9695

• UCREFRP Catalog Number: 9695
• Author: Gold, J.R., L. Ma, E. Saillant, P.S. Silva and R.R. Vega.
• Title: Genetic effective size in populations of hatchery-raised red drum released for stock enhancement.
• USFW Year: 2008.
• USFW - Doc Type: Transactions of the American Fisheries Society
• Copyright Material: YES

1332

mately 34% less than the expected maximum N of
4.55 had all possible mating combinations occurred at
each spawn; and third, that the varying number of
progeny generated per mating combination further
reduced the average Ne per spawn by about 9% (to
2.59). Overall, the average Ne per spawn was
approximately 43% less than the maximum possible,
with nearly 80% of the reduction being due to the
number of mating combinations that actually occurred.
Because the latter represent the number of mating (dam
X sire) combinations irrespective of sex, the observed
reduction in Ne per spawn appears to be due primarily
to nonspawning dams. Finally, using simulation
analysis and TPWD spawning and release records in
2003, the average number of spawns from different
spawning tanks mixed in a released population was
estimated to be 2.84.
Spawning and release records at the TPWD MDC
indicated that in 2003 a total of 62 release populations,
roughly equivalent in size, were stocked into different
localities in each of four different bays or estuaries. The
number of released populations per bay or estuary
ranged from 7 (Aransas Bay) to 27 (upper Laguna
Madre) and included offspring from 11 (Aransas Bay)
to 18 (upper Laguna Madre) different spawning tanks.
Considering the average maximum Ne of 4.55 for each
spawn, the estimated maximum average NeR of all
released fish per bay of estuary potentially would range
from about 50.1 (Aransas Bay) to about 81.9 (upper
Laguna Madre). These estimates assume that the
contribution of individual spawning tanks and survival
probabilities per released population were equivalent.
Considering the average Ne per spawning tank of 2.59
when accounting for variation in the number of
progeny generated per observed mating combination,
the average NeR per bay or estuary would be reduced to
approximately 28.5 (Aransas Bay) and 46.6 (upper
Laguna Madre). These values of NeR would be
underestimates if the contribution of progeny of
individual dams, sires, or both tended to equalize over
time; the values would be overestimates if the
contribution of different spawning tanks and survival
probabilities per released population varied. Based on
present data, the latter seems more likely than the
former.
Based on a coalescent approach, Turner et al. (2002)
estimated the long-term genetic effective size (Net) of
wild red drum populations in each of seven bays or
estuaries in the northern Gulf of Mexico. Their
estimates ranged from 183 to 517 (average, approxi-
mately 263 per bay or estuary). We used the data from
the study of Turner et al. (2002) and estimated the
contemporaneous variance of the genetic effective size
(NeV)' using the temporal method (Nei and Tajima

GOLD ET AL.

1981; Waples 1989) for the same seven bays of
estuaries. The estimates of NeV per bay or estuary
ranged from 166 to 356 and averaged 272. The range
of NeR estimates (from 28.5 to 46.6) per bay or estuary
for all fingerlings released from the TPWD MDC in
2003 are smaller, on average, than both the Nel and NeV
estimates per bay or estuary. Moreover, the estimates
of NeR are very likely inflated given that the
assumption of equivalent survival probability per
released population is probably not met. Briefly,
Karlsson et al. (2008) genotyped yearling or older red
drum from two "stock-enhanced" bays or estuaries
along the Texas coast and unequivocally identified 30
of 321 fish (9.3%) sampled from Aransas Bay as being
of hatchery origin. The contribution of brood dams,
sires, and dam X sire combinations to the hatchery-
assigned fish, however, was nonrandom, as was the
distribution of hatchery-assigned and wild fish with
respect to sampling localities within each bay. Karlsson
et al. (2008) interpreted these results as indicating
variation in survival probability among releases, which
clearly would result in decreases in total NeR over all
releases. The above considerations suggest that a
Ryman-Laikre effect could occur in Texas bays or
estuaries supplemented with hatchery-reared fish.
There are three approaches that might be employed
to increase the NeR of TPWD-released fish and
decrease the probability of a Ryman-Laikre effect:
(1) increase the number of mating combinations per
spawn, (2) equalize the number of progeny generated
per mating combination, and (3) increase the number of
spawns from different spawning tanks in each released
population. The first could be accomplished by using
two dams and three sires in each spawning tank, given
that the proportion of spawning sires appears greater
than the proportion of spawning dams. This approach
might be problematic, however, as total egg output per
spawning tank could be compromised significantly,
particularly as far fewer dams than sires appear to
participate in individual spawns. The observed spawn-
ing activity of dams and sires also raises the question as
to whether over a spawning season there are dams (or
sires) that contribute few or no progeny to any released
population. We are currently studying this issue and, to
date, it appears that there are dams (but not sires) in
TPWD spawning tanks that contribute negligibly, if at
all, over a spawning season. Monitoring and replacing
non- or low-contributing dams might be a strategy to at
least marginally increase the number of mating
combinations per spawn, although this might prove
difficult given the need for a "conditioning" period
prior to spawning activity. The second approach,
equalizing the number of progeny generated per mating
combination within a spawning tank, would seem in

practice difficult to in
likely be unprodu(
reduction in Ne per s
produced per mating
(about 9%).
The third approach,
from different spawni
would seem to be the
released population w
the average Ne per Spi
the number of spawn
included in the mixtUl
five different spawnil
single release, the ave
fold increase in N eR' Jl
17.2) could be achiev,
progeny from differeJ
released population. E
ized mixture as simul
increased to 16.1. G
number of release pop!
in 2003, the maximum
(i.e., using a pseudoe
from more than 110 to :
relative to that estin
program and closer to
(NeV) and long-term (!
Saillant and Gold (un
(2002), respectively.

Ackn,

We gratefully acknO\
provided by personnel,
Gamez of the Texas Pi
CCNCPL Marine Deve
We also thank C. Bradfi(
the laboratory, D. Gatli!
obtaining tissues, and T.
on an early draft of
supported by the Texas
(award NA16RGlO78), 1
of the Texas Parks ar:
Coastal Conservation ,
Texas Agricultural Exp
6703). This paper is num
Studies in Marine Fishes
Center for Biosystemati(
A&M University.

Ref,

Colura, R. L. 1987. Saltwa!
48-50 in G. Chamberlai
editors. Manual of red