THE SEXES IN 2-EGG CLUTCHES OF RINGNECKS, Streptopelia risoria.
by
Wilmer J. Miller and Paul Hollander

A myth concerning the sexes in the 2-egg clutches of doves and pigeons has never been erased completely. In W. M. Levi's 1957 book The Pigeon, he covers the topic well for the domestic pigeon, Columba livia. But Aristotle before 322 B.C. wrote "The pigeon as a rule, lays a male egg and a female egg and generally allows a day's interval to ensue and then lays the second egg." Aelian (about 222 A.D.) is more extreme. "The female brings forth two eggs, the first of which always produces a male, the second a female."

Later authorities follow in a similar manner, although not quite so absolute, as can be noted in Fulton and Lumley (1895) "The two eggs which a pigeon lays are almost invariably cock and hen, the cock being almost invariably the first one".

Levi, after quoting such authorities, correctly indicates the expected clutch ratio of 25% males, 50% males/females, and 25% females. This follows from the statistical knowledge of the laws of sex chromosomal segregation. In most species of birds the actual presence of the W is not proven and W could represent the absence of a chromosome leading to the same generalizations anyway. The slash represents the chromosome and is often omitted for brevity. One chromosome comes from the father and the other from the mother. Likewise one is transmitted to the offspring from the father and one from the mother. ZZ controls production of a male and ZW a female in birds (interacting with many other genes, of course).

 

Of course, other factors might influence the ratio, but those known or suspected in organisms in general are minor. And why not actually look to see what it really is?!

Levi cites several such experiments including some tabulations of the Palmetto Pigeon Plant data by Dr. W. F. Hollander for sex-linked matings in the pigeon. Small deviations from the exact expected are expected, of course. But they are all within simple chance deviations.

With numbers in the thousands Levi slays the dragon of this myth. But what about ringnecks, Streptopelia risoria, which is another genus and species. Some people on the internet have wondered about the sex distribution in the clutches. So I have started compiling the results of over 20 years of results of genetic breeding of ringnecks for other purposes. (That number could be expanded to 60 years if small numbers of mating were included.)

I start here with results from the Iowa State University dove colony of 200 matings of birds descended from albino ringnecks imported from Japan. These doves were originally designated Streptopelia douraca. But they are placed within risoria now. And they are completely fertile with our familiar ringnecks. In fact the original birds in the importation of 3 males and 3 females were so infertile among themselves (from inbreeding?) that outcrosses had to be made to save the albino gene. Thus, this population is more regular risoria than the original stock.

One bias in the study is that males evince their sex by bow-cooing at variable times but weeks to months before the females are evident by laying. Since many of the offspring had to be disposed of (usually donated), before being sexed, more males were known than females. Also keeping BOTH members of a clutch was infrequent, so that the numbers from this group are smaller.

Fortunately, ringnecks also have sex-linked colors allowing sex to be noted at hatching from particular appropriate matings (Cole, 1930). The sex-linked alleles controlling the dilution colors blond and white and the wild type dark allow the sex to be determined at hatching if a lighter colored male has a darker colored mate. The genes controlling these are labeled D+ for dark, dB for blond and dw for white in order of decreasing dominance.

Some other matings can be termed partial sex-linked matings, since half the daughters can be distinguished at hatching. For example, the mating of dBdw x dB is one of this latter type. Half the daughters will be white and, therefore, female. The other half of the daughters will be blond and not distinguishable from blond males. So the results are divided into sex-linked matings, sex-linked partial and non-sex-linked (autosomal) matings.

The other methods of sexing were behavior (bow-coo from males, 100% trustworthy only in context of the presence of several doves as in a holding pen, and other associated behaviors); the "look-see" method using an infant nasal speculum inserted into the cloacae of mature individuals (Miller and Wagner, 1955); and the actual mating success in producing offspring.

Since records were taken 3 times a week, and with experience in noting the squab's development including the segregating color differences available, the first squab versus the second was almost always known. Color differences were usually evident at hatching so that the older squab could almost always be detected. Segregation and assortment were the general situation, since genetic numbers to fit Mendelian ratios were being studied as well as blood types and other characters such as silky (Miller 1956). In the minority of cases without color differences evident, still the older squab could almost always be detected since the growth of the feathers filling in the area about the corner of the mouth was more advanced in the older squab. So Aelian's notion that the first egg was the male could also be checked.

To reiterate, in doves the triple allelic series of sex-linked alleles are dark, blond, white in order of decreasing dominance. For example, a dark female mated to a blond male resulted in dark sons and blond daughters detectable by the bill and skin color at hatching.

We also had what may be termed partial sex-linked matings. These consisted of the males which were heterozygous for dark, blond or white carrying a second sex-linked color controlling gene which was recessive to that of the mother. For example, if a blond male which was heterozygous for white were mated to a blond female, then half the daughters would be white. All such whites would be female. The blond daughters could not be detected as females since the males also would be blond (Cole, 1930, Miller, 1989).

Table 1. Two egg clutches that have been sexed.

Duration: March 1974 to May 1992 Hypothesis: A 1:1:1:1 ratio of these classes reflects expected equal chance distribution of the sexes.

Ringnecks of Albino (Japanese) extraction.

Egg #

1 2

1 2

1 2

1 2

male/male

male/female

female/male

female/female

Totals

Non-sex-linked Autosomal [Regular methods*]

40

40

31

35

= 146

Sex-linked

17

16

20

19

= 72

partial matings.

6

3

4

5

= 18

Totals

63

59

55

59

= 236

expected:

59

59

59

59

236

deviation= d

4

0

4

0

d2

16

0

16

0

c 2 = S d2/e = 0.2712 + 0 + 0.2712 + 0 = 0.5424

df = 3 P~ .90 Therefore, the hypothesis is supported very well.

 

Table 2. Results in a single large family 436 from 408Y, the long-lived female. 

7 7 11 6 = 31 

One expects a 1:1:1:1 ratio for these four classes. By inspection one can note the results are a good fit even without a X2 test of significance. The sex-linked results are almost too good. 

Levi listed two families purporting to show a sex preference in some pigeon families. This is attributable to chance. If you have a large number of families, you can find deviations in such directions. My numbers are not so large but two samples of raw data follow. Two eggs in a clutch are the rule in most dove/pigeon species. 

Table 3. Excess sex in a non-sex-linked mating and a sex-linked mating. 

Family J187

Family J187

Eggs

Sex

Eggs

Sex

E, F

F, M

G, H

M, M

J, K

F, F

J, K

F, F

L, M

F, M

L, M

M, M

Q, R

M, F

N, P

F, M

S, T

F, M

Q, R

M, F

W, X

F, F

S, T

M, M

4 males: 8 females

U, V

M, F

W, X

M, F

10 males: 6 females

 

The results in the ISU colony were subject to 3 episodes of paratyphoid, Salmonella typhimurium epidemics brought in by wild house mice. This is not believed to have influenced the sex ratio. But it did reduce the numbers of 2-egg clutches successfully sexed.

 

A Chi Square test for the total results for J mating type birds yields a X2 of .54 with 3 degrees of freedom and a Probability of approximately .91 which is almost too good a fit. 

The results for regular ringnecks, Streptopelia risoria, are summarized in tables 4 and 5, separating the sex-linked, partially sex-linked and autosomal results.

Table 4. Two egg clutches that have been sexed.

Duration: November 1955 to April 1965 Families 1-300

Hypothesis: A 1:1:1:1 ratio of these classes reflects expected equal chance distribution of the sexes.

 

male/male

male/female

female/male

female/female

Totals

Regular Methods*

75

67

58

53

253

Sex-linked

24

42

35

34

135

S-L Partial

11

19

20

15

65

Totals

110

128

113

102

453

expected

113.25

113.25

113.25

113.25

= 453

deviation= d

3.25

14.75

0.25

11.25

d2

10.5625

217.5625

0.0625

126.5625

c 2 = S d2/e = 0.0933 + 1.9211 0.0006 + 1.1175 = 3.1325

df = 3 P ~ .37 > .05 Therefore, the hypothesis is supported.

 

Table 5. Two egg clutches that have been sexed.

Duration: February 1968 to July 1980 Families 308-500

Hypothesis: A 1:1:1:1 ratio of these classes reflects expected equal chance distribution of the sexes. 

male/male

male/female

female/male

female/female

Totals

Regular Methods*

57

58

48

59

222

Sex-linked

18

18

17

20

73

S-L Partial

6

9

8

12

35

Subtotals

81

85

73

91

330

expected

82.5

82.5

82.5

82.5

deviation= d

1.5

3.5

9.5

8.5

d2

2.25

12.25

90.25

72.25

c 2 = S d2/e = 0.0273 + 0.1485 + 1.0939 + 0.8758 = 0.9764

df = 3 P = ~ 0.80 in good agreement with the hypothesis. Therefore, there is no significant difference from the expected equally likely frequency of sex distribution. 

*Regular methods consist of behavior (especially bow-coos by adult males in mixed populations), vent sexing (Miller and Wagner 1955 , observed copulation, actually laying eggs, production of young by pairs isolated in cages. The first two methods predominantly were used.

 

Table 6. Two egg clutches that have been sexed.

Duration: July 1977 to Oct 1980 ; Families 501-599 -671 June 92

 

male/male

male/female

female/male

female/female

Totals

Regular Methods

19

27

15

15

76

Sex-linked

8

14

13

14

49

S-L Partial

12

18

5

15

50

Subtotals

39

59

33

44

175

expected

43.75

43.75

43.75

43.75

deviation= d

4.75

15.25

10.75

0.25

d2

22.5625

232.5625

115.5625

0.0625

 

c 2 = S d2/e = .5193 + 5.3157 + 0.0407 + 0.0014 = 5.88 

df = 3 P ~ .12 = a low probability but not significant at the 5% level. 

Combining all these results of the ringnecks sexed 2-egg clutches results in 1,194 clutches so analyzed.

 

Table 7. Two-egg clutches that have been sexed.

Duration:1955 to 1992 Families 1-671

Hypothesis: A 1:1:1:1 ratio of these classes reflects expected equal chance distribution of the sexes.

 

male/male

male/female

female/male

female/female

Totals

Regular Methods*

191

192

152

162

697

expected

174.25

174.25

174.25

174.25

697

deviation

16.75

17.75

2.25

12.25

d2

280.5625

315.0625

495.0625

150.0625

c 2 = S d2/e = 1.6101 + 1.8081 + 2.8411 + 0.8612 = 7.1205

df = 3 P~ .068 > .05 Therefore, the hypothesis is supported.

* * * * * * * * * * * *

male/male

male/female

female/male

female/female

Totals

Sex-linked

67

90

85

87

329

e expected

82.25

82.25

82.25

82.25

d deviation

15.25

7.75

2.75

4.75

d2

232.5625

60.0625

7.5625

22.5625

c 2 = S d2/e = 2.8275 + 0.7302 + 0.0919 + 0.2743 = 3.9239

df = 3 P~ .34 Therefore, the hypothesis is supported.

* * * * * * * * * * * *

male/male

male/female

female/male

female/female

Totals

S-L Partial

35

49

37

47

168

e expected

42

42

42

42

168

d deviation

49

49

25

25

d2

232.5625

60.0625

7.5625

22.5625

c 2 = S d2/e = 1.1667 + 1.1667 + 0.5952 + 0.5952 = 3.5238

df = 3 P ~ .46 Therefore, the hypothesis is supported.

* * * * * * * * * * * *

male/male

male/female

female/male

female/female

Totals

Totals

293

331

274

296

1194

e expected

298.5

298.5

298.5

298.5

d deviation

5.5

32.5

24.5

2.5

d2

30.25

1056.25

600.25

6.25

c 2 = S d2/e = 0.1013 + 3.5385 + 2.0109 + 0.0209 = 5.6716

df = 3 P ~ .14 Therefore, the hypothesis is supported.

 

There does seem to be an excess of male/female clutches. It does not reach the significance level considering all 4 classes. However, comparing only the male/female versus female/male classes which should be in a 1:1ratio, we have 331 versus 274 with a surplus of male/female clutches.

The X2 of 5.37 (df=1) is significant with a probability of about .02 . There could be some bias, but it is possible that a slight excess of males in the first egg occurs. It is known in several species for male offspring to be in excess. ( ref. ) If so, then it also should occur in pooling the male/male male/female classes and the female/male female/female classes. Then we have 624 versus 570 yielding a chi-square of about 2.44 with one df for a probability of approximately 0.12 which is not significant at the 5% level. If there is a bias for male offspring in the first egg laid, then it requires data well in excess of 1,000 2-egg clutches.

The overall sex ratio in these data is 1191 males to 1197 females. 

 

References: 

Cole, L. J. 1930. A triple allelomorph in doves and its interspecific transfer. Anatomical Record 47: 389. Aviculture 2:27-30 Anat. Rec. 47: 389 

Cole, L. J., T. J. Painter, and A. Zeimet 1928. Studies on pigeon hybrids I. Aberrant sex ratios. Anat. Rec. 41: 112 

Levi, W. M. 1957. The Pigeon. Levi Publishing Co., Inc. Sumter, S.C. pp 667. Lumley, W. F. 1895 Fulton's book of pigeons edited by Lewis Wright

London: Cassell and Co. Ltd. 

Miller, W. J. and Wagner 1955. Sexing mature Columbiformes by cloacal characters. The Auk 72: 279-285. 

Miller, W. J. 1956. Silky plumage in the ring neck dove. Jour. Heredity. 47: 37-40. 

Miller, W. J. 1989. Dark-Blond-White: Sex-linked alternatives in ringneck doves, ADAN Jan./Feb. pp. 6-10.