California. Dept. of Fish and Game.

California fish and game (Volume 38, no. 4) online

. (page 4 of 15)
Font size
QR-code for this ebook

mercially as many others of the genus, it is easily available in sufficient
numbers to permit an outline of its life history.

This study was carried out at the Hopkins Marine Station of Stanford
University, Pacific Grove, California. Work began in the summer of 1929,
at which time 280 fish were tagged. Nothing more was done until June,
1930, when tagging was resumed for a few months along with regular
sampling, which was continued for an entire year.

When the investigation was started, no work had been carried out on
the life histories of the fishes of the Monterey Bay region and there was
little or no knowledge of the seasonal changes in the phj^sicochemical
conditions of the waters. As a consequence we had no ideas in regard to
the difficulties inherent in the problem. As the work progressed and as


Blue rockfish from Monterey Bay, April 1, 1931 . Phoiograph by J. H. Wales and
Wm. A.Dili, April, 1931.

1 Submitted for publication May, 1952. Modified from a typewritten thesis submitted to
Stanford University, June, 1932, in partial fulfillment of the requirements for the
degree of Master of Arts. The common name "blue roclvfish" replaces the depart-
ment's old official name of "priestfish." It is believed that the new name more nearly
conforms with popular usage.




the data of the hyclrobiological survey of Hopkins Marine Station in-
creased in number, the true problem began to be realized.

The hydographical work (Skogsberg, 1936) disclosed that while the
deeper waters of this region were characterized by a very distinct rhythm,
the superficial waters were distinguished by the apparent lack of regu-
larity in their changes, in other words b}' irregular fluctuations. Figure 2
illustrates the fact that there is less similarity in surface temperatures
from one year to another than between deep temperatures of different
years. It seems altogether possible that this fact accounts for the remark-
able differences in scale features and in the growth of blue rockfish of
different year classes. Because this is a shallow water species, its growth
mirrored very closely the lack of regularity in its environment. Those
features which usually express the rhythm of growth in fishes were
found to be very poorly developed.

The writer wishes to express his appreciation for the help and encour-
agement of the late Dr. Tage Skogsberg of Stanford University, at whose
suggestion this investigation was undertaken.












55 4-^



53.6 (^


51.8 I


50 D















Water temperatures of Monterey Bay in 1929 and 1930. Surface temperatures are
represented by solid lines; temperatures at 50 meters by broken lines.


]\Iost of the specimens were secured by hook and line from a skiff just
off the rocks at Hopkins ^larine Station, which is situated on ]\Iussel
Point, one and one-fifth miles west of Monterey, California. Nets and rock-
cod lines, consisting of many hooks, were tried, Avithout reasonable suc-
cess. The best gear for catching the fish was found to be a short pole with
a line of equal length and one or two hooks. The bait was squid, blue
rockfish, Oxyjulis (the senorita, a common shore fish) or sardine. The
method is not only extremely time consuming but undoubtedly tends to
be strongly selective. As a consequence, the material collected was largely
of one year class and even this was so poorly represented that a good
length frequencj^ analysis was impossible.


Some samples of the commercial catch were taken at the Monterey fresh
fish markets. Since these fish were known to have been caught close to
the main sampling area, the writer did not hesitate to include them with
the rest of the specimens.

Scale samples were taken from all fish handled. In the younger age
classes about 98 percent of the samples could be read. On the other hand,
many of the scales of the older, market-caught fish could not be read.
In *S*. mysfinns the scales seemed to be more easily read than the otoliths
and opercular bones but in some of the other commercially important
species of this genus well-defined rings could be found in the otoliths.
This peculiarity is apparently correlated with the circumstance that the
more important commercial species occur in deep Avater. where there
is a distinct seasonal rhythm in the temperature during the year.
>S'. mystinus, on the contrary, spends all or most of its life in shallow
water, which in Monterev Bav is not characterized bv such regular
changes, but by irregular fluctuations (Figure 2). If, as is generally
assumed, the rhythm expressed by the bony structures actually is the
material record of a metabolic rhvthm. then this differential behavior
of the bony structures of these species must be considered as very strong
evidence that the temperature is at least the fundamental factor in bring-
ing about this rhythm. It should be noted that no season of the year is
characterized bv searcitv of food in the Monterev region.

In an effort to find out something about the rate of growth, the rate
of addition of scale circuli was determined simply by counting the num-
ber of circuli on each scale beyond a certain point which Avas established
for each year class separately. For the zero class, all the circuli were
counted (Figure 6) ; for class l-\-. all beyond ''secondary ring" C
(Figure 7) ; for class 2+, all beyond annulus D (Figure 8) ; and for
class 3+, those beyond annulus F (Figure 9).

As has already been noted, 280 fish were tagged in September 1929,
and 770 in the summer and fall of 1930. The fish were caught bv hook
and line, the fork length obtained, a sample of scales taken, a tag ^
attached to the operculum and the fish liberated. In the recaptured fish
the tag had worn the hole in the opercle slightly larger and in some cases
evidently had irritated the skin. The tags themselves often became cov-
ered with a fine growth of algae. The loss of tags is not believed to have
been great.

A number of fish were retaken after having been tagged a week or
two but only those which had been out for at least a month were con-
sidered in the following analysis. Seven fish tagged in 1929 were retaken
by staff members of the Hopkins ^Marine Station from one month to
nine months and twenty-five days after liberation. Ten fish tagged in
1930 were retaken by the writer within the following five months. Xone
of these fish were out during the period of annulus formation. In other
words, out of 1,050 tagged fish only 17 were recaptured and examined.
Not onh^ was this a very small number but its significance was further
decreased by the facts that the specimens belonged to three different
year classes and that they were recaptured during different seasons of
the year and after varying periods following their liberation. However,
the data obtained from the tagged fish were valuable in checking the
general scale and length frequencj^ studies.

2 The tags were No. 3, noncorrosive metal fish tags made by the Salt Lake Stamp Com-
pany, Salt Lake City, Utah.


The average monthh^ growth of the tagged specimens was 2.46 mm.
Assuming that there is no differential seasonal growth, this would cor-
respond to an average yearly increment of 29.52 mm. (Ig^ inches)
within the first three year classes. The supposition of uniform growth
is not unreasonable. The fact that the annuli are very poorly marked
indicates but a slight decrease in growth during the time of their for-

The average increase in the number of scale circuli was 4.34 each month
or 52.08 a year. The first three year classes showed no significant differ-
ence in this regard. Class 1+ added 5.7 circuli; class 2-\-, 4 circuli; and
class 3+, 4.9 circuli per month. These latter values probably do not
represent the average trend, since it would be most unusual to find a
greater growth rate in class 3-(- than in class 2-\-.

Despite this apparent aberration, a comparison with the results pre-
sented in the f olloAving pages will reveal a fair agreement between tagging
and scale studies, thus showing that the former strongly substantiate
the general results and conclusions of the report.

No definite information concerning the movements of this species was
gained b}^ the use of tags. This subject will be mentioned later. A few fish,
both with and without tags, were kept in an aquarium for some months,
but it Avas believed unwise to include observations made under such con-


The form of the blue rockfish scale can be seen in the photographs. The
most noticeable and bv far the most noteworthv variations in the distance
between circuli occur at the "annuli," which usually appear as more or
less definite bands across the scale. As will be seen from Figure 3, the
relative depth of the body in this species gradually decreases with age.
Since the number of scales is the same throughout life, it follows that
the growth of these structures must conform to this change in body pro-
portions. In other words, as the scales grow larger they must assume a
somewhat more elongate shape. In Figure 10 we can see what is accom-
plished by noticing the distance between side checks No. 1 and No. 2 as
compared with the distance between annuli G and X.

70 80 90 100 110 120 130 14-0 150 160 170 180 190 200 210 220 230 240


FIGURE 3. Length-depth relationship of Sebosfoc/es mystinus



After the first annulus has been completed, there is a tendency for the
succeeding* annuli to occur as pairs of rings. The first member of each
pair is the more constant ( Table 1 ) , and the more conspicuous laterally.
In general, the two rings of late pairs are more widely separated than
those formed earlier. Indeed, in classes 4+, 5+, and 6-\- it is often
impossible to tell whether two successive rings are members of a pair or
true annuli of different years. I cannot explain why there are two rings
in some annuli and but one in others. As a matter of fact, the cause or
causes of annulus formation in mystimis admittedly are unknown.

Occurrence of Scale Rings in tlie Year Classes

Year class


1 +


2 +















F _ ...




The first member of a pair is considered as a " true ' ' annulus because
of its more frequent occurrence. The others are termed "secondary"
because of their irregularity. Where A or B alone was present it was
impossible to tell them apart.

It was noted earlier in this discussion that there are other variations
in the regular sequence of circuli besides the annuli. These irregularities
may be in the form of bands (B in Figure 7) almost indistinguishable
from true annuli except that they seldom have side checks ; or they may
occur as abrupt changes from areas of closely set circuli to areas of
widely placed circuli.

No difference was found between scales of males and females.

The photographs are of selected scales and therefore give a somewhat
idealized picture. However, they do not present any characters not found
in the majority of these structures.

The fact that the various rings were designated by letters suggests that
they could always be homologized with certainty, but this was not en-
tirely the case. Thus in class 2-{- it was very questionable whether ring
A was the homologue of A, B, or C in class 1+. However, there w^as
seldom any doubt about the years in which the annuli were formed, thus
making it possible to age fish of the first four classes with very little
hesitation, once the positions of the annuli were well established.

This assurance in the recognition of classes 0+, 1+, 2-|-, and 3+ does
not apply to the older groups. There are a number of reasons for this.
One has already been given, i.e., the difficulty in telling whether two
successive rings represent two annuli or whether they are simply members
of a paired annulus. Another reason is that beyond class 3-|- the annuli


often fail to appear altogether, or they may be so poorly defined that
their presence becomes a matter of conjecture.

Soon after the writer began the study of mystinus scales, he became
aware that the establishment of similarities rather than difference was
the first and most important thing. Every scale seemed to be different.
Then slowly, one by one, points in common became apparent ; and as the
writer grew more and more familiar with the set of characters indicative
of the various classes, it became easier to sort the scales, even though many
cases always remained highly perplexing.

Some work was done in the matter of measuring scales from the nucleus
to the various annuli. However, there was so much variation in the size
of the scales that the results were nearly useless, even when a correction
was made by plotting an average scale length — body length curve and
then interpolating from this in the equation :

distance to margin average distance to margin

distance to check X

The scale's length was measured from the center of the nucleus to the
middle of the anterior margin.

After having classified the fish into preliminary age groups, the next
step taken was to determine the age of the smallest class, the value thus
established to be used as a basis of age computation.

Of the specimens taken during June through September, 1930, the
great majority formed a very pronounced size and scale group, the mode
of the length frequency curve being located at 135 mm. (5^% inches)
and the extremes at 113 mm. and 148 mm. Below 113 mm. there was a very
decided gap in the frequency, the next largest specimen being only
76 mm. This fish according to the scales evidently belonged to another
group, of which only six fish were taken, the smallest being 68 mm. These
six fish were taken in July, 1929, and others presumably of this class
were observed in tidepools in ^lay, 1931. In other words, this smaller
class occurred at the same time as the larger. The difference in age is
almost certainly one year, since this species is known to spawn but once
a year and the spawning season is quite short. Absolute proof that their
age difference is one year is lacking, however. According to actual obser-
vations the larger group increased 40 mm. in one year. By subtracting
this length from the moclal length of 135 mm., we arrive at a modal length
of 95 mm. for the smaller group, thus decidedly higher than the one actu-
allv observed. Tn regard to the number of circuli on the scales, the smaller
class exhibited an average of 24. the larger, of 80. As a matter of com-
parison, it may be noted that the larger group added 67 circuli in one
year, an unexpectedly large addition when the difference between these
two length classes is taken into account.

As a matter of convenience the smaller group was called class O-j- ;
the larger one, class l-j-.

In regard to the age of class 0+, we have no direct evidence. Assuming
that the growth of this class was greatest during the early part of its
development, we reached the conclusion that it was derived from the
previous spawning season. In other words, this class would be slightly
more than six months old.

Besides the two classes just considered there were two others which
could easilv be distinguished on the basis of scale and size differences.



In June, 1930, the average lengths of these classes were 175 mm. (6|-
inches) and 200 mm. (l-%2 inches), respectively. The nnmber of scale
rings in the first class was 95. while that in the larger class was 150.
Undoubtedly they were older than classes O-f- and 1^. but just how
much ■? It has already been said that the methods of sampling for a length
frequency analysis were highly selective ; but even though the various
age groups were not sampled in a manner which would show their relative
abundance, it is the writer's belief that representatives of the first four
year classes were collected. From the length modes, such as they are.
and from a consideration of the scales this belief seems correct beyond
doubt. As no sexual differences were found in the scales and as there
seems to be just one breeding season a year, we are forced to call these
year classes 2+ and 3+, in other words to consider them as one and two
years older than class 1+. It has been said previously that age determina-
tion at the present time cannot be carried beyond class 3+ because of
lack of data and the increasingly numerous scale difficulties.

Figure 4 shows the dominant class of the 1929 collecting season falling
midway between class 2-\- and class 3+ of the 1930 season. Upon exam-
ination of the scales it is seen that this class is also age 2+, but it has
grown faster than class 2-1- of 1930.





u 21



cr I 8


< 12

AGE 0+


— ^r - if'^'-

60 70 80 90 100 MO 120 130 140 150 160


190 200 210 220

FIGURE 4. Length frequencies of the blue rockfish caught during June through September, 1929
(broken line) and during 1 930 (solid line). Smoothed by a moving average of threes.

Figure 5 shows a steadv increase in circuli in class 14-. but the length
curve is very peculiar, doubtlessly because of the small number of fish
measured. It might be asked why the circulus curve is so much better than
that of the length measurements. This may be explained if we take for
illustration two fish of class l-j-. Fish A is 170 mm. long while B is 185


mm. In six months A has grown to a length of 185 mm. and has added
25 circuli while B has attained the length of 200 mm. and also has added
25 circuli. If small samples were made at these times they might contain
fish nearlj^ all of which were like A or nearly all like B. In either case the
circulus averages would be similar but length averages would be much





50 o


30 c




FIGURE 5. Increase in length of fish and in number of scale circuli for class 1 + during 1930-31

When regular sampling began in June, 1930, class 1+ was the young-
est to be found (Figure 4). Class 1+ continued in the same relative
abundance for the ensuing 13 months. In April, 1931, fish which were a
year younger than this dominant class made their appearance in the
area but in very small numbers. This group had very probably been
present all of the time but the fish were too small to be caught. It is re-
markable how much less numerous this class was than the class of the
same age a year before. From this fact, it may be concluded that there is
great variation in the mortality of the different year classes.

When the writer began examination of the scales of the blue rockfish
in June, 1930, ''annulus" C of class 1+ was forming or had recently
been completed. During the ensuing year, circuli were apparently added
in a very regular manner. However, at the end of the collecting in June,
1931, the scales of class 1+ taken in May and June were examined criti-
cally for irregularities, and although the irregularities found were very
poorly defined, the examination shoAved that 31 percent had previously
formed ' * annuli. ' '

A little more than half of these annuli were of the single type, the rest
were paired. The average number of circuli added since the completion
of these annuli was seven, indicating that the annulus forming period
for the fish of class 1-|- which had annuli present was in March, April,
and May. Contrasted with the year during which most of the fish failed
to form annuli, we have the previous year and a half's growth of the
same group of fish, when there were typically three "annuli" (Figure
8) . Looking at the other scale photographs we can establish that the scale
configurations of each year class are so unlike that we must conclude that


each year class has its own characteristic peculiarities of growth, but
only in a very general way can it be said that an ' ' annulus ' ' is added once
a year and that by counting the annuli can the age of the fish be deter-
mined. Figure 8 is an exceptional case, since it appears that each of the
heavy bands is a true annulus.

This situation has alreadv been considered, and it need scarcelv be said
again that we believe the irregularities in the environment of this species
produce very undependable scale features for use in age determination.


Realizing that the scale features in this species are somewhat confus-
ing and that an understanding of age determination in the blue rockfish
may assist in "aging" other species of this genus, it appears desirable
to describe the scale photographs used in this paper more fully.

After a thorough study of the scales collected from all of the age groups
found off the Hopkins Marine Station in 1929-30 it was apparent that
in most cases fish of the same year class had similar scale characteristics,
so that the observer could group his scales. However, it was not always
possible to assign an age to these groups. Some were characterized by
scales which seemed to have annular rings, but other groups did not.

In Figure 6 (1928 class 0-|-) the first annular ring or band started to
form about twelve circuli out from the nucleus. The corresponding an-
nulus A can be seen in Figure 7. Also in Figure 7 can be seen a "secon-
dary" ring B, the cause of which is unknown. Frequently this so-called
secondary ring is not present. In Table 1 it will be noted that in the 1927
year class this secondary ring B was present in 67 percent of the scales
examined, while in the 1926 and 1925 classes this ring was altogether
absent. In general it was found that when one member of such a pair was
absent it was the second one formed. Therefore, I have referred to the
first ring of a pair as the true annulus.

Looking again at Figure 7 it can be seen that both rings of the pair
A, B were formed in the first year following the birth of the fish. The
Ring C was formed in the second year. It was not followed by a secondary

Now let us examine Figure 8. This is a scale from a 1926 class 2+ fish.
In that year 94 percent of the fish showed the true annulus A but not
the secondary ring B. Only 25 percent showed an annulus in the second
year C. In its third year this class developed a pair of rings in most
instances. Figure 8 shows the first of the pair D and the secondary ring E
can be seen in Figure 9.

Referring to Figure 9, we see a scale from a 1925 class 3+ fish. Nat-
urally this is a far better than average scale, selected to show all of the
characters useful in aging this class. Ring A was the first of the pair com-
monly formed in the first year. Ring C was the annulus formed by some
fish in the second year. Rings D and E make the pair formed in the third
year (2-|-). Of this pair ring D was formed in 100 percent of the scales,
while E was formed in only 17 percent of the scales. Ring F was the first
or true annulus formed in 69 percent of the scales.

Figure 10 is given to show the growth between time of tagging and




FIG. 9

FIG. 10



Figure 11 shows a scale from a 1925 class 5+ taken in February, 1931.
This is an exceptional scale. Rarely are the annuli shown this clearly.
However, even in this scale the annulus for the first year is not present.
However, a single broad annulus was formed in each succeeding year.
It is only in rare cases that blue rockfish of this age show the annuli as
clearly as does this particular specimen and even here the age determi-
nation would have been a vear less than it should have been if all the
younger age classes had not been studied. In conclusion, we may say that
age determination from scale reading is not impossible in this species
but that it is very difficult and must be accompanied by a length-fre-
quency analysis.


Some three-year-old males produce sperm, and it is probable that

1 2 4 6 7 8 9 10 11 12 13 14 15