JOURNAL OF CRUSTACEAN BIOLOGY, 35(2), 185-190, 2015
DECREASED TEMPERATURE RESULTS IN DAYTIME LARVAL RELEASE BY THE
FIDDLER CRAB UCA DEICHMANNI RATHBUN, 1935
Kecia A. Kerr 1,2,3,∗
1 Department of Biology, McGill University, 1205 Docteur Penfield, Montréal, Québec, Canada H3A 1B1
2 Smithsonian Tropical Research Institute, Apartado Postal 0843-03092, Panamá, República de Panamá
3 McGill-STRI Neotropical Environment Option (NEO), Faculty of Science, McGill University,
Montréal, Québec, Canada H3A 2T6
A B S T R A C T
Many crabs release their larvae during large amplitude high tides at night to reduce predation by fishes. These “safe” tides often occur within
a few hours of dawn, therefore crabs targeting these tides are vulnerable to releasing larvae during daylight if a decrease in temperature
extends incubation by a few days. Uca deichmanni Rathbun, 1935 release larvae during the large amplitude tides, but the timing of larval
release with respect to the tidal and diel cycle was previously unknown. This species released larvae exclusively during the high tide,
approximately two hours before sunrise, in warm conditions in the laboratory. In cold water, females released larvae primarily during the
high tide, but many released during daylight (55%). During cold conditions in the field, 35% of females released larvae on days when the
morning high tide occurred during daylight. These females or their larvae may therefore have been exposed to higher risk of predation by
diurnally feeding predators.
KEY WORDS: larvae, predation risk, reproductive cycles, temperature, timing of reproduction, upwelling
DOI: 10.1163/1937240X-00002334
INTRODUCTION
Many species of brachyuran crabs demonstrate strong cycles
of larval release with an impressive ability to target larval
release with particular moments in tidal and diel cycles.
The most prevalent timing of larval release, especially for
intertidal crabs (83% of species surveyed), is during large
or maximum amplitude (spring) high tides at night (Christy,
2011). Predation by fishes on newly released larvae is
higher during the day in many habitats (Hovel and Morgan,
1997; Morgan and Christy, 1997; Kerr et al., 2014a). This
timing of larval release therefore likely reduces the risk
of predation on females and larvae by diurnally feeding
fishes that are abundant nearshore (Morgan, 1990, 1995;
Morgan and Christy, 1995, 1997; Hovel and Morgan, 1997;
Kerr et al., 2014a). A few crab species nevertheless release
larvae during both day and night, or release during the
day, if the largest amplitude tides do not occur at night
(Morgan and Christy, 1995; Kellmeyer and Salmon, 2001;
Yamaguchi, 2001; Christy, 2011; Rasmuson et al., 2014).
In addition, there are conditions under which females may
be unable to match the timing of larval release with the
best tides, the nocturnal high tides with large amplitude,
and must release during suboptimal tides. For example, as
temperature changes, crabs may not be able to compensate
for the effects of temperature on incubation period and the
timing of hatching may shift by several days (Kerr et al.,
2012).
∗ Current address: Australian National University, Research School of Biology, Division of Evolution, Ecology and Genetics, Canberra,
ACT 2601, Australia. Mailing address: 9807 78 Avenue NW, Edmonton, AB, Canada T6E 1N4. E-mail: kecia.kerr@mail.mcgill.ca
The safest tides for larval release, the maximum or large
amplitude nocturnal high tides (Christy, 2011), occur for a
few days approximately biweekly in areas with semidiurnal
or mixed semidiurnal tides, such as along much of the
coasts of the Americas. Maximum amplitude high tides, the
targeted period for larval release, tend to occur just before
dawn and dusk in Panama. Since the high tide occurs approx.
40 min to 1 hour later each day, high tide releases that occur
several days after the maximum amplitude tides will occur
after sunrise or sunset. Embryos that develop at relatively
low temperatures have longer incubation periods and may
not be ready to hatch until several days after the maximum
amplitude tide. Once the morning high tides occur after
sunrise, crabs could potentially avoid releasing larvae during
the day by switching to the early evening high tide, thus
maintaining release during the cover of darkness.
The ecology and behavior of Uca deichmanni Rathbun,
1935 is relatively unknown. This species exhibits cycles
of courtship and larval release that is synchronized with
the largest amplitude tides, but the timing of release of
larvae relative to tidal and diel cycles has not been reported
(Zucker, 1983; Kerr et al., 2012, 2014b). If U. deichmanni
release their larvae at the time of the high tide as expected,
they might avoid hatching during daylight when conditions
cool by switching from the morning high tide to the evening
high tide. I examined the timing of the release of larvae
by females kept in the laboratory relative to tidal and diel
cycles. Females were held in: 1) ambient, warm conditions
© The Crustacean Society, 2015. Published by Brill NV, Leiden DOI:10.1163/1937240X-00002334
186 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 35, NO. 2, 2015
that are the norm for about 8 months of the year, and 2) a
cold treatment that corresponds to temperatures that females
and embryos experience during the coldest part of the annual
cycle. Using the results of the laboratory experiment, I also
re-examined field data on the dates of larval release across
seasonal changes in temperature and inferred the timing of
larval release relative to the diel cycle based on the time of
the morning high tides.
MATERIALS AND METHODS
Study System
The study area is located at the Pacific entrance to the Panama Canal in
the Gulf of Panama, Central America. Tides are semidiurnal and of large
amplitude (range = 2-6 m). The tidal amplitude cycle ranges in period
length from 12-17 days. The spring high tides, the tides of large amplitude
that occur during a 3-5-day period surrounding the maximum amplitude
tide, usually occur between 4:00 and 7:00 a.m. and p.m. Sunrise and sunset
occur at approximately 6:00-6:30 a.m. and p.m. in this tropical location.
Water temperature in this area is typically 28-29°C and is constant during
much of the year, but the Gulf of Panama experiences seasonal upwelling of
cold, deep-source waters between the months of December and May, which
decreases surface temperatures to approx. 20-25°C (D’Croz and O’Dea,
2007).
Uca deichmanni is a common species of fiddler crab found in the
mid intertidal of low-energy sandy beaches in the tropical eastern Pacific
from northern Costa Rica to northern Colombia. Females carry embryos
on pleopods below their abdomen until they are ready to hatch. During
incubation of the embryos, females spend the majority of their time in the
sediment in closed burrows, but they are occasionally seen on the surface
of the sand cleaning out the burrow and producing a mound of balls of sand
(0.5-1 cm diameter), which they then use to cover the burrow entrance (Kerr
et al., 2012). Burrows of ovigerous females were identified by the presence
of this mound of sand balls covering a burrow entrance. Ovigerous females
were collected from burrows at “Puente,” a beach on the eastern side of the
Panama Canal below the Bridge of the Americas (see Kerr et al., 2014b),
during late afternoon low tides on 3 to 4 consecutive days for each of 3 runs
of the laboratory experiment. Burrows of ovigerous females were identified
by the presence of a mound of sand balls covering a burrow entrance.
Laboratory Experiment
To determine the effects of temperature on the timing of larval release
relative to the tidal and diel cycles, I recorded the timing of larval release in
two temperature treatments in the lab. Upon collection, ovigerous females
were placed in plastic containers with sub-compartments, each with a few
drops of seawater to keep the broods moist, and kept in a small cooler
for transport to the Smithsonian Tropical Research Institute (STRI)’s Naos
Marine Laboratories located on Naos Island, Amador Causeway, Panama
City, Panama. Upon arrival to the lab, a small sample of eggs was taken
from each brood for assessment of developmental stage and females were
placed in aquaria as described below.
I observed eggs under a dissecting microscope and estimated the
approximate percentage of yolk remaining. The majority of crabs started the
experiment with broods of eggs containing high levels of yolk (average ±
SE = 79.0 ± 3.13%, median = 90%, range = 25-100%), and carried
out most of their development in the treatments. Incubation periods for
crabs starting the experiment with broods with 90-100% yolk averaged
8.9 days (SD = 0.99 days) in the warm treatment and 12.4 days in the
cold treatment (SD = 1.93 days). Each crab was individually labelled by
affixing a numbered tag made of vinyl tape to her carapace with a tiny drop
of cyanoacrylate glue.
Experiments were conducted during the wet (non-upwelling) season
between June and August in 2010 in the outdoor seawater pavilion at
Naos Laboratories. The pavillion is covered with a roof but is open
on the sides, shielding the aquariums from rain, but providing natural
light. For each water temperature treatment, a 190-liter insulated drum,
receiving constant inflow from the laboratory seawater system, supplied
water to the experimental aquariums through insulated PVC tubes. Ambient
temperature in the seawater system pumped from the Bay of Panama was
28.9 ± 0.52°C (average ± SD) during the course of the experiments. For
the cold treatment, water was chilled using a Tradewind 1/5 HP Drop-in
chiller. Average temperature ± SD in the cold treatment aquariums was
23.0 ± 0.42°C. Small aquariums (30 cm × 30 cm × 20 cm) were randomly
assigned to temperature treatments with three replicate aquariums for each
temperature treatment (ambient or chilled). Aquariums were covered on all
sides with black plastic and then insulated using 1.27-cm-thick polystyrene
foam. Tops of aquariums were left open to expose females to ambient light
but the air-water interface between the tubes that contained the females was
insulated with polystyrene foam to minimize the effect of air temperature on
water temperature. Aquariums were equipped with air stones and constant
water flow. Females were randomly assigned to treatments and replicate
aquariums and placed in PVC tubes (10 cm long × 3 cm inner diameter)
with mesh on one end (Christy, 1982). The tubes were placed in the
aquariums so that the females were approximately half covered by water.
An additional tube containing an iButton temperature logger was placed in
each aquarium in the same manner as those containing the females. iButtons
measured water temperature every hour.
To determine when U. deichmanni females released larvae relative to
the tidal and light/dark cycle, the females were checked for hatched larvae
multiple times per day during three experimental runs. The time between
consecutive checks on the females varied, so the time of each check and
which females, if any, had released larvae was recorded. The time of release
for each hatching event was estimated as the midpoint between the time
when the release was recorded and the time of the previous check on
the female. Precision of the estimated of time of release (the time period
between consecutive checks) was determined for each release. Females
whose hatching events were not recorded within a 3-hour time period were
excluded from the analysis.
Timing of Release Relative to the Diel Cycle in the Field
Larval release relative to the biweekly tidal amplitude cycle across seasonal
temperature variation in field enclosures was reported in Kerr et al. (2012).
The dates of larval release by U. deichmanni for 278 individuals over 8 tidal
amplitude cycles between 3 February and 15 April 2009 during upwelling
and 21 May and 29 July 2009 during non-upwelling were recorded in that
study. The present laboratory study inferred the time of release relative to
the diel cycle for the Kerr et al. (2012) field data based on the time of the
morning high tide on each day of release. These times of the morning (a.m.)
high tides on the days when larvae were released (estimated times of larval
release) were pooled for 3 tidal amplitude cycles during the non-upwelling
season and 5 tidal amplitude cycles during the upwelling season. I used
these times of the morning high tide, when females likely released larvae,
to compare the timing of larval release relative to the light/dark cycle during
the two seasons.
Data Analysis
Timing of release of larvae relative to the time of day was analyzed using
circular statistical methods described in Zar (2010). Hatching times were
categorized into hourly bins. For example, all releases that were estimated
to have occurred between 6:00 and 6:59 were categorized as 6:00. Angles,
or phase, for each hour of the day were determined by dividing 360° by
24 such that each hour of a day represents 15°. Binned hourly hatching
times were converted to angles in degrees and plotted on circular plots using
the GGplot2 package in R (Wickham, 2009) and subjected to descriptive
and inferential statistics using the circular statistics package CircStat in
Matlab (Berens, 2009). Mean angle (a) and resultant vector length or
dispersion (R, range from 0 for randomly distributed to 1 for to identical
values for all observations) are reported. To determine if release timing
was significantly clustered, I used the omnibus or Hodges-Ajne test for
uniformity of distribution and the v-test to test for deviation from a specified
angle. The omnibus or Hodges-Ajne test (Zar, 2010) for circular uniformity
provides a probability that the data come from a population with a uniform
distribution of angles without the constraint of an underlying assumption of
a unimodal distribution. A p(o) < 0.05 indicates that there is significant
clustering of data. The v-test allowed me to test the alternative hypothesis
that the data points are clustered around a mean angle not significantly
different from 90° (6:00 a.m., approximate timing of sunrise). A significant
result (p(v) < 0.05) indicates that the data are significantly clustered
around a mean value not different from 90°, but a p-value > 0.05 could
indicate that the data are uniformly distributed or that the mean angle is
different from 90°. When the omnibus test is significant but the v-test
is not, one can conclude that the releases are clustered but the timing is
significantly different from that of 6:00 a.m. (around sunrise).
KERR: TIMING OF LARVAL RELEASE IN UCA DEICHMANNI 187
RESULTS
Laboratory Experiments
The times of 66 releases of larvae were recorded in the lab
with a precision of 3 hours or less (time between consecutive
checks on the females). Of all releases in both the warm
and cold treatments 88% occurred during a time interval
that included the high tide (Table 1). Only 2 larval releases
occurred near the time of the low tide and the remaining 6
occurred during the ebbing tide. All U. deichmanni females
kept in the warm treatment released their larvae during the
high tide. The majority of these hatches (88%) occurred at
night or near dawn (Table 1). Only 3 individuals in the warm
treatment released larvae after sunrise. These all occurred
during the high tide on days when the high tide occurred
between 12:00 and 2:20 p.m. and were “early” rather than
“late” relative to the day of the maximum amplitude tide
(i.e., the tides were still increasing in amplitude rather than
decreasing). One of these individuals carried eggs with
only 25% yolk at the beginning of the experiment but the
other two carried broods with 80% yolk when collected and
released larvae after a short incubation period relative to the
other females in the same treatment. Of the 42 releases of
larvae in the cold treatment, 81% occurred during high tide,
4 of which were at night, 20 at dawn and 10 during the
day. All larval releases that occurred after high tide (19%)
were by females in the cold treatment on days when the
morning high tide occurred during daylight. Of all females
in the cold treatment (all tidal states), 33% released larvae
at least one hour after sunrise and 57% released at dawn.
Since sunrise occurred between 5:58 and 6:10 a.m. during
the experiments but releases that occurred between 6:00 and
6:59 were categorized as 6:00 a.m., 9 of the larval releases
categorized as 6:00 a.m. actually occurred just after sunrise.
When these releases are included, 55% of females released
larvae after sunrise in the cold treatment.
Using the hourly bins of estimated times of release, I
employed circular statistics to calculate the average time of
day of hatching and the dispersion or synchrony of hatching
times. Under ambient, warm conditions the average timing
of release was between 4:00 and 5:00 a.m., while in chilled
conditions the average release occurred at approximately
7:00 a.m. (Fig. 1). In both temperature treatments, the time
of release was highly clustered (p(o) < 0.001) and was
not significantly different from the time of sunrise, which
ranged from 5:58 to 6:10 a.m. during the study (approx. 90°,
p(v) < 0.001). The majority of releases therefore occurred
during the large amplitude high tides that occurred within a
couple of hours of the time of sunrise. Of particular note is
the fact that U. deichmanni females always released larvae
during the morning high tide, even when hatching occurred
on days when the high tides were at dawn and dusk or shortly
thereafter, rather than switching to the high tide at dusk or
early evening (Fig. 1).
Timing of Release in the Field
The average timing of larval release in the field enclosures
during 3 tidal amplitude cycles of the non-upwelling (warm)
season occurred 2-3 days before the maximum amplitude
tide (Kerr et al., 2012). Assuming that females released their
larvae during the morning rather than evening high tides in
the field, as they did in the lab, I estimated the time of larval
release to be the time of the morning high tide on the days
of release. The time of the high tide on the mean day of
larval release across the pooled 3 tidal amplitude cycles was
about 4:00 a.m. (Fig. 2). During these three cycles only a few
individuals released larvae on days when the morning high
tide happened at or just after sunrise (0-22% per cycle, high
tide occurred at 7:51 a.m. on the day of the latest release,
Fig. 2), but the majority of releases would have occurred
during darkness.
In contrast, the time of the morning high tide on the
average day of release was 5:30 a.m. during the upwelling
season (Fig. 2). Of all releases recorded during the upwelling
season (5 tidal amplitude cycles) 65% occurred on days
when the high tide occurred at night or just before dawn and
another 7% released on days when the high tide occurred
within a half hour of sunrise (Fig. 2). Nonetheless, the
average days of release when only the 3 coldest tidal
amplitude cycles are considered, had morning high tides
that occurred after sunrise (6:40 to 7:40 a.m.), and 47 to
81% of all releases for these cycles occurred on a day
when the morning high tide was after sunrise. The latest
release occurred on a day when the morning high tide was at
11:19 a.m. (Fig. 2).
DISCUSSION
This study establishes that U. deichmanni releases larvae
mainly during nocturnal, spring, high tides. This timing
is the most common pattern (83%) for the 36 species
of intertidal crabs for which data are available for all 3
environmental cycles (diel, tidal, tidal amplitude) (Christy,
2011, supplementary table). Uca deichmanni continued to
release larvae during the morning high tides even when the
period of incubation increased by several days in the cold-
water treatment and the morning high tide occurred after
sunrise. Some species, including Uca terpsichores Crane,
1941, a species of fiddler crab that is closely related to,
and often lives on the same beaches as U. deichmanni,
Table 1. Number of females of the fiddler crab Uca deichmanni releasing larvae during each tidal and diel state in ambient (approx. 29°C) and cold
(approx. 23°C) temperature treatments in the laboratory. Crabs with broods of all yolk stages at the beginning of the experiment (25-100%) were included
in the analysis but the majority of crabs started the experiment with eggs containing high levels of yolk remaining (median = 90% yolk).
Treatment N Tidal State Diel State
High Ebb Low Flood Dusk Night Dawn Day
Ambient 24 24 0 0 0 0 12 9 3
Cold 42 34 6 2 0 0 4 24 14
188 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 35, NO. 2, 2015
Fig. 1. Uca deichmanni larval release in the lab. Time of day of the release of larvae by ovigerous fiddler crabs U. deichmanni in ambient (29°C) and cold
(23°C) temperature treatments in the laboratory. The grey arrow indicates the approximate time of sunrise (approx. 6:00 a.m.) during the experiments. Crabs
with broods of all yolk stages (25-100%) were included in this analysis but the majority of crabs started the experiment with eggs containing high quantities
of yolk (median = 90% yolk). Releases by crabs beginning the experiment with 100% yolk are indicated by gray dots (10 ambient, 13 cold) while black
dots represent releases by females that began the experiment with broods containing all other yolk quantities. Times of the high tides were categorized into
hourly bins prior to analysis (ex: tides from 5:00 to 5:59 were categorized as 5:00). Sample sizes (N), summary statistics (a, mean angle; R, resultant vector
length) and p-values for the omnibus test (o) and v-test (v) are shown for each temperature treatment.
maintain the timing of larval release during darkness by
switching from the morning to the evening high tide when
the morning tide occurs after sunrise (Christy, 2003; Christy
and Backwell, unpublished data). Other crabs that live in
colder water and have low synchrony of larval release among
individuals also release during either evening or morning
high tides, thus maintaining release during darkness or
twilight for the most part (Rasmuson et al., 2014). A few
other species do release larvae, at least partially, during
the day. Yamaguchi (2001) found that 40 to 60% of Uca
lactea (De Haan, 1835) females released their larvae near
the time of high tides that occurred during daylight. Morgan
and Christy (1995) reported that Uca beebei Crane, 1941
releases larvae during the higher high tide of each 24-
hour period even though those tides often occurred during
the day. Based on experiments of predation on larvae of
differing coloration and their patterns in the timing of larval
release, they concluded that U. beebei larvae may be more
protected than other crab larvae due to their coloration and
the habitat in which they are released (Morgan and Christy,
1997). Uca thayeri Rathbun, 1900 releases larvae during the
night-time large or maximum amplitude high tides where
tides are semidiurnal, but releases larvae during daytime
large or maximum amplitude high tides in a mixed tidal
regime where larger amplitude tides do not occur at night
(Kellmeyer and Salmon, 2001; Weaver and Salmon, 2002;
Fig. 2. Uca deichmanni estimated timing of larval release in the field. Time of day of the morning high tide on days when ovigerous U. deichmanni fiddler
crabs released larvae in field enclosures during non-upwelling (average sediment temperature = 28.8°C) and upwelling (24.2°C) conditions. Data are from
3 tidal amplitude cycles during the non-upwelling season (23 May to 29 July 2009) and 5 tidal amplitude cycles during upwelling (3 February to 15 April
2009). The grey arrow on the outer margin of the circle indicates the approximate time of sunrise during data collection for that season (non-upwelling:
5:58-6:08 a.m., upwelling: 6:09-6:39 a.m.). Times of the high tides were categorized into hourly bins prior to analysis (ex: tides from 5:00 to 5:59 were
categorized as 5:00). Since releases were assumed to have occurred during the morning high tides, afternoon and evening high tides were excluded and those
time periods have been shaded in gray in the plots. Sample sizes (N), summary statistics (a, mean angle; R, resultant vector length) and p-values for the
omnibus test (o) and v-test (v) are shown for each temperature treatment.
KERR: TIMING OF LARVAL RELEASE IN UCA DEICHMANNI 189
Christopher et al., 2008). Christy (2011, supplementary
table) compiled 8 additional reports of intertidal crabs
releasing during the day out of 64 for which the diel period
was reported. For all of these instances where crabs release
larvae during the day, the crabs also release larvae at night.
This preference, or constraint, by U. deichmanni from
switching to the evening tide may have detrimental effects
on the survival of newly hatched larvae since experiments
have shown that risk of planktonic predation is higher during
the day than at night (Kerr et al., 2014a). In Panama, dusk
was the safest time period overall for tethered plankton and
risk was highest during the daytime for small plankton (Kerr
et al., in press). Furthermore, risk of predation on small
plankton was higher during the upwelling season than during
non-upwelling (Kerr et al., in press). Switching the timing
of release of larvae to the evening tide, especially during
upwelling when predation risk is higher and errors in timing
are more likely, should therefore result in increased survival
of larvae.
While the diel cycle is important for the timing of
larval release of many species, tidal amplitude and tidal
height also play roles, especially if there are constraints
on releasing larvae during large amplitude high tides at
night (Kellmeyer and Salmon, 2001; Yamaguchi, 2001;
Christopher et al., 2008; Christy, 2011). Larger amplitude
tides result in stronger tidal currents and increased potential
of movement of larvae offshore during subsequent ebb tides.
Higher high tides cover more of the intertidal habitat and
allow females in the upper intertidal to release larvae from
the safety of their shelters. Of the two high tides per day
that occur in semidiurnal areas, the high tide with higher
height often also has a larger amplitude. This is not always
the case though. High tides on the Pacific coast of Panama
occur near the time of sunrise and sunset on two days of each
semi-lunar tidal amplitude cycle. On these days, neither high
tide occurs during complete darkness. During this study,
sunrise high tides were slightly higher in height, but were
smaller in amplitude, than high tides occurring at sunset.
On all other days of the tidal amplitude cycle, high tides
during daylight, including those just after sunrise, were of
both larger amplitude and higher height than the high tides
occurring during darkness. Therefore, larvae released during
tides before dawn were released under the cover of darkness,
but were released during tides with smaller tidal amplitude
and height than the afternoon tide, and they had little time
to move offshore before daylight. In contrast, post-sunrise
releases of larvae had the advantage of larger amplitudes
and higher heights of the tide relative to post-sunset releases,
although the absolute amplitudes and heights on these days
were lower than those occurring just before sunrise and
sunset. Thus, when U. deichmanni females are forced to
release larvae later in the tidal amplitude cycle due to the
effects of temperature on development of embryos, the larger
amplitude of these daylight high tides relative to the evening
tides may somewhat diminish the risk of predation during
daylight by sweeping the larvae away from shore and the
higher densities of predators.
Uca deichmanni targets the release of larvae during high
tides that occur just before dawn during ambient, warm con-
ditions. The timing of the maximum amplitude tide ranges
from ∼4:00 a.m. to 7:00 a.m. throughout the year. The max-
imum amplitude high tides during the three tidal ampli-
tude cycles of the experiments occurred at 5:06, 5:08 and
5:20 a.m. Larval release in the ambient treatments was there-
fore very well timed with a maximum amplitude high tide
that occurred during the cover of darkness. This near-dawn
timing, however, leaves little leeway for error and means that
temperature-induced increases in incubation period of only
a day or two may result in larvae being released during day-
light. Kerr et al. (2012) found that the average day of lar-
val release in the field during warm conditions was 1-3 days
before the maximum amplitude tide and was 1-2 days after
the maximum amplitude tide during the coldest conditions.
Re-examination of this data, with the consideration that U.
deichmanni does not switch to the afternoon/early evening
high tides that occur after sunset, revealed that during the
coldest conditions, the average timing of larval release cor-
responded with tides that occurred just after sunrise and that
a large proportion of releases would have occurred during
daylight. The proportion of individuals that released larvae
during daylight in cold conditions was, however, quite a bit
lower in the field (36%) than in the lab (55%).
The reason that U. deichmanni does not switch to the
evening tide while other species do remains unknown.
Increased risk of predation associated with larval release
during the day, and higher predation risk overall during
upwelling when these crabs are most likely to release
larvae late (Kerr et al., 2012, 2014a, in press), may provide
strong selection for targeting the largest amplitude tides
that occur before dawn, or releasing larvae during darkness
on slightly lower amplitude tides, as does U. terpsichores.
Kerr et al. (2014b) showed that Uca terpsichores adjust
to changes in temperature and reduce the magnitude of
potential errors in timing of hatching by changing the timing
of courtship. In contrast, U. deichmanni does not shift timing
of courtship (Kerr et al., 2014b). Despite this, during cold
conditions in the field, U. deichmanni released larvae more
synchronously and closer to the maximum amplitude tides
than did U. terpsichores (Kerr et al., 2012). Differences in
timing between the lab and the field, high synchrony and
relatively low rate of errors in timing of release indicate
that U. deichmanni may use some method of adjustment to
decrease the magnitude of timing errors (Kerr et al., 2012).
One possibility is the regulation of incubation temperature
via movements within their burrows. Release of larvae
among female U. terpsichores is of relatively low synchrony
but a switch to releasing during the evening tide when the
morning tide occurs after sunrise may afford them some
respite from potential risk of predation associated with errors
in timing. In contrast, errors in timing by U. deichmanni,
while rare, result in releases of larvae that occur during
daylight and may prove costly in terms of larval survival,
potentially increasing selection to hit the targeted timing.
In other words, the penalty, in terms of fitness, for missing
the targeted timing may be larger in for U. deichmanni
than for U. terpsichores. On the other hand, the higher
synchrony of releases by this species may mean that the
need to switch to the evening tide may be relatively low. The
coldest conditions that resulted in relatively high proportions
of daytime releases of larvae in the field occur during only
190 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 35, NO. 2, 2015
a few of the approximately 8 biweekly tidal amplitude
cycles that occur during an upwelling season. Errors in
timing induced by temperature variation may prove quite
costly during a small number of reproductive bouts per year,
potentially resulting in reduced recruitment for a few tidal
amplitude cycles, and affecting the size of these population
cohorts. Nonetheless, the low frequency of these events, in
the context of a species that reproduces up to 26 times per
year, may limit the overall impact of these errors on the
population. The large scale effects of errors in timing of
larval release have yet to be determined.
ACKNOWLEDGEMENTS
I am grateful to J. Christy, R. Collin and F. Guichard for guidance
throughout the research and to J. Christy for encouragement to pinpoint
the timing of release of this species relative to the tides and diel cycle;
to the Collin Lab, A. Cornejo in particular, for help with collecting
ovigerous females; to C. Bonilla, J. I. Sánchez and A. Verlade for
assistance with the setup of the laboratory system; and to G. Álvarez, M.
Batista and A. Bilgray for logistical support. Funding was provided by
a Smithsonian Tropical Research Institute (STRI) Short-term Fellowship
granted to K.A.K., Natural Sciences and Engineering Research Council of
Canada (NSERC) Discovery Grant to F. Guichard (McGill University), Post
Graduate Scholarship (NSERC PGS D) to K.A.K., Quebec Ocean funding
of F. Guichard and K.A.K. and Smithsonian Marine Science Network
funding of R. Collin (STRI). The Autoridad del Canal de Panamá (ACP)
and the Unidad Administrativa de Bienes Revertidos of the Ministerio de
Economía y Finanzas, the Servicio Nacional Aeronaval, provided access to
the collection site and the Autoridad de los Recursos Acuáticos de Panamá
(ARAP) provided the research permit. J. Christy, F. Guichard, R. Collin,
J. Luque and two anonymous reviewers provided valuable comments on
earlier versions of the manuscript.
REFERENCES
Berens, P. 2009. CircStat: a MATLAB toolbox for circular statistics. Journal
of Statistical Software 31: http://www.jstatsoft.org/v31/i10.
Christopher, C. E., M. Salmon, and R. B. Forward Jr. 2008. Is the hatching
clock of fiddler crab larvae (Uca thayeri) phenotypically plastic? Journal
of Crustacean Biology 28: 328-333.
Christy, J. H. 2003. Reproductive timing and larval dispersal of intertidal
crabs: the predator avoidance hypothesis. Revista Chilena de Historia
Natural 76: 177-185.
. 2011. Timing of hatching and release of larvae by brachyuran
crabs: patterns, adaptive significance and control. Integrative and Com-
parative Biology 51: 62-72.
Crane, J. 1941. Crabs of the genus Uca from the west coast of Central
America. Zoologica; Scientific Contributions of the New York Zoolog-
ical Society 26: 145-208.
D’Croz, L., and A. O’Dea. 2007. Variability in upwelling along the Pacific
shelf of Panama and implications for the distribution of nutrients and
chlorophyll. Estuarine, Coastal and Shelf Science 73: 325-340.
De Haan, W. 1835. Crustacea, pp. 1-255. In, P. F. V. Siebold (ed.), Fauna
Japonica sive Descriptio Animalium, Quae in Itinere per Japoniam,
Jussu et Auspiciis Superiorum, qui Summum in India Batava Imperium
Tenent, Suscepto, Annis 1823-1830 Collegit, Notis, Observationibus et
Adumbrationibus Illustravit. Lugduni-Batavorum, Leiden.
Hovel, K. A., and S. G. Morgan. 1997. Planktivory as a selective force for
reproductive synchrony and larval migration. Marine Ecology Progress
Series 157: 79-95.
Kellmeyer, K., and M. Salmon. 2001. Hatching rhythms of Uca thayeri
Rathbun: timing in semidiurnal and mixed tidal regimes. Journal of
Experimental Marine Biology and Ecology 260: 169-183.
Kerr, K. A., A. Cornejo, F. Guichard, and A. C. Crespi Abril. In press.
Planktonic predation risk: effects of diel state, season and prey life history
stage. Journal of Plankton Research, DOI:10.1093/plankt/fbv006.
, J. H. Christy, R. Collin, and F. Guichard. 2012. Reducing error
in reproductive timing caused by temperature variation: interspecific
differences in behavioural adjustment by fiddler crabs. Marine Ecology
Progress Series 459: 1-16.
, A. Cornejo, F. Guichard, and R. Collin. 2014a. Planktonic
predation risk varies with prey life history stage and diurnal phase.
Marine Ecology Progress Series 503: 99-109.
, J. H. Christy, Z. Joly-Lopez, J. Luque, R. Collin, and F. Guichard.
2014b. Reproducing on time when temperature varies: shifts in the timing
of courtship by fiddler crabs. PLoS ONE 9: e97593.
Morgan, S. G. 1990. Impact of planktivorous fishes on dispersal, hatching,
and morphology of estuarine crab larvae. Ecology 71: 1640-1652.
. 1995. The timing of larval release, pp. 157-191. In, L. McEdward
(ed.), Ecology of Marine Invertebrate Larvae. CRC, Boca Raton, FL.
, and J. H. Christy. 1995. Adaptive significance of the timing of
larval release by crabs. The American Naturalist 145: 457-479.
, and . 1997. Planktivorous fishes as selective agents for
reproductive synchrony. Journal of Experimental Marine Biology and
Ecology 209: 89-101.
Rasmuson, L. K., J. L. Couture, and S. G. Morgan. 2014. Weakly
synchronized larval release maintained in the presence of predatory
fishes. Journal of Experimental Marine Biology and Ecology 454: 26-
31.
Rathbun, M. J. 1900. Results of the Branner-Agassiz Expedition to Brazil. I.
The decapod and stomatopod Crustacea. Proceedings of the Washington
Academy of Sciences 2: 133-156.
. 1935. Preliminary description of six new species of crabs from
the Pacific coast of America. Proceedings of the Biological Society of
Washington 48: 49-52.
Weaver, A., and M. Salmon. 2002. Hatching rhythms of Uca thayeri:
evidence for phenotypic plasticity. Journal of Crustacean Biology 22:
429-438.
Wickham, H. 2009. Positioning, pp. 115-137. In, GGplot2: Elegant Graph-
ics for Data Analysis. Springer, New York, NY.
Yamaguchi, T. 2001. Daytime larval release of the fiddler crab, Uca lactea
(Decapoda, Brachyura, Ocypodidae). Crustaceana 74: 545-555.
Zar, J. 2010. Biostatistical Analysis. 5th Edition. Pearson Prentice Hall,
Upper Saddle River, NJ.
Zucker, N. 1983. Courtship variation in the neo-tropical fiddler crab Uca
deichmanni: another example of female incitation to male competition?
Marine Behaviour and Physiology 10: 57-79.
RECEIVED: 9 October 2014.
ACCEPTED: 16 February 2015.