- Email: [email protected]
Laboratory Prediction of the Fertilizing Capacity of Cock Semen
Laboratory Prediction of the Fertilizing Capacity of Cock Semen D. M.
COOPER
Houghton Poultry Research Station, Huntingdon, England AND J. G. ROWELL Agricultural Research Council Statistics Group, School of. Agriculture, Cambridge, England (Received for publication September 4, 1956)
T
EXPERIMENTAL METHODS
The following qualities were measured in the laboratory: initial motility (rated from 0 to 4), initial pH, sperm density (duplicate counts on a haemocytometer), percentage of dead sperm (nigrosin-eosin staining technique with duplicate slides and duplicate counts of 100 on each slide) and time taken to reduce resazurin to the pink and white end points. Fertility estimates from each sample were obtained from inseminations of 4 hens; each hen received 0.1 ml. of undiluted semen immediately after collection. To obtain sufficient semen for insemination and the laboratory tests it was necessary to pool semen from 3 cocks; there were 8 such groups. Eggs were collected for a fortnight following insemination, set once weekly in a forced-draft incubator and individually candled at 96 hours for evi-
dence of embryonic development. The experiment was designed as an 8 X8 Latin square, letters corresponding to groups, columns to dates of collection and rows to times of collection. Semen was collected twice weekly for 4 weeks and collections and inseminations were carried out with about 12 minute intervals between groups. BIOMETRICAL METHODS AND RESULT
Studies of variations between collections, which were possible where replicated observations were made, suggest that sperm density variations within collections are small compared with variations between collections, so that if an estimate of the average density for a particular cock is required, there is little to be gained by replicating counts within collections. For percentage dead sperm, on the other hand, appreciable gains in precision are possible both by replicating counts on a given slide and by replicating slides from a given collection. In the regression analyses, fertility (transformed by arc sin \/p) was the dependent variable and the following independent variables were considered: initial pH, density, live density (density X proportion live), percentage dead sperm (transformed by arc sin \/p) and the resazurin reduction times (transformed by log /). The measurement of initial motility was subjective and was not used in the statistical analyses; Group G, Table 1 shows, however, that semen with a very low fertilizing capacity exhibited poor
284
Downloaded from http://ps.oxfordjournals.org/ at NERL on June 15, 2015
HE purpose of the experiments was to examine the possibility that certain laboratory characteristics of cock semen could be used to predict its fertilizing capacity. Shaffner and Andrews (1948) found that fertility in the chicken was significantly correlated with initial motility, methylene blue reduction time and survival time at 4 ° C ; the correlations between fertility and volume of ejaculate and density were not significant. McCartney (1956), working with the turkey and measuring initial motility, initial pH, density and volume of ejaculate, found no significant correlations with fertility.
285
FERTILIZING CAPACITY OF SEMEN
TABLE 1.—Group means of fertility and the laboratory characteristics, residual standard deviations and %of variation in fertility explained by each laboratory measurement Group of cocks Fertility (%) Resazurin reduction time to pink (mins.) Resazurin reduction time to white (mins.) Initial pH Density (millions/mm.*) Percentage dead Live density (millions/mm.>) Initial motility
Residual standard deviation
A
B
C
D
£
F
G
H
75.7
68.2
64.6
71.5
77.3
76.2
34.9
75.1
16.6
%
variation explained
9.3
12.8
11.6
14.3
12.2
12.5
26.3
14.5
2.55
67.2
31.2 7.47 4.41 60.6
49.3 7.48 4.64 56.0
42.5 7.44 3.90 48.0
52.3 7.47 4.02 49.6
43.6 7.42 3.33 46.6
43.6 7.53 3.38 49.4
89.4 7.58 2.91 73.4
51.7 7.42 3.21 47.4
8.40 0.134 0.700 13.1
63.8 37.5 0.0 60.6
1.69 3.7
2.05 3.8
2.03 3.6
1.96 3.8
1.81 3.9
1.74 3.6
0.82 2.9
1.81 3.9
0.569
55.5
—
—
lizing capacity (75%), measured over a 14 day period, and those whose semen has a moderate fertilizing capacity (65%). SUMMARY
(1) Laboratory tests cannot differentiate between collections within cocks. (2) By measuring any one of resazurin reduction time, initial pH or percentage dead sperm, it is possible to identify cocks whose semen has an abnormally low fertilizing capacity. After allowing for variations in resazurin reduction time (pink), variations in fertility between cocks are considerably reduced. The small residual variations can to some extent be explained by percentage dead sperm. (3) In view of the small number of groups of cocks in this experiment, the conclusions reached are somewhat tentative. A further experiment is proposed, in which individual cocks will be studied and fewer laboratory tests used. ACKNOWLEDGMENTS
The authors are grateful to Lord Rothschild, F.R.S. for his interest in this work and to Mr. S. Lipton, Rothamsted Experimental Station, for carrying out most of the calculations on the Elliott 401 electronic computer. REFERENCES McCartney, M. G., 1956. Relation between semen quality and fertilizing ability of White Holland turkeys. Poultry Sci. 35: 137-141. Shaffner, C. S., and F. N. Andrews, 1948. The influence of thiouracil on semen quality in the fowl. Poultry Sci. 27:91-102.
Downloaded from http://ps.oxfordjournals.org/ at NERL on June 15, 2015
motility. When semen from different cocks is pooled, lower correlations between fertility and the laboratory measurements may be expected between groups than would be obtained using the semen of individual cocks, since variations between groups are less than variations between individuals. Regression analyses of fertility on the laboratory measurements were carried out on the residuals of the Latin square and on the group means. None of the residual regressions was significant, suggesting that it is not possible to discriminate between the fertilizing capacity of different collections of a given cock by means of these laboratory tests. Regressions on the group means were significant for all variables except density. The group means, the residual standard deviations which measure collection differences, and the percentages of variance in fertility accounted for by the regressions on the independent variables are given in Table 1 [percentage accounted for = 100 (1 —residual mean square/total mean square)]. Bivariate regression analyses, carried out to see if better predictions are possible by using more than one characteristic, showed that significant improvement was only obtained when reduction time (pink) and percentage dead sperm were used as independent variables, in which case 80.7% of variance in fertility was accounted for. This suggests that it may be possible to distinguish between cocks whose semen has a high ferti-
COOPER
Houghton Poultry Research Station, Huntingdon, England AND J. G. ROWELL Agricultural Research Council Statistics Group, School of. Agriculture, Cambridge, England (Received for publication September 4, 1956)
T
EXPERIMENTAL METHODS
The following qualities were measured in the laboratory: initial motility (rated from 0 to 4), initial pH, sperm density (duplicate counts on a haemocytometer), percentage of dead sperm (nigrosin-eosin staining technique with duplicate slides and duplicate counts of 100 on each slide) and time taken to reduce resazurin to the pink and white end points. Fertility estimates from each sample were obtained from inseminations of 4 hens; each hen received 0.1 ml. of undiluted semen immediately after collection. To obtain sufficient semen for insemination and the laboratory tests it was necessary to pool semen from 3 cocks; there were 8 such groups. Eggs were collected for a fortnight following insemination, set once weekly in a forced-draft incubator and individually candled at 96 hours for evi-
dence of embryonic development. The experiment was designed as an 8 X8 Latin square, letters corresponding to groups, columns to dates of collection and rows to times of collection. Semen was collected twice weekly for 4 weeks and collections and inseminations were carried out with about 12 minute intervals between groups. BIOMETRICAL METHODS AND RESULT
Studies of variations between collections, which were possible where replicated observations were made, suggest that sperm density variations within collections are small compared with variations between collections, so that if an estimate of the average density for a particular cock is required, there is little to be gained by replicating counts within collections. For percentage dead sperm, on the other hand, appreciable gains in precision are possible both by replicating counts on a given slide and by replicating slides from a given collection. In the regression analyses, fertility (transformed by arc sin \/p) was the dependent variable and the following independent variables were considered: initial pH, density, live density (density X proportion live), percentage dead sperm (transformed by arc sin \/p) and the resazurin reduction times (transformed by log /). The measurement of initial motility was subjective and was not used in the statistical analyses; Group G, Table 1 shows, however, that semen with a very low fertilizing capacity exhibited poor
284
Downloaded from http://ps.oxfordjournals.org/ at NERL on June 15, 2015
HE purpose of the experiments was to examine the possibility that certain laboratory characteristics of cock semen could be used to predict its fertilizing capacity. Shaffner and Andrews (1948) found that fertility in the chicken was significantly correlated with initial motility, methylene blue reduction time and survival time at 4 ° C ; the correlations between fertility and volume of ejaculate and density were not significant. McCartney (1956), working with the turkey and measuring initial motility, initial pH, density and volume of ejaculate, found no significant correlations with fertility.
285
FERTILIZING CAPACITY OF SEMEN
TABLE 1.—Group means of fertility and the laboratory characteristics, residual standard deviations and %of variation in fertility explained by each laboratory measurement Group of cocks Fertility (%) Resazurin reduction time to pink (mins.) Resazurin reduction time to white (mins.) Initial pH Density (millions/mm.*) Percentage dead Live density (millions/mm.>) Initial motility
Residual standard deviation
A
B
C
D
£
F
G
H
75.7
68.2
64.6
71.5
77.3
76.2
34.9
75.1
16.6
%
variation explained
9.3
12.8
11.6
14.3
12.2
12.5
26.3
14.5
2.55
67.2
31.2 7.47 4.41 60.6
49.3 7.48 4.64 56.0
42.5 7.44 3.90 48.0
52.3 7.47 4.02 49.6
43.6 7.42 3.33 46.6
43.6 7.53 3.38 49.4
89.4 7.58 2.91 73.4
51.7 7.42 3.21 47.4
8.40 0.134 0.700 13.1
63.8 37.5 0.0 60.6
1.69 3.7
2.05 3.8
2.03 3.6
1.96 3.8
1.81 3.9
1.74 3.6
0.82 2.9
1.81 3.9
0.569
55.5
—
—
lizing capacity (75%), measured over a 14 day period, and those whose semen has a moderate fertilizing capacity (65%). SUMMARY
(1) Laboratory tests cannot differentiate between collections within cocks. (2) By measuring any one of resazurin reduction time, initial pH or percentage dead sperm, it is possible to identify cocks whose semen has an abnormally low fertilizing capacity. After allowing for variations in resazurin reduction time (pink), variations in fertility between cocks are considerably reduced. The small residual variations can to some extent be explained by percentage dead sperm. (3) In view of the small number of groups of cocks in this experiment, the conclusions reached are somewhat tentative. A further experiment is proposed, in which individual cocks will be studied and fewer laboratory tests used. ACKNOWLEDGMENTS
The authors are grateful to Lord Rothschild, F.R.S. for his interest in this work and to Mr. S. Lipton, Rothamsted Experimental Station, for carrying out most of the calculations on the Elliott 401 electronic computer. REFERENCES McCartney, M. G., 1956. Relation between semen quality and fertilizing ability of White Holland turkeys. Poultry Sci. 35: 137-141. Shaffner, C. S., and F. N. Andrews, 1948. The influence of thiouracil on semen quality in the fowl. Poultry Sci. 27:91-102.
Downloaded from http://ps.oxfordjournals.org/ at NERL on June 15, 2015
motility. When semen from different cocks is pooled, lower correlations between fertility and the laboratory measurements may be expected between groups than would be obtained using the semen of individual cocks, since variations between groups are less than variations between individuals. Regression analyses of fertility on the laboratory measurements were carried out on the residuals of the Latin square and on the group means. None of the residual regressions was significant, suggesting that it is not possible to discriminate between the fertilizing capacity of different collections of a given cock by means of these laboratory tests. Regressions on the group means were significant for all variables except density. The group means, the residual standard deviations which measure collection differences, and the percentages of variance in fertility accounted for by the regressions on the independent variables are given in Table 1 [percentage accounted for = 100 (1 —residual mean square/total mean square)]. Bivariate regression analyses, carried out to see if better predictions are possible by using more than one characteristic, showed that significant improvement was only obtained when reduction time (pink) and percentage dead sperm were used as independent variables, in which case 80.7% of variance in fertility was accounted for. This suggests that it may be possible to distinguish between cocks whose semen has a high ferti-