP -LOW HATCHABILITY GROUP 5-8 HOLDING TIME IN DAYS HOLDING 5-8 9-12 TIME IN DAYS FIG. 4. The relationship between the hatchability group of the dams and pre-incubation storage time on hatchability after 21^4 days of incubation.

630 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING -HIGH HATCHABILITY GROUP -MEDIUM HATCHABILITY GROUP LOW HATCHABILITY GROUP HOLDING TIME IN DAYS FIG. 5. The relationship between the hatchability group of the dams and pre-incubation storage time on hatchability after 22J4 days of incubation. to confirm the observation by Kosin (1954) that high hatchability strains of turkeys are more resistant to storage effects than low hatching strains. This observation does not appear to hold, however, if incubation time is restricted, and would not be expected to hold if storage time was extended. The average hatchability and average hatching time of the hens in the three hatchability groups are shown in Table 9 and the analysis of variance of hatching time in Table 10. It is clear that at 21 days of incubation a distinct difference in hatching time exists between all hatchability groups. The difference between the high and medium hatchability groups is reduced after 21% days of incubation, but the difference between the medium and low groups remains about the same. All differences are reduced after 22% days of incubation but an effect still appears to be present, especially between the medium and low hatchability groups. Thus, again a close association is noted between the effects of storage and the hatching time of the hen on hatchability. The correlation between hatching time and percent hatch of fertile eggs, based on the 31 individual hens, was -.847 (P <.01) after 21 days of incubation. This value was reduced to -.557 (P <.01) after 21% days and to -.313 (N.S.) after 22% days of incubation. The effects of the average egg weight of the hen on the hatchability of the hen appeared to be relatively constant for the three lengths of incubation. The correlations between these variables were.485 (P <.01), -.486 (P <.01) and -.481 (P <.01) for the 21, 21% and 22% day incubation lengths respectively. The egg weights for the three hatchability groups at the three incubation times are shown in Table 9. The low hatching group has the largest egg weights at all three incubation times. The high and medium hatching groups were similar in egg weight. For eggs hatched up to and including hatch periods TABLE 9. Average hatchability, hatching times and egg weights for hens classified arbitrarily into three hatchability groups, at three incubation periods Hatchability defined after hatch period no. Hatchability group Hatchability percent 12 Ave. hatch time in periods Ave. egg wt. in grams Hatchability percent 18 Ave. hatch time in periods Ave. egg wt. in grams Hatchability percent 27 Ave. hatch time in periods Ave. egg wt. in grams High Medium Low 51.36 40.04 20.44 11.02 12.69 15.28 60.8 59.6 63.8 74.15 63.73 48.85 11.72 12.41 14.98 60.5 59.8 63.9 76.77 69. SO 56.82 12.46 12.85 13.76 58.9 60.7 63.6

HATCHING TIME AND HATCHABILITY 631 TABLE 10. Analysis of variance of hen mean hatch times as affected by arbitrary classification into three hatchability groups, at three periods of incubation Source of variation Between hatchability groups Within hatchability groups Total Highly significant (P<.01). 2 28 30 Mean squares for analysis after hatch period no. 12 50.62** 1.09 18 29.70** 2.56 27 4.42 4.39 12 or 18 the high hatching group of dams had the larger egg weight, while for eggs hatched through hatch period 27, the high hatching group of dams had the lower egg weight. For eggs hatched through hatch period 27 the relationship between egg weight and hatchability appears to be essentially linear. To test for the linearity of the regression of hatchability on egg weight, the 31 hens were ranked according to egg weight and divided into egg weight groups of three hens. Analysis of variance of hatchability showed significant differences between egg weight groups and significant mean squares due to regression at all three lengths of incubation. The deviations from regression, however, were significant only after 21^4 days of incubation (period 18). While there is evidence here of curvilinearity, it is clear that predicted adjustments of hatchability for egg weight would not be greatly different whether made on the basis of the linear or the curvilinear model. DISCUSSION It is clear from this analysis that hatchability defined at 21 days of incubation is a markedly different trait than hatchability defined at 22% days of incubation. It follows then that any effect on hatchability of a treatment, such as storage, which also affects hatching time will have different effects on hatchability depending upon the length of the incubation time after which hatchability is determined. Thus it would appear that in any experiments involving hatchability, the procedure for determining hatchability should be explicitly defined with respect to time. "True" hatchability would be best defined after sufficient time for all, or almost all, eggs in the upper tail of the distribution to hatch. The close association between the effects of pre-incubation storage and hatching time noted here is of interest in connection with the report of Abplanalp and Kosin (1953). They obtained increased heritability estimates for hatchability in stored eggs. The data presented here suggests that hens with the fastest hatching times are more resistant to the effects of storage than those with slower hatching times, even over the extended incubation intervals. Furthermore, if hatchability were determined at a fixed incubation interval, it could be that the stored eggs had lowered hatchability and greater variation between hens simply because the eggs of some hens did not have sufficient time to hatch. Hatchability of the longer stored eggs would then be reflecting more differences between hens in hatching time than would hatchability of those eggs stored for shorter periods. The report of Smyth and Howes (1949) suggests that the heritability of hatching time is quite high compared to the heritability of hatchability as such. This effect could then explain the increased heritability of hatchability of stored eggs as observed by Abplanalp and Kosin (1953). A similar ar-

632 B. B. BOHREN, L. B. CRITTENDEN AND R. T. KING gument could be applied to the results of Davis (1955) who observed a higher heritability of hatchability in eggs hatched at high altitudes without supplemental oxygen, than when the eggs were hatched with supplemental oxygen. The exact time of embryo deaths was not determined in this study. It is clear, however, that the slow hatching hens are more susceptible to the effects of pre-incubation storage and have a larger number of late dead embryos, especially if the incubation time is restricted. Kaufman (1938) found that storage delayed the initiation of growth after resumption of incubation, but that after growth started, the rate of growth was unaffected by storage. A somewhat similar observation was made by Neel (1942). It would be interesting to determine if the susceptibility of the slow hatching hens to the effect of storage was in part due to their inability to initiate growth after being placed in the incubator and/or to their inability to sustain growth after starting to develop, resulting in an increased mortality in the early stages of incubation. With such information it would be possible to associate the inherent hatching times of individual hens or strains to the much studied early and late "peaks" of embryonic mortality. A complex interrelationship exists between the variables examined in this preliminary study. These interrelationships can be further clarified with a multivariate analysis on more adequate data, and the preliminary results presented here suggest that such an effort would be justified. SUMMARY Eggs saved for a period of 12 days from 100 mass-mated yearling White Leghorn hens provided 474 fertile eggs for study. Hatchability was defined as the percent hatch of fertile eggs after 21 days of incubation, after 21J^ days and after 22*4 days of incubation. The relationship between hatching time, egg weight, preincubation storage time, and hatchability was studied at each of the three incubation periods. The distribution of hatching times was essentially normal with a mean hatching time of 21.125 days and a standard deviation of 8.96 hours. The regression of hatchability on days of holding was found to be significant and linear at all incubation times, but decreased sharply as the length of the allowed incubation time increased. The effect of holding time on hatching time would account for most of the effects of holding time on hatchability after 21 and 21J4 days, but not after 22^4 days of incubation. Hatching time was linearly associated (depending on interpretation) with holding time in the range observed. The regression coefficient was 1.28 hours. Within holding periods, egg weight showed a small but near significant correlation with hatch time. The correlation between egg weight and hatching time within hens was small and non-significant. Between hen means, however, these two variables were significantly correlated. A relationship between holding time and hatching time was observed within hens but not between hen means. Hens having the shortest hatching times were highest in hatchability at all three lengths of incubation and showed the greatest resistance to the effects of storage, particularly at the longer incubation times. This study suggests that hatching time and its interactions with holding time and egg weight may play an important role in the inheritance of hatchability and merits more extensive analysis on a larger volume of data. REFERENCES Abplanalp, H., and I. L. Kosin, 1953. Genetic variation of fertility and hatchability in the Broad Breasted Bronze turkey. Poultry Sci. 32: 321-331.

HATCHING TIME AND HATCHABILITY 633 Asmundson, V. S., 1947. Time held prior to incubation and hatchability of turkey eggs. Poultry Sci. 26: 305-307. Bartlett, M. S., 1947. The use of transformations. Biometrics, 3: 39-52. Bernier, P. E., L. W. Taylor and C. A. Gunns, 1951. The relative effects of inbreeding and outbreeding on reproduction in the domestic fowl. Hilgardia, 20: 529-628. Buss, E. G., 1956. Some factors which affect hatchability of chicken eggs at higher altitudes. Unpublished Ph.D. Thesis, Purdue University. Byerly, T. C, 1933. Some factors affecting the length of the incubation period. Proc. 5th World's Poultry Congress. Rome. Byerly, T. C., and S. J. Marsden, 1938. Weight and hatchability of turkey eggs. Poultry Sci. 17: 298-300. Coles, R., 1956. The influence of the hens egg weight on hatching. Poultry Sci. 35: 817-822. Cruz, S. R., and A. L. Romanoff, 1944. The effect of oxygen concentration on the development of the chick embryo. Physiol. Zool. 17: 184-187. Davis, G. T., 1955. Influence of oxygen concentration on hatchability and on selecting for hatchability. Poultry Sci. 34: 107-113. Ells, J. B., and L. Morris, 1947. Factors involved in hatching chicken and turkey eggs at high elevations. Poultry Sci. 26: 635-638. Funk, E. M., 1934. Factors influencing hatchability in the domestic fowl. Missouri Agr. Exp. Sta. Bui. 341. Hays, F. A., 1941. The importance of length of incubation period in Rhode Island Reds. Massachusetts Agr. Exp. Sta. Bui. 384. Henderson, E. W., 1950. Breed and egg weight variation in incubation period. Michigan Agr. Exp. Sta. Quart. Bui. 32: 520-525. Kaufman, L., 1958. Entwinklung and wachstum von huhnerembryonnen in frischen und gelagarten eiern. Archiv fiir Geflugelkunde, 12: 65-77. Kosin, I. L., 1950. A relationship between the length of storage and incubation period in Broad Breasted Bronze eggs. Poultry Sci. 29: 620-621. Kosin, I. L., 1954. Length of pre-incubation storage and heredity as factors affecting the hatchability of turkey eggs. Poultry Sci. 33: 24-28. Landauer, W., 1948. The hatchability of chicken eggs as influenced by environment and heredity. Connecticut (Storrs) Agr. Exp. Sta. Bui. 262. Lerner, I. M., 1951. Natural selection and egg size. Amer. Nat. 85 : 365-372. Lerner, I. M., and C. A. Gunns, 1952. Egg size and reproductive fitness. Poultry Sci. 3,1: 537-544. McCartney, M. G., and C. S. Shaffner, 1949. Chick thyroid size and incubation period as influenced by thyroxine, thiouracil and thyroprotein. Poultry Sci. 28: 223-228. McNally, E. H., and T. C. Byerly, 1936. Variation in the development of embryos of hens eggs. Poultry Sci. 15 : 280-283. Neel, J. K., 1942. A calibration of the development of the chick under improved conditions of incubation and the relationship of developmental rate to age of hen and hatchability record. Poultry Sci. 21: 294-300. Olsen, M. W., 1942. The effect of age and weight of turkey eggs on the length of the incubation period. Poultry Sci. 21: 532-535. Riddle, O., 1924. On the necessary gaseous environment for the bird embryo. Ecology, 5 : 348-362. Rogler, J. C, H. E. Parker, F. N. Andrews and C. W. Carrick, 1959. The effects of iodine deficiency on embryo development and hatchability. Poultry Sci. 38: 398^110. Scott, H. M., 1933. The effect of age and holding temperature on hatchability of turkey and chicken eggs. Poultry Sci. 12: 49-54. Smyth, J. R., and C. E. Howes, 1949. The inheritance of length of incubation period in chickens. Poultry Sci. 28: 782. Snedecor, G. W., 1940. Statistical Methods. Iowa State College Press, Ames, Iowa. Stephenson, A. B., 1950. Supplemental oxygen increases the hatchability of eggs at high altitudes. Farm and Home Sci., Utah Agr. Exp. Sta. 11: 65, 95-96. Taylor, L. W., 1949. Fertility and Hatchability of Chicken and Turkey Eggs. John Wiley & Sons, New York. Thompson, R. L., 1952. Incubation at high altitudes. The effects of wind, barometric pressure and humidity on foetal mortality in the hen's egg. Poultry Sci. 31: 497-509. Waite, R. H., 1919. The effect of age of eggs on their hatching quality. Maryland Agr. Exp. Sta. Bui. 233. Wilgus, H. S., and W. W. Sadler, 1954. Incubation factors affecting hatchability of poultry eggs. 1. Levels of oxygen and carbon dioxide at high altitudes. Poultry Sci. 33: 460-471. Williams, C, G. F. Godfrey and R. B. Thompson, 1951. The effect of rapidity of hatching on growth, egg production, mortality and sex ratios in the domestic fowl. Poultry Sci. 30: 599-606. Worth, B. C, 1940. Egg volumes and incubation periods. Auk, 57: 44-60.