After fertilization occurred (either due to selfing
or outcrossing), males were returned to rearing tanks
while the hermaphrodites were kept in isolation so that
eggs could be collected. Eggs remained in the brood
chamber for approximately 24 hours. Hermaphrodites
were returned to the rearing tanks after they dropped
their eggs. On one occasion, a hermaphrodite was seen
dropping its clutch and then selfing another clutch within one hour.
Discussion
Sex ratios
The majority of Conchostracan species have a 50:50
sex ratio between males and females (Sassaman, 1989).
The skewed sex ratio of E. texana can be attributed to the nature of sex determination for this system.
Comparisons of sex ratios for thirteen E. texana populations across the United States showed an average
sex ratio of approximately 75-80% hermaphrodites,
20-25% males (Sassaman, 1989). The sex ratio in
the LTER playa is fairly characteristic for the species.
One population, SWP3, was entirely hermaphroditic.
The percentage of males within the SWP4 and SWP5
populations (40-42% respectively) are somewhat high
when compared to the range of sex ratios in populations examined by Sassaman (range = 0-35% males)
(Sassaman, 1989).
Population genetics modeling of the E. texana
breeding system has suggested that the costs associated with sexual reproduction would make the gradual
elimination of males a natural occurrence if some type
of inbreeding depression were not present to offset
the benefits of selfing (Otto et al., 1993). Sassaman
(1989) reported probable all-hermaphrodite populations of E. texana in Florida and Texas. What makes
the observed absence of males in SWP3 surprising,
however, is the close proximity (35 m) of this population to other male-containing popUlations. Eggs of
E. texana have a diameter of 220 p, (Belk, 1989) and
could be dispersed by wind action or on the feet of
cattle and birds that drink from the three small neighboring populations. If hermaphrodites bury their eggs
within burrows, however, as current observations suggest, these mechanisms of dispersal may be limited
in effectiveness. Other factors which maintain males
within some of these E. texana populations and seemingly inhibit gene flow between these populations are
as yet unknown.
79
The number of viable eggs hatching from the LTER
playa popUlation were significantly less than those
hatching from the three SWP popUlations. Data are
unavailable on the length of time E. texana eggs can
remain viable in the field, but in general the population
densities of all of the species living within the LTER
Playa appear to decrease with an increase in length of
time between floodings (MacKay et aI., 1990). The
LTER Playa has not seen a significant flood in ten
years (October, 1984) and this may account for lower
hatch rates from this soil when compared to the samples taken from the other three populations. The SWP
populations typically flood at least once every year
and were flooded during the summers of 1992 and
1993 (pers. obs.). Also the LTER Playa clam shrimp
share the playa with other ephemeral shrimp. These
species may provide competition for food resources or
predation pressures which lower E. texana population
densities. Such pressures would be absent from the
SWP popUlations because no other ephemeral shrimp
live within these sites. The bare mud substrates of the
SWP sites may also be better suited for clam shrimp
burrowing and egg laying than the vegetated LTER
Playa substrate.
The tadpole shrimp Triops longicaudatus also
hatched within the LTER Playa rearing tanks. Previous research had suggested that the tadpole shrimp,
T. longicaudatus only feeds on dead prey (Medland,
1989; MacKay etal., 1990; Weeks, 1990; Loring, pers.
comm.). This conflicts with the behavior of T. longicaudatus, in our laboratory where they were avid
predators; thus all tadpole shrimp were removed from
rearing tanks to avoid predation. When tadpole shrimp
had been left within a tank, they ate all of the clam
shrimp within a matter of days. Similar predation by
T. longicaudatus has been reported for other laboratory
populations of E. texana (Sassaman & Weeks, 1993).
This may not occur as often in the field because T. longicaudatus typically remain near the bottom of a playa
while E. texana remains around the periphery (Medland, 1989; pers. obs.). Tank size constraints in the
laboratory may make contact between the two species
more common and thus predation more prevalent.
Behavioral observations
This study supports several predictions of a recent population genetics model designed to study the relative
costs of outcrossing and selfing (Otto et aI., 1993).
One prediction of the model is that male mortality rates
will be higher than hermaphrodite mortality rates. High
or outcrossing), males were returned to rearing tanks
while the hermaphrodites were kept in isolation so that
eggs could be collected. Eggs remained in the brood
chamber for approximately 24 hours. Hermaphrodites
were returned to the rearing tanks after they dropped
their eggs. On one occasion, a hermaphrodite was seen
dropping its clutch and then selfing another clutch within one hour.
Discussion
Sex ratios
The majority of Conchostracan species have a 50:50
sex ratio between males and females (Sassaman, 1989).
The skewed sex ratio of E. texana can be attributed to the nature of sex determination for this system.
Comparisons of sex ratios for thirteen E. texana populations across the United States showed an average
sex ratio of approximately 75-80% hermaphrodites,
20-25% males (Sassaman, 1989). The sex ratio in
the LTER playa is fairly characteristic for the species.
One population, SWP3, was entirely hermaphroditic.
The percentage of males within the SWP4 and SWP5
populations (40-42% respectively) are somewhat high
when compared to the range of sex ratios in populations examined by Sassaman (range = 0-35% males)
(Sassaman, 1989).
Population genetics modeling of the E. texana
breeding system has suggested that the costs associated with sexual reproduction would make the gradual
elimination of males a natural occurrence if some type
of inbreeding depression were not present to offset
the benefits of selfing (Otto et al., 1993). Sassaman
(1989) reported probable all-hermaphrodite populations of E. texana in Florida and Texas. What makes
the observed absence of males in SWP3 surprising,
however, is the close proximity (35 m) of this population to other male-containing popUlations. Eggs of
E. texana have a diameter of 220 p, (Belk, 1989) and
could be dispersed by wind action or on the feet of
cattle and birds that drink from the three small neighboring populations. If hermaphrodites bury their eggs
within burrows, however, as current observations suggest, these mechanisms of dispersal may be limited
in effectiveness. Other factors which maintain males
within some of these E. texana populations and seemingly inhibit gene flow between these populations are
as yet unknown.
79
The number of viable eggs hatching from the LTER
playa popUlation were significantly less than those
hatching from the three SWP popUlations. Data are
unavailable on the length of time E. texana eggs can
remain viable in the field, but in general the population
densities of all of the species living within the LTER
Playa appear to decrease with an increase in length of
time between floodings (MacKay et aI., 1990). The
LTER Playa has not seen a significant flood in ten
years (October, 1984) and this may account for lower
hatch rates from this soil when compared to the samples taken from the other three populations. The SWP
populations typically flood at least once every year
and were flooded during the summers of 1992 and
1993 (pers. obs.). Also the LTER Playa clam shrimp
share the playa with other ephemeral shrimp. These
species may provide competition for food resources or
predation pressures which lower E. texana population
densities. Such pressures would be absent from the
SWP popUlations because no other ephemeral shrimp
live within these sites. The bare mud substrates of the
SWP sites may also be better suited for clam shrimp
burrowing and egg laying than the vegetated LTER
Playa substrate.
The tadpole shrimp Triops longicaudatus also
hatched within the LTER Playa rearing tanks. Previous research had suggested that the tadpole shrimp,
T. longicaudatus only feeds on dead prey (Medland,
1989; MacKay etal., 1990; Weeks, 1990; Loring, pers.
comm.). This conflicts with the behavior of T. longicaudatus, in our laboratory where they were avid
predators; thus all tadpole shrimp were removed from
rearing tanks to avoid predation. When tadpole shrimp
had been left within a tank, they ate all of the clam
shrimp within a matter of days. Similar predation by
T. longicaudatus has been reported for other laboratory
populations of E. texana (Sassaman & Weeks, 1993).
This may not occur as often in the field because T. longicaudatus typically remain near the bottom of a playa
while E. texana remains around the periphery (Medland, 1989; pers. obs.). Tank size constraints in the
laboratory may make contact between the two species
more common and thus predation more prevalent.
Behavioral observations
This study supports several predictions of a recent population genetics model designed to study the relative
costs of outcrossing and selfing (Otto et aI., 1993).
One prediction of the model is that male mortality rates
will be higher than hermaphrodite mortality rates. High