California. Dept. of Fish and Game.

California fish and game (Volume 54, no. 4) online

. (page 1 of 10)
Online LibraryCalifornia. Dept. of Fish and GameCalifornia fish and game (Volume 54, no. 4) → online text (page 1 of 10)
Font size
QR-code for this ebook


QMJFDRNIA
FISH-GAME




California Fish and Game Is a journal devoted to the conser-
vation of wildlife. If its contents are reproduced elsev/here, the
authors and the California Department of Fish and Game would
appreciate being acknowledged.

The free mailing list is limited by budgetary considerations
to persons who can make professional use of the material and
to libraries, scientific institutions, and conservation agencies. In-
dividuals must state their affiliation and position when submitting
their applications. Subscriptions must be renewed annually by
returning the postcard enclosed with each October issue. Sub-
scribers are asked to report changes in address without delay.

Please direct correspondence, except regarding paid subscrip-
tions, to:

LEO SHAPOVALOV, Editor
California Fish and Game
1416 9th Street
Sacramento, California 95814

Individuals and organizations who do not qualify for the free
mailing list may subscribe at a rote of $2 per year or obtain
individual issues for $0.75 per copy by placing their orders with
the Office of Procurement, Documents Section, P.O. Box 20191,
Sacramento, California 95820. Money orders or checks should
be made out to Office of Procurement, Documents Section. In-
quiries regarding paid subscriptions should be directed to the
Office of Procurement.



u







D



V



VOLUME 54



OCTOBER 1968



NUMBER 4




Published Quarterly by

STATE OF CALIFORNIA

THE RESOURCES AGENCY

DEPARTMENT OF FISH AND GAME



STATE OF CALIFORNIA

RONALD REAGAN, Governor



THE RESOURCES AGENCY

NORMAN B. LIVERMORE, JR., Adminhtrator



FISH AND GAME COMMISSION

WILLIAM P. ELSER, President, San Diego

JAMES Y. CAMP, Vice President C. RANSOM PEARMAN, Member

Los Angeles Huntington Pork

HENRY CLINESCHMIDT, Member SHERMAN CHICKERING, Member

Redding San Francisco



DEPARTMENT OF FISH AND GAME

WALTER T. SHANNON, Director

1416 9th Street
Sacramento 95814



CALIFORNIA FISH AND GAME

Editorial Staff

LEO SHAPOVALOV, Editor-in-Chief Sacramento

PAUL M. HUBBELL, Editor for Inland Fisheries Sacramento

CAROL M. FERREL, Editor for Wildlife Sacramento

HERBERT W. FREY, Editor for Marine Resources Terminal Island

DONALD H. FRY, JR., Editor for Salmon and Steelhead - Sacramento



CONTENTS

Page

Mortality Rates in the California Striped Bass Population

Harold K. Chadwick 228

Some Encloparasites of Fishes From the Sacramento-San Joaquin

Delta, California Stephen B. Edwards and F. M. Nahhas 247

A Probability Sea Survey Plan for Estimating Relative Abun-
dance of Ocean Shrimp Norman J. Ahramson 257

First Report of the Crab Family Chirostylidae Off California, and

Description of a New Species of Chirostylus Janet Haig 270

Prionotus xenisma Jordan and Bollman, a Searobin New to Cali-
fornia BicJiard S. Lee 278

Fecundity of the Northern Anchovy, Engraulis mordax Girard

John S. MacQregor 281

Summer AVater Requirements of Desert Bighorn in the Santa
Rosa Mountains, California, in 1965

Bonnor Blong and William Pollard 289

Third Cooperative Survey of the California Condor

Fred C. Sibley, Bolert D. Mallctte, John C. Borneynan, and

Baymo7id S. Dalen 297

Notes

Harbor Seal Censuses in Humboldt Bay During 1966 and 1967

Bichard J. Bosenthal 301

Deformed Lateral Line in a Jack Mackerel, Trachurus sym-
metricus (Ayres) John M. Duffy 306

Observation of Striped Bass Spawning in the Sacramento River

Lee W. Miller and Bohert J. McKechnie 306

Book Beviews 308

Index to Volume 54 315



(227)



Calif. Fish and Game, 54(4) : 122.S-24G. 1968.



MORTALITY RATES IN THE CALIFORNIA STRIPED

BASS POPULATION^

HAROLD K. CHADWICK

Inland Fisheries Branch ^

California Department of Fish and Game

Mortality rates for the Sacrcsmento-San Joaquin River system striped
bass (Roecus saxatilis) population were calculated from tag returns. An-
nual variations and variations among different parts of the population
were examined.

INTRODUCTION

A gradual decrease iu sport fishing success for striped bass during
the late lOiO's and early 1950's (Chadwick, 1962) caused considerable
concern over the population's status. As part of a program to evaluate
changes in the population and regulations, a 4-year tagging program
was started in 1958 to measure mortality rates. This paper describes
and evaluates these mortality rates. Angling regulations will Ik^ ovnl-
uated in light of these estimates in a subsequent pa])er.

An earlier tagging study (Calhoun, 1952) did not yield satisfactory
mortality estimates, largely because of the loss of tagged bass in com-
mercial salmon and shad gill nets. These commercial fisheries were
closed in 1957, and bass in the present study were not affected signifi-
cantly by any commercial fishery.

This study involves only California's main striped bass population,
which inhabits the Sacramento-San Joaquin river system. Adults
spawn in the Saeramento-San Joaquin Delta and its main tributaries
in the spring, migrate downstream to brackish and saltwater bays and
the Pacific Ocean immediately aftci-wards, and return to the Delta
either in the fall or the following spring (Calhoun, 1952; Cliadwick,
19G7).

METHODS

Most striped bass tagged in this study'' were caught during April,
May, and early June in drift gill nets fished at several locations near
the junction of the Sacramento and San Joaquin Eivers (A through
E, Figure 1). These tagging areas are referred to as the western
Delta throughout this paper. Small groups of bass were also tagged
there in the falls of 1957 and 1958. Additional fish were caught for
tagging in wire fyke nets (Hallock et ah. 1957) at Fremont Weir on
the upper Sacramento Eiver (F, Figure 1) in IMay 1958. in drift gill
nets at Prisoners Point in the eastern San Joaquin Delta (G, Figure 1)

1 Accepted for piiblicatioti IMay 196S. This study was performed as part of Dinsell-

Johnson Project California F-O-R, "A Study of Sturgeon and Striped Bass", sup-
ported by Federal Aid to Fish Restoration Funds.

2 Now with Delta Fish and Wildlife Protection Study.

3 See Chadwick (1960, 1963, and 1967) for more detailed descriptions of fishing gear,

fishing methods, tags, and tagging methods.



( 22S)



STRIPED BASS MORTALITY RATES



229




FIGURE 1— Map of study area showing tagging sites. A, Sacramento River at Chipps Island;

B, Sacramento River at Chain Island; C, Sacramento River at Sherman Island; D, Broad

Slough; E, San Joaquin River at False River; F, Sacramento River at Fremont Weir; G, San

Joaquin River at Prisoners Point; H, San Pablo Bay.



in the spring of 1959, and by angling in San Pablo Bay (H, Figure 1)
in the falls of 1958 and 1959.

Bass were tagged with several kinds of tags, but most were tagged
with disk dangler tags. This tag proved to be superior or equal to the



230 CALU'UKXIA Fl.SlI AND GAME

other types (ChadAviek. 1063), and was the only tj'-pe used in esti-
matinj; mortality rates. Disk dan<rler tajrs are modified Atkins tajrs
(Calhoun, lf)5;i. Fijj:ure 3). In this study, these tags consisted of a cel-
lulose nitrate disk attached with pure tantalum or soft stainless steel
wire through the back below the first dorsal tin (Chadwick. 1003).

Some of the disk dangler tags had the wording "$o REWARD"
printed on the face of the tag. In 1058 this wording was red. Avliile
the remainder of the printing was black and the tags were white. In
subsequent years, some of the reward tags were orange witli bhick
letters. Nonreward tags were white with black letters.

In 1058, reward tags were applied in a ratio of 1:2 witli ;i 'jnni]) of
nonreward tags. In subsequent years the ratio was 1:4. In ijhIi case,
the two types of tags were applied systematically, so the two groups
were comparable. For example, in 1058 one fish received a reward tag.
the next two received nonreward tags, the next a reward tag, etc., until
the number quota was reached.

In addition to the type of tag, the date, the fork length to the near-
est inch, the tagger, the type of fishing gear, and the fish's condition
were recorded for each tagged fish. Fish which swam away immediately
upon release were considered to be in good condition. Those which
floated at the surface for a period ranging from a few seconds to sev-
eral minutes after release were considered to be in poor condition. Fish
whicli bled profusely from the tagging wound were ])laced in a third
condition category. This bleeding was caused by puncturing blood ves-
sels in the vicinity of the neural spines.

Since bass were tagged over approximately 2 months each spring
while sportsmen were fishing, the year of recovery for each tag was
determined by the number of elapsed days between tagging and re-
capture. When anglers did not report date of recapture, a postcard
questionnaire was sent to try to determine this date. In the few cases
when this failed, the date of recapture was assumed to be the post-
mark date on the angler's letter. Tags with to 365 elapsed days were
recorded as first-year recoveries, those with 366 to 730 days as second-
year recoveries, etc. Thus, recovery years do not pertain to any precise
calendar ])eriod, but generally run from spring to spring.

Designating year of recovery in this manner could affect mortality
estimates significantly for a fishery that is highly seasonal. However,
in this case relatively few bass were recaptured while tagging was in
progress. In 1061, 8.2% of the first-year recoveries were made while
tagging was in progress, while in other years this percentage ranged
from 2 to 3.5%.

Also, an analysis later in this paper (see Table 6) shows that per-
centage returns are not related to time of tagging. Hence, apportion-
ing recoveries in this manner has no significant effect on mortality
estimates.

In general mortality rates were estimated following Ricker (1058),
but several modifications were made.

The proportion of all recaptured tags which were returned was es-
timated from the formula :



STRIPED BASS MORTALITY RATES 231

" = M^' Ml



A



where :

r = proportion of recaptured tags returned,
M ' = number of reward tags applied,

R' = number of tags returned during first year from group M',
M" = number of nonreward tags applied to a group of fish com-
parable to group receiving reward tags,
R" = number of tags returned during first year from group M".

The variance of this estimate of response was estimated from the
formula :

V(r) = {^y (^ + ^) [2]

where the new symbols are :

A

V(r) = estimated variance of response,
P" = proportion of tags M" returned during first year,
Q" = proportion of tags M" not returned during first year,
and P' and Q' are comparable proportions for tags M'.

This formula for variance is adapted from formulas 10.4 and 18.1
(Hansen, Hurwitz, and Madow, 1953, Chapter 4). The proportion of
nonresponse is equal to 1 — r, and the variance of nonresponse is equal
to the variance of response. Confidence limits were estimated assum-
ing a normal distribution.

Annual response estimates were weighted by the reciprocal of their
variances when estimating the mean response for the four years ac-
cording to the formula:



4

s

i=l


A

ri


V(r)i


4

i=l


1


A

V(r)i



r = '-^^ [3



where the subscript i denotes time in years and ranges from to t. In
all cases where i is used to designate time for a process occurring over a
period of time, the time it designates is from i — 1 to i.



232 CALIFORNIA FISH AND GAME

Rates of exploitation were llicii estimated using the fonnma:

(i'i = -^<1=1^ [4]

where new symbols are:

u'i = rale of exploitation from i — 1 to i.

Mi = number of nonreward tags applied at i.

Rij = numbcn- of nonreward tags applied at time i and recaptured

during time j, with j ranging iVom 1 to t.

The varianee of the proportion of tags returned was calculated from
the formula:

where ui is the proportion of M(i _ d tags actually returned from i — 1
to i.

The approximate formula for the variance of tlie rate of expknlation
was developed following Lindley (19G5, p. 134-139). The formula is:

n A

, 1 ,2 A /,V-\ 2 V(r)i

V(u'0 ^:(^^)V(U0 +r' '^^^^^— [6]



Confidence limits were calculated assuming a normal distribution.

Returns indicated that survival was not constant so survival and its
variance in 1958, 1959, and lOdO were estimated following Jolly (19G5),
using the following formulas:

A

yi Ao+i+...+(.i-i))

= M- (-^0(('+l) + ) + ^^^l((i+l) + ) + • • • + R(i-l)((i+l) + )) r_,

Ri((i+i)+)



STRIPED BASS MORTAIiTTY RATES



233



A



A

Si =



]\J.(0+l+...+(i-l))



A



(M(i_;jj^<'+i+-+('-i) > + M(i_i))



V(si) = Si'^



^1.(0+ . . . +(i-i) )(Aii{o+ . . . +(i-i)4- Mi)



A



(Mi



(o+...+(i-l))^2



y



[8]



(—1 J-)

VI^i((i+i)+) ^li /



+ M(i_i)(»+-4>^^^^ / 1 1__\

A -^.-/j \ R(i_i)i+ jM(i_i)/



A

(1 - Si)



+ -

]\J.(0+...+(i-l))

where the new symbols are



^VvO:^



[9]



' \o



ny\



L



]y[.(o+i+ . . . +(i-i) ) = Estimate of number of tagged fish sm-vivmg

from releases at times 0, 1, . . . , i — 1
to time i.
Rij+ = number of bass tagged at time i which are
recaptured in year j and all subsequent
years.

Si = estimate of probability of survival from time
i — 1 to time i.

These formulas are analogous to Jolly's formulas 22, 24, and 27,
respectively. Tliej- differ in not having the number of tagged fish
recovered (Jolly's mi) subtracted from estimates of the numbers of
tagged fish alive, and in having the notation changed to correspond
more closely to Kicker's notation. The numbers of tagged fish recov-
ered were not subtracted because tag recoveries were not made coinci-
dent with subsequent tagging operations.

Since both natural mortality and fishing mortality varied, precise
estimates of survival and natural mortality could not be made after
the third year and fishing mortality could not be estimated ])recisely
after the fourth year. Mortality rates through the spring of 1965 M^ere
approximated by:



234 CALIFORNIA FISH AND GAME

1) Estimating survival for 1961 by dividing the estimated total
recaptured in 1962 by the estimated total in 1961. This was
multiplied by the number of fish estimated to be alive at the be-

A

ginning of 1961 (M3^°+^+^^ + M3) to estimate the number alive
at the beginning of 1962.

2) Estimating the number cauglit during 1962 by dividing the num-
ber of tags returned in 1962 by the mean response estimate (equa-
tion 3). Then estimating the rate of exploitation during 1962 by
dividing the estimated number caught during the year by the
number estimated to be alive at the beginning of the year.

3) Averaging expectations of deaths from natural causes for 1958
through 1961, and assuming that this remains constant from
1962 on.

4) Subtracting the estimated rate of exploitation plus the mean
expectation of deaths from natural causes from 1 to estimate
survival rate, and multiplying this by the number alive at the
beginning of the year to estimate the number surviving at the
end of the year.

5) Repeating steps 2 and 4 for each subsequent year.

VALIDITY OF ESTIMATES

Mortality estimates based on tag returns are accurate only if the
tags do not cause mortality, do not affect vulnerability to fisliin<r.
are not shod, are returned when the fish are caught, and are randomly
distributed in the population. Study results do not permit a precise
evaluation of all of these, but pertinent findings will be reviewed to
give the reader an understanding of the estimates' limitations.

Effects on Mortality

The gill nets were not drifted longer than 20 minutes and fish obvi-
ously in poor condition were not tagged, to minimize immediate mor-
tality from handling and tagging. Nevertheless some fish died sliortly
afterward, presumabl.v as a direct result of handling and lagging.
Each year anglers and the tagging crew picked up a few dead bass
wdthin a week after tagging. For example, in 1958, 8 of 4.386 tagged
bass were found dead. This is the only direct evidence of mortality.

The mean blood lactic acid level * in 6 bass cauglit in the gill net
in June 1959 which were in good condition was 630 mg/1. In contrast,
the mean level in 12 bass caught at the same time, but which were
in poor condition, w^ts 1,170 mg/1. Only one fish in the first group
fell within the range of the latter group. The mean level in 10 bass
caught by angling in October 1959 was 230 mg/1. These levels indicate
that the arbitrary classification by condition reflects real physiological
differences, and that even fisli in good condition when removed from

* Determined bv meUiod of Barker and Summerson (Hawk, Oser, and Summerson,
1949).



STRIPED BASS MORTALITY RATES



235



the net have increased blood lactic acid, presumably as a result of
fatigue.

Returns from fish in different conditions at release provide indirect
evidence of mortality (Table 1). In 1958 and 1960 returns differed
significantly among the three groups (x" = 8.66, P = 0.014, and
y- = 6.69, P = 0.038, respectively). This was caused primarily by low
returns from fish in poor condition. This did not occur in 1959 and
1961, and returns did not differ significantly among the three groups
in those years.

TABLE 1

Relationship Between Condition at Tagging of Striped Bass
and Percentage Returns *





Returns from each year's tags


Condition


1958


1959


1960


1961


Good.


23.2
(3653)

16.2
(321)

21.0

(238)


14.9

(2870)

15.6
(308)

22.1

(95)


14.4

(3262)

8.8
(205)

18.8
(96)


11.4


Poor.


(1565)
12.9


Bleeding from tagging wound


(85)

18.2
(44)



* Includes only first year returns from nonreward tags on striped bass tagged in the
western Delta during the spring. Numbers in parentheses are numbers of tagged
bass.

The low 1958 and 1960 returns from fish in poor condition were
presumably caused by mortality resulting from tagging and handling.
I have no explanation for the mortality being inconsistent. Because
of this evidence, fish in poor condition when tagged were eliminated
from all mortality estimates.

Bleeding at the time of tagging obviously did not increase mortality.

Bass tagged with disk dangler tags generally grew more slowly than
untagged bass (Chadwick, 1963), but this would not necessarily have
caused greater mortality.

Disk dangler tags sometimes caused considerable irritation, often as
the result of hydroids attached to the tag (Chadwick, 1963). There is
no w^ay of knowing whether this caused mortality, but fish were seen
with large scars near the tag, indicating recovery from considerable
tissue damage. Observations indicate that tissue damage from the tags
was greatest in the first j^ear after tagging.

Vulnerability to Fishing

Tagging and handling reduced the vulnerability of bass to angling
for about a month after tagging (ChadAvick, 1963). The monthly ratios
of returns from bass tagged in one year to those tagged tlie previous
year during 4 months after the later group was tagged indicate that
vulnerability was reduced by about a third to a half during the first
month after tagging.



236



CALIFORNIA FISH AND GAl^IE



To estimate tlie iiiiixiinum error from tliis source, vulnerability dur-
ing tlie first month Avas assumed to have been reduced by lialf. To
approximate tlie v;ff^< cts of this tlie number of tags received in tlie
first 30 days after tagging (Chadwick, 1963. Table 11) was doubhul.
This Avould have increased angling exploitation rates in 1958, 1959.
1960, and liKil hy about 3. 4, 4. and 8%, respectively. Th(> actual
increase "would have been less because some of the fish Avhicli had been
saved by the lower vulnerability would have been caught later in the
year, and because this assumes the maximum observed error. Hence,
errors fi-om this source are small.

Tag Shedding

"While disk dangler tags have proved equal or su])prior to several
other types of tags tested, a few have been shed under various cir^
cumstances (Chadwick, 1963).

Kesults from the earlier years of the study suggested that apparent
survival was greatest for the most recently tagged fish, indicating that
either shedding rate increases Avitli time after tagging or mortality
is greater for older fish (Chadwick. 1963). However, more recent re-
sults (Table 2) suggest that sur\ival is not related to Icngtli of time



TABLE 2

Ratios of Tag Returns in Successive Years From Tags
Applied in Different Years *





Year tagged


Years for which r.itios
were calculated


1958


1959


1960


1961


1960/19.59

1961/1960. . ..-


0..50 (103)
0.50 (52)
0..52 (27)
0.52 (14)
0.86 (12)
0.58 (7)
1.14 (8)


0.59 (266)
0.42 (111)
0.80 (89)
0.69 (61)
0.33 (20)
0.60 (12)
1.42 (17)


0.52 (261)
0.61 (1.58)
0.92 (145)
0.41 (.59)
0..59 (35)
0.95 (33)




1962/1961

1963/1962


0.70 (131)
0.89 (117)


1964/1963


0.45 (53)


1965/1964.

1966/1965 . ...


0.40 (21)
1 . 19 (25)







* Includes only nonreward tags from fi.^h in good condition ■which were tag-g-ed in the
western Delta during the spring. Numbers in parentheses are the number of tags
returned in the second year on which the ratio is based. No corrections were
made for variable exploitation rates, so ratios are biased estimates of survival.

tags have been out. Estimates are quite variable each calendar year,
but there is no consistent trend of decreasing survival with increased
time at large.

This finding and the fact that returns have been received through
8 years, indicate that shedding afifects mortality estimates insignifi-
cantly.

Reporting of Tagged Fish

The striped l);iss fishery is so Midespreiid that there is no })ractical
way of assuring that anglers return all tags from fish they catch.
Eeturns from the $5 reward tags were compared with returns from
comparable groups of nonreward tags to estimate the proportion of
tags which were not returned.



STRIPED BASS MORTALITY RATES



237



This assumes tliat anglers return all $5 reward tags. No check was
made of this assumption. Undoubtedly some were not returned, caus-
ing nonresponse to be underestimated.

First-year returns of reward tags were significantly greater than
the return of nonreward tags in all 4 years. However, except for the
1961 tags, diiferences were much smaller or nonexistent in subsequent
years (Table 3).

TABLE 3





Comparisor


of Returns From Nonreward Tag
Colors of Reward Tags


s and Different






Tag
description


Number
tagged


Percentage returned each year


Total


Year
tagged


1


2


3


4


5


percentage
returned


1958—


Nonreward

White reward


299
150


22.4
40.0


7.0

6.7


3.3
2.7


0.0
1.3


0.3
1.3


33.1
52.0


1959„.


Nonreward

Wliite reward

Orange reward


1006
125
125


13.8
2.3.2
20.0


9.7

11.2

7.2


3.7
2.4
3.2


2.9
0.8

4.8


2.0
1.6
4.0


32.2
39.2
39.2


1960...


Nonreward

White reward

Orange reward


1587
199
201


14.2
22.6
17.9


7.6

9.5

13.4


4.4
7.0
5.0


4.0
3.5
2.5


1.8
0.5
2.0


31.9
43.2
40.8


1961...


Nonreward

Orange reward


995

248


10.6
17.3


9.2
14.5


7.6
10.9


2.7
5.6




30.2

48.4



One factor undoubtedly contributing to this is that the red printing
on the white reward tags faded. It sometimes became illegible after
about a year, making the tags in effect nonreward tags. This led to the
decision to use the orange reward tags. However, returns from these
were generally lower than returns from the white tags, although dif-
ferences were not significant (x^ for 1960 first-year returns ^= 1.37,
P = 0.24).

Colors have been shown to influence tag returns (Lawler and Smith,
1963), and red tags have stimulated an antagonistic resiDonse in trout
(German and LaFaunce, 1955). Thus, the orange tags may have in-
fluenced returns by some mechanism such as making bass more vul-
nerable to predation by sea lions.

While the cause for the smaller difference after the first year cannot
be determined, it is not reasonable to attribute it to a true difference
in nonresponse. Hence, only first-year returns were used in calculating
nonresponse. Whatever caused the decreased nonresponse after the
first year may well have commenced during the first year, causing non-
response to be underestimated.

Calculated nonresponse varied considerably from year to year (Table
4). Since migration patterns varied annually (Chadwick, 1967), rela-
tionships between location of capture and nonresponse might explain
this variation. Nonresponse tended to be lowest in San Pablo Bay and
highest in the Pacific Ocean (Table 5). Tliis appears reasonable since


1 3 4 5 6 7 8 9 10

Online LibraryCalifornia. Dept. of Fish and GameCalifornia fish and game (Volume 54, no. 4) → online text (page 1 of 10)