Sexual Dimoqphism in Vertebral Number in the CapelinMallotus oillosus (Muller)

Bv JonN LewsoN HenrPacific Biological Station

(Recei,'ued for fwblicati'on Septeruber 1, 1937)

ABSTRACTVertebral counts of caoelin soawning in autumn on Vancouver Island were: males 65.65,

females 65.13. Difiering significantly, these demonstrate sexual dimorphism of a type rare

among fishes. The sexual difierence is not primarily associated with length, although thereis some relationship between length and vertebral number among females. The high degreeof sexual specialization in the fins is limited to size and integument and does not include meristicfeatures.

At spawning time the capelin shows a degree of sexual dimorphism unusual

among isospondylous fishes. Most noticeable are the four rows of modified scales

on the males, two on either side, which produce the villous ridges from which the

specific name is derived. Ma[es are distinctly larger than females-some five per

cent longef on the average in our material-and much more robust. Other differ:

ences between the sexes are to be found in the fins. These are larger in'the males,

especially those on the under surface of the body, and most noticeably the pectorals

and anals. These fins in the spawning male are otherwise modified. On the

anterior surface of the pectoral fins the rays bear membranous fidges and on the

posterior the rays are lined with tubercles. The first ten to twelve rays of the male

anal fin are much enlarged, a condition which is most noticeable approximately

between rays seven to ten, which are so enlarged at the distal ends as to practically

eliminate the intervening membrahe. The males among spawning fish are dis-

tinguished also by having tubercles on the lower part of the head including snout,

jaw, lower part of the operculum, branchiostegals, and isthmus.

Vertebral counts feported in the pfesent paper show that sexual dimorphism

has extended to vertebral number, the male capelin having on the average approxi-

mately one-half a vertebra more than the female. An exploratory investigation of

fin ray number gave no evidence of sexual dimorphism in other meristic characters'

Sexual difierences in meristic character in general afe fare among fishes.

Hubbs (1918) in referring to a sexual difference in anal fin ray number in certain

species of viviparous perches says, "But one other instance of this kind in the whole

class of fishes has come to the writer's attention; it is that of the South American

Poeciloid genus Cynolebias. Regan . . . ' I9I2 . . . ."

Vertebral count has been used so generally in the study of fish populations

that sexual difierences in that charactef, if present, should certainly have been

417

J. Bror. Bo. CeN. 3 (5) 1937.

J. B

iol.

Boa

rd C

an. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

ER

SIT

Y O

F M

ICH

IGA

N o

n 11

/18/

14Fo

r pe

rson

al u

se o

nly.

418

recorded and taken into consideration in analysing the data. This has not been

done by rnost investigators. On the other hand, Schaefer (1936) records theabsence of any significant sexual difference in vertebral number for surf smelt,Ilypomesws pretiosus; Tester (1937) states that there is no sexual difference forPacific herring, Clupea pallasii; and the writer obtained negative results on an-alysing data for some 3,@0 vertebral counts for pilchard, Sardinops caerulea.

In two species evidence of sexual dimorphism in vertebral number is availablein the literature. Hubbs (1925) found the males of Pacific anchovy, Engraulisiltordar tnordar to have a slightly higher vertebral number (45.77) than thefemales (45.65). This he points out to be "of probable but hardly certain signif-icance" (P c.0.02). Punnett (1904) working on the elasmobranch Spinar nigerfound the number of "whole vertebrae" to be significantly smaller among adultnales (44.62) than among adult females (45.06) (P much less than 0.01) and thenumber of total vertebral segments also to be smaller among the males (65.04)than among the females (65.47) (P much less than 0.01).

MATERIALS AND METHODS

The material used in making vertebral counts was collected about Vancouverisland with beach seines in four years as follows: Departure bay, October ?,1930; October 5, 1932; September 28, 1934; Hammond bay, October 6, 1936.The fish are readily taken only on the spawning grounds and the numbers of thesexes there are so unequal (333 males to 82 females in 1936) that samples forseveral years had to be examined in order to have suffrcient females for a satisfac-tory analysis.

After being preserved in formalin the vertebral column was exposed by re-

alizarin (Tining) 1927) to facilitate accurate counting. The counts were made

moving the flesh from one side of the body and the specimens were stained with

under tl-re lo'lv power of a binocular microscope. In making these counts the ter-minal ossicle (Ford 1928) (the hypural of many authors) was not included.

I-engths were determined on a measuring board, the caudal end point being

the end of the silvery area.In measuring paired fins, the distance from the insertion to the end of longest

ray was taken; for unpaired fins the height recorded was the length of the longesttay. The lengths and heights of fins have been expressed in parts per thousandof the length. All formed rays were counted in paired fins; in unpaired fins only

those anterior rays were counted which were half as long as the longest ray andthose posterior rays which rose from a separate pterygiophore.

RESULTS AND DISCUSSION

The results of the vertebral counts are shown in table I. For each group ofcounts is given the sum of the squares of the deviation of the counts from their

rnean (S(x-i)'9of Fisher 1932) trom which all other constants may be calculated,

and the standard error of the mean (oM).

J. B

iol.

Boa

rd C

an. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

ER

SIT

Y O

F M

ICH

IGA

N o

n 11

/18/

14Fo

r pe

rson

al u

se o

nly.

Yea

r9r0

L912

L9r4

L915

AIteara

4I

lill

se! 5, 64 65 66 5'l

6 2 5 t O 4 4 1 20 2 7 2 1 1 '

6 2 L 2 L 5 19 2 r 4 5

d r r 5 8 8 1 1 2 4o l E l , 1 7

d 1 i t l l L L to , l E t 2 2 0 5

3 , 29 r7r L1' 50e 8 ) 5 ? o t 6 r 5

T,c.er,r I. Vertebral number in capelin

l{alee

S@ of6 8 6 9 ' l o l v . s q u a r e a [ o . o M

, 55 .15 94 .L 98 o . ro

419

laEel s I

su otlT. aqusea [o. oU

6 5 . 0 5 " . 9

4 0 0 . 1 '

65 .50 r r .5 '0 o . I2I 65 . r ) tL .1 L5 0 . r9

| 65 .59 L ' .15 .9 20? 0 .0546 5 . 2 + 4 2 . L t ' 0 . 2 2

! 65 . ' , l2 10r .5 Lo? 0 ,105 r . 1 1 7 1 . L ? 9 o . r r

l r 65 .65 '9? .0 142 0 .045r 55 ,L t 185.4 16? 0 .082

From the table it may be calculated that the difference (0.52 excess for males)

betlveen the numbers of vertebrae of all the males and all the females is very highly

significant. (P much less than 0.01). For the separate years 1930 and 1936

highly significant <lifferences between the sexes are indicated. (P less than 0.01 )'Tlre difierences for the smaller 1932 and 1934 samples were not significant. How-

evef, they are in the same clirection, and when the counts for these two, years are

combinecl a difference of probable significance (P c.0.016) is indicated. More-

over, the lowest vertebral average for males is higher than the highest for females

by a cluarter of a vertebra. It appears, accordingly, that the difference in vertebral

number between the sexes is a regular occrlrrence and is independent of year class

variation.Tl-rere are indications of a relationship between length and vertebral number

sinrilar to that recorded by Tester (1937) for herring. (For female capelin less

than97 rnm. length, avefage vertebral number, 64.71; number of specimens, 34;

sunr of squafes, 2I: lor those more than 96 rnm. length, avefage vertebral number,

65.23 number of specimens, 133;sum of squares, 158: probabil ity of such a differ-

ence occurring by chance, 0.01). llor,vever, the arrangement of the data as in table

II inclicates that length is not the principal factor in producing the difference be-

tlveen the sexes.

Tenp II. Average vertebral numbers of capelin in each length group. The length distributionis given by the numbers in brackets indicating the number of counts entering into each average.

Ser sdyeF{ .

Males

1 a ? 2

t 9 t+L9t6 65.'.l

F e n a l e s

I I U -

2?:2,1'.)65 .90 (e I )5 r . 50 (10 )

- 9 5 9 ? - 1 o o

65 .6 r \ 6 ). 6 b ( I )

6 5 . 5 6 ( 9 )( r ) 5 6 . 0 0 ( 5 )

64.' \ ') er.ez(I2) '2..ltlrzJ li..irtl 64 ( 1)z,l,Ezl"i\ ii.i\\til it'iz\'ai .ii,'iii 6' o ( z,

Length 1a nilllnetres.

Lol - 10, 1o/ t - to6 Ig? - 109

Z!,:X\l'il 3i:33tt3J ti:9iti6tzi'iiitl, z3:?iiiil 2?'.;1iz'l

r930I a 1 2

19 t4 5 5 ( 1 )

J. B

iol.

Boa

rd C

an. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

ER

SIT

Y O

F M

ICH

IGA

N o

n 11

/18/

14Fo

r pe

rson

al u

se o

nly.

420

Averages for the proportional fin lengths of ten males and eleven females are:pectoral, males 186, females 133; ventrals, males 164, females l4l; anal, males 90,females 55; dorsal, males 127, females 722 (nine only). In spite of the stronglymarked sexual dimorphism in the size of fins on the ventral surface of the bodythere is no evident difference in meristic features, as is shown by the followingray counts on the same material: pectoral, males 17.8, females 17.8; ventrals, males8, females 8; anal, males2O.2, females 20.2; dorsal, males 12.3, females 12.4. Thesecounts agree with those given for fish taken in northern Europe by Schnakenbeck(1931) for the dorsal (10-13) and pectoral (15-20) fins, but disagree with thoseof Sleggs (1933) for the dorsal (13-15) for Newfoundland capelin. on the otherlrand, the anal ray counts agree with that of Sleggs (19-22).rather than those ofthe former author.

REFERENCES

Frsnnn, R. A. Statistical methods for research workers. oliver and Boyd, Edinburgh, 4thed., 307 pp., 1932.

Fonn, E. !. Mar. BioL Ass. t5 (1), 237-266, 1928.Hrysss, C. L. Proc. Biol. Soc. Wash. 31, 9-13, 1918.

Calif . Fish Game Comm. Fi,sh ButI. 8, 23 pp., 1925.PuNNrrr, R. C. Biometriko 3 (4), 313-362, 1904.RrcaN, C. T. Ann. Mag. Nat. Fftsr. Ser. 8, tO, 641-642, 1912.Scrrarron, M. B. Wash. State Dep. Fish. BioI. Rcp.35b, l-45, 1936.Scnwernltnecr, W. Mallotus aillosus. In Joubin, I. Fauna Ichthyologique de I'Atlantique

Nord, No. 8. Cons. Perm.. Expl. Mer., 1931.Srnccs, G. F. Rep. Nfld. Fish Res,. Comm. 1 (3), 1-66, 1933.Ti,xrNc, Y. J. Conseil 2, 50-52, 1927.Trsren, A. L. I. Biol, Bd. Can.3 (2), 108-144,1937.

J. B

iol.

Boa

rd C

an. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

ER

SIT

Y O

F M

ICH

IGA

N o

n 11

/18/

14Fo

r pe

rson

al u

se o

nly.