Journal of Plankton Research Vol.20 no.9 pp.1781-1796, 1998
Relationships between geotaxis/phototaxis and diel vertical
migration in autotrophic dinoflagellates
D.Kamykowski, EJ.Milligan and R.E.Reed
Department of Marine, Earth & Atmospheric Sciences, North Carolina State
University, Raleigh, NC 27695-8208, USA
Abstract. Marine dinoflagellate diel vertical migrations are often conceptually explained by a species'
geotactic and phototactic preferences, but actual simultaneous measurements are rare. Newly
collected simultaneous measurements on Heterocapsa (Cachonina) illdefina (Herman and Sweeney)
and Gymnodinium breve (Davis) are combined with similar literature information on Amphidinium
carterae (Hulbert), Peridinium faeroense (Paulsen) and Prorocentrum micans (Ehrenberg) to explore
several examples of the actual relationships between diel vertical migration and geotaxis/phototaxis.
Amphidinium carterae does not migrate, but it exhibits a negative geotaxis that may counter a small
sinking velocity. The four other species all exhibit diel vertical migrations that yield surface aggregations during daylight, but the associated combinations of geotaxis and phototaxis precision (which is
strongest when every cell in a population exhibits the same response to a stimulus and weakest when
the response is random) and sign [which is positive (negative) when motion is toward (away from)
the stimulus] are different in each case. These different taxis combinations may be related to speciesspecific sensor structure and/or placement. Furthermore, variations in the different biochemical pools
over a species* cell cycle may contribute to structural/mechanical changes that influence how a given
sensory array functions at a given time. If so, this coupling may be an important link in the growth
optimization mechanisms and occasional bloom successes of different autotrophic dinoflagellate
species under varying environmental conditions.
Introduction
Field (Hasle, 1950, 1954; Eppley et al., 1968; Berman and Rodhe, 1971; Blasco,
1978; Heaney and Tailing, 1980; Kamykowski, 1980; Staker and Bruno, 1980) and
laboratory (Eppley et al., 1968; Heaney and Furnass, 1980; Staker and Bruno,
1980; Heaney and Eppley, 1981; Kamykowski, 1981) observations of dinoflagellate diel vertical migration (DVM) often report surface aggregations during the
day and subsurface dispersal or aggregations at night. Vertical gradients in
various environmental factors like temperature and nutrients can modify this
basic pattern (Cullen, 1985; Maclntyre et al., 1997). These occurrences of DVM
are often attributed to phototactic ascent in response to light and to geotactic
descent in the dark in response to gravity without direct measurement of the
actual taxes. On the other hand, laboratory studies of geotaxis (Levandowsky and
Kaneta, 1987) and phototaxis (Hand, 1970; Forward, 1974, 1976; Hand and
Schmidt, 1975; Kreimer, 1994) are often used to infer DVM patterns without
actual measurement of DVM. Studies that undertake measurements of both
DVM and taxes are rare.
Weiler and Karl (1979) and Cullen and Horrigan (1981) obtained supplementary observations during laboratory studies of DVM that explored the occurrence
of geotaxis and/or phototaxis. Eggersdorfer and Hader (1991a,b), however,
provided the most complete set of complementary observations. They periodically collected samples from a 3 m column. Cell counts monitored DVM, while
motion analysis techniques determined the precision and sign of geotaxis and
© Oxford University Press
1781
D.Kamykowski, E JJVIilligan and R.E-Reed
phototaxis. Precision was strongest when each individual cell in a population
exhibited the same response to a stimulus and weakest when the response was
random, and sign was positive (negative) when the motion was toward (away
from) the stimulus. DVM appeared to be related to an antagonism between geotaxis and phototaxis. The present paper will provide new simultaneous observations of DVM and taxes, and summarize the results in the context of the
available data.
Method
Culture techniques
Non-axenic, unialgal stock cultures of Heterocapsa (Cachonina) illdefina
(Herman and Sweeney) (hereafter Heterocapsa illdefina) and Gymnodinium
breve (Davis) were maintained in 125 ml Erlenmeyer flasks filled with Ll/4
medium (Guillard and Hargraves, 1993), except that the Cu addition was eliminated and soil extract (0.5 ml I"1 obtained by autoclaving 100 ml of Hyponex
potting soil with 11 of deionized water) was added (i.e. modified-Ll/4). The flasks
were kept in an incubator held at 20°C and were illuminated with fluorescent
bulbs (mixed cool white and daylight) that provided 150 mmol quanta m~2 s"1
photosynthetically available radiation (PAR) on a 12 h light:12 h dark cycle.
Larger volumes of these species were grown from the stock cultures by inoculating 20 1 carboys filled with the modified-Ll/4 medium. The carboys were kept in
a temperature-controlled room set at 22°C and were illuminated by fluorescent
bulbs (shoplite) that provided 200 umol quanta m~2 s"1 PAR on a 12 h light:12 h
dark cycle.
Mesocosm design
A translucent fiberglass column (44 cm diameter by 155 cm height, holding -2251)
was washed with a weak HC1 solution, rinsed with tap water, and then placed in
a wooden frame positioned in a temperature-controlled room set at 22°C. A
300 W tungsten halogen lamp suspended over the column on the wooden frame
illuminated the column surface from 06:00 to 18:00 at -400 umol quanta m~2 s"1
PAR through a 2 cm water bath heat filter with a sand-blasted 1 cm glass diffuser
plate and neutral screens placed in the bottom of the bath. Since the mesocosm
was sampled without volume replacement, the distance between the water
surface and the light source increased by -2.5 cm per sample over the course of
each experiment. In both cases, the surface PAR decreased to -125 umol quanta
m"2 s"1 by the end of the experiment. The column was simultaneously filled with
pumped 0.2-um-filtered sea water enriched to modified-Ll/12 and with gravityfed cultures from the 20 1 carboys. Samples were collected from the mesocosm
using a weighted 5 m length of TYGON tubing capable of collecting water from
and returning water to the bottom of the mesocosm. A peristaltic pump cycled
water through the 40 cm3 of tubing (0.32 cm internal diameter) in -10 s. Profiling
at a rate of 20 cm per 10 s disturbed, but did not mix, the mesocosm population
distribution, and the sampled cells remained motile. During profiling, the exit
1782
DinoflageUate taxis and migration
tube was positioned -10 cm below the entrance port to return the sea water to
within a few centimeters of the sample depth, based on the transit time through
the tubing. In addition to looping through the peristaltic pump, the tubing outside
the mesocosm entered and exited a Turner Designs fluorometer with a 5-60 excitation filter, a 2-64 emission filter and a FT45 blue lamp, and incorporated a Yjoint for water collection. A thermistor connected to a YSI control box was bound
to the tubing in the mesocosm with the sensor at the depth of the inlet. Both the
in vivo fluorescence and temperature at pre-determined depths were digitized
from line graphs obtained with a dual-channel Weathermeasure chart recorder.
PAR extinction in the mesocosm depended on population density, population
distribution and chlorophyll a per cell. Approximate extinction coefficients,
calculated using the Riley (1956) equation k = 0.04 + 0.0088C + 0.054C2'3 where
C is chlorophyll a concentration for mid-experiment conditions at lights on, averaged 0.7 m"1 for H.illdefina and 2.4 nr 1 for G. breve, but this equation ignores a
package effect due to the fact that a G.breve cell contains 5-10 times the chlorophyll a of a H.illdefina cell.
Schedule
Heterocapsa illdefina was first introduced into the mesocosm on 6 May 1997. A
28 h sample set with in vivo fluorescence and temperature profiles every 2 h and
water collections every 4 h began at 22:00 on 7 May 1997 and ended at 02:00 on
9 May 1997. Gymnodinium breve was first introduced into the mesocosm on 1
June 1997. [See Kamykowski et al. (1998) for a detailed description of the column
manipulations.] Briefly, partial drainings of the mesocosm on 10 and 18 June 1997
while cells were aggregated at the surface yielded a quantized population that
divided synchronously about every 3 days. A 72 h sample set with in vivo fluorescence and temperature profiles every 2 h, and water collections every 6 h, began
at 06:00 on 22 June 1997 and ended at 06:00 on 24 June 1997.
Water collection
During the water collections, samples from the surface and mid-column (-0.75 m
below the surface) of the 1.5 m mesocosm were collected in two separate 21 polycarbonate Fernbach flasks through the Y-joint. The in vivo fluorescence and the
temperature were recorded after the readings stabilized during the sample collection at each depth. Each Fernbach flask was shaken and a subsample was poured
into a 125 ml Erlenmeyer flask which was then placed briefly in the dark for subsequent cell number and taxis determinations. Cell counts on 0.5 ml were made on
a Coulter Multisizer II Particle Analyzer with a 100 (am orifice that was calibrated
using 20 um polystyrene beads and blanked with the filtered sea water used to
make the medium. The remaining sample in the Fernbach flask was divided into
100 ml aliquots for either filtration (chlorophyll a and carbohydrate) or centrifugation (DNA, RNA, lipid and protein) as reported in Kamykowski et al. (1998).
The measurements for H.illdefina were limited to chlorophyll a, carbohydrate,
DNA and protein.
1783
D.Kamykowski, EJ.MiOigan and R.E.Reed
Taxis preferences
Since initial attempts at video analyses like those used by Eggersdorfer and
Hader (1991a,b) were compromised by low cell populations, an alternate method
was applied. Two nearly identical 'behavior incubation boxes' (Figure 1) adapted
from the basic plan reported in Kamykowski et al. (1988) were used to monitor
geotaxis and phototaxis, respectively. For each unit, the large outer box (Figure
1 A) was a six-sided 12 x 6.5 x 12.5 cm rectangle constructed of black acrylic. Six
4 x 0.5 x 10 cm (-20 ml) small chambers also constructed of black acrylic were
suspended in the box with attachment only at the surface. The narrow width was
designed to inhibit convection as described in Eggersdorfer and Hader (1991a,b).
These six, five-sided chambers were open at the top (Figure IB) to form slits
1.3 cm apart in the top of the large outer box. The large outer box controlled
temperature in the small chambers using water circulated around their walls
through input and output ports connected to a thermostat-controlled water bath.
A 12.5 X 7.2 X 1.7 cm hollow cover (Figure 1C; where the 1.7 cm total height
consists of a 1.2 cm enclosure height and a 0.5 cm lip) with input and output ports
connected to the same thermostat-controlled bath as the box was placed over the
top of the box to cover the slits and to control temperature. The water-jacket
cover for the geotaxis unit was made of black acrylic, and the unit was used with
the slits facing upward. The water-jacket cover for the phototaxis unit was made
of clear acrylic, and the unit was used on its side (slit length dimension horizontal)
with the slits facing a light source. A sub-cover of clear acrylic was used to seal
the phototaxis unit after thick silicone grease was smeared on the slit-side surface
of the box. For G. breve, black tape was added over slits 3 and 4 on this sub-cover
to provide a dark blank for these chambers, but these data were not used because
of the lack of replication. The water-jacket cover was placed over the sub-cover
and secured with elastic bands.
The phototaxis light source was a tungsten halogen projector lamp rated at
250 W, but powered at only 80 V. The PAR intensity at the surface of the cover
ranged between 325 and 425 umol quanta m~2 s"1 due to light dispersion, but replicates distributed over the PAR gradient generally showed the same pattern.
The spectrum of the projector lamp closely matched the spectrum of the growth
light used over the 1.5 m migration column as measured with an ASD spectroradiometer. Both the surface and the mid-column samples from the migration
column were exposed to the same PAR intensity in the phototaxis unit to facilitate intercomparisons between sample depths.
Aliquots were pipetted into successive triplicate slots in the geotaxis unit and
into alternate triplicate slots in the phototaxis unit. After the small chambers were
filled in both units, the geotaxis unit was covered with a black cloth to ensure
darkness, the secured phototaxis unit was turned on its side with slits toward the
light source, the projector lamp was turned on, and a stopwatch was started. After
25 min, the projector lamp was turned off, the phototaxis unit was turned slit side
up, and each small chamber in the phototaxis unit was emptied with a disposable
pipette into three scintillation vials representing the upper, middle and lower
thirds, respectively. All of the upper thirds of the chambers were removed first,
1784
Dinoflagellate taxis and migration
Water Bath Inlet
Cell Migration Chamber
10cm
12.5 cm
Constant Temperature
Water Bath
12 cm
A. Side Cut-Away of Behavior Incubation Box Showing Six Migration Chambers
Water Bath Outlet
Water Bath Outlet
Knob for Elastic
Band Tie-Back
3 Water Bath Inlet
12 cm
Cell Migration
Chamber
Constant Temperature
Water Bath
0 5 cm
6.5 cm
B. Top View of Behavior Incubation Box
C
Water Bath Inlet
C. Lid of Behavior Incubation Box (opaque
black for geotaxis or dear for phototaxis
analysis)
Fig. L A diagram representing the behavior incubation box depicting a side cut-away view (A), a top
view (B) and the lid (C). The geotaxis configuration is the side cut-away orientation [table at bottom
of box in (A)] with an opaque lid. The phototaxis configuration is the top view orientation [table at
bottom of box in (B)] with a transparent lid oriented toward the light source.
followed in turn by all of the middle thirds, and then by all of the lower thirds.
After 35 min, the geotaxis unit was similarly emptied. The incubation times were
selected to allow full column redistribution by swimming (-1.5 cm min"1), but not
by sinking (-0.15 cm min"1) (Kamykowski et al., 1992). The vial contents (-6.7 ml
in each) were counted using the Coulter Multisizer II Particle. The cell number
and average cell size in each vial were recorded in the 14-28 um diameter band
1785
D.Kamykowski, EXMilligan and R.E.Reed
that represented the examined species within the recorded size spectrum. The
averages were based on triplicate determinations, except for G. breve phototaxis,
which was based on duplicate determinations.
Comparisons with literature data
See Eggersdorfer and Hader (1991a,b) for a detailed description of their techniques as applied to Amphidinium carterae (Hulburt), Peridinium faeroense
(Paulsen) and Prorocentrum micans (Ehrenberg). Note that Eggersdorfer and
Hader (1991a,b) referred to the Amphidinium species as lcaterea\ but we assume
they meant 'carterae'. For DVM, they periodically obtained samples through 18
outlets spaced along the side of a vertical Plexiglas column (7 cm diameter and
300 cm length) that was illuminated from 06:00 to 22:00 h. They determined cell
densities in triplicate on an image-analysis system. For taxis determinations,
Eggersdorfer and Hader (1991a,b) incubated samples in narrow glass cuvettes
(4 X 4 X 0.017 cm) designed to decrease convection. They measured phototaxis
in the x,y plane with a white stimulating light oriented perpendicular to the IR
measuring beam, and geotaxis in the x,z plane using only the IR measuring beam.
The automated image analysis of the digitized video images handled up to 200
cells in real time. Since these measurement techniques differed from those used
here on H.illdefina and G.breve, intercomparisons of taxis responses and DVM
among the species used percent normalized data. Since the cuvette design used
in the present study was conceptually similar to that described by Eggersdorfer
and Hader (1991a,b), the assumption is made that the cumulative response
measured with the Coulter Counter approach is a result of the more instantaneous response measured with the video approach.
Precision and sign
The magnitude of the percentages for both the original H.illdefina and G.breve
data reported in this paper and the Eggersdorfer and Hader (1991a,b) data read
off their figures represented the strength of taxis precision. Again, this precision
increased in strength as more cells in a population responded in the same way at
the same time. For the data original to this paper, precision was measured by the
percentage of the inoculated population that occurred in each third of the small
chamber. A random distribution occurred when ~33% of the cell population
resided in each third of a small chamber. For the Eggersdorfer and Hader
(1991a,b) data, precision was measured based on a Rayleigh test ('0' represented
random and '1' represented all cells in one direction). The sign of the taxis indicated directionality relative to a stimulus. For the original data, this represented
accumulation (negative geotaxis and positive phototaxis) or depletion (positive
geotaxis or negative phototaxis) in the top third, and vice versa in the bottom
third of the small chambers compared to 33%. For the species from Eggersdorfer
and Hader (1991a,b), the direction represented the quadrant selected by a
majority of the cells in a population interpreted as appropriate for a positive or
negative taxis.
1786
Dinoflagellate taxis and migration
Statistical test
The distributions obtained from the replicate measurements obtained from the
behavior incubation box were tested statistically for significant differences using
SigmaStat. The results of a Tukey one-way ANOVA with pairwise comparisons
are reported in the figures to quantify the patterns.
Results
The DVM pattern of H.illdefina (Figure 2) was monitored using in vivo fluorescence (F), which was closely related to cell concentration (N in cells ml"1) as
F = 0.41 + 5.53 X lO^Af (r2 = 0.90). A strong surface aggregation (marked by the
2 contour level) occurred between 06:00 and 23:00 h. The water column below
20 cm showed decreased in vivofluorescenceas cells left this layer for the surface.
In order to emphasize the time course relationships between DVM and taxis
preferences at the sample depths, percent surface in vivo fluorescence values
[calculated as (Fj/Fm) X 100, where F, is the in vivo fluorescence value at sample
time i and Fm is the highest in vivofluorescencevalue recorded] were plotted with
the taxis measurements in Figure 3.
Heterocapsa illdefina
-120,,
6
8
10 12 14 16 18 20 22 24 26 28
Time (h)
Fig. 2. Heterocapsa illdefina. The diel vertical migration displayed as contours of in vivo fluorescence
profiles obtained from the mesocosm at 2 h intervals over a 28 h period. The closed bars along the
x-axis mark the dark period.
1787
D.Kamykowski, E J.Milligan and R.E-Reed
Heterocapsa illdefina
Mid-Column Geotaxis
18 22 02 06 10 14 18 22 02 06
Mid-Column Phototaxis
"D
Surface Phototaxis
- B
18 22 02 06 10 14 18 22 02 06
i TOP
1MID
C3BOT
18 22 02 06 10 14 18 22 02 06
18 22 02 06 10 14 18 22 02 06
•MIG
Fig. 3. Heterocapsa illdefina. Summary plot of the temporal relationship (nominally 4 h intervals)
between geotaxis (A and C) and phototaxis (B and D) precision strengths (bar chart of cuvette
positions represented by a trio for each sampling time where the top third is left, the middle third is
center and the bottom third is right) and in vivo fluorescence (connected open circles) expressed as
percents for mesocosm populations collected from the surface (A and B) and mid-column (C and D)
over a 28 h period. Error bars represent + 1 SD; n = 3 for both geotaxis and phototaxis. The numbers
above the highest part of a trio of histogram bars refer to statistical significance codes for that set. For
example, 1 refers to 'B:TM', which means the bottom (B) is significantly different from the top (T)
and the middle (M). The other codes are 2 - 'B:T'; 3 - T:M,B:TM'; 4 - T:M'; 5 - T:MB\ If there is
no number above a histogram trio, the bars are not statistically different at the 5% level. The closed
bars along the *-axis mark the dark period.
The surface population of H.illdefina exhibited strong positive geotaxis
between 14:00 and 02:00 h that weakened during the first half of the light period;
a neutral or slightly negative geotaxis occurred at 10:00 h (Figure 3A). Negative
phototaxis occurred at night, but weakened during the light period; a slightly positive phototaxis occurred at 10:00 h (Figure 3B). The mid-column population
extended the neutral or weak positive geotaxis through 14:00 h (Figure 3C), but
generally followed the tendencies of the surface population's phototactic pattern
(Figure 3D).
The weakened positive geotaxis (Figure 3A and C) and negative phototaxis
(Figure 3B and D) corresponded to the ascent of the population toward the
surface that was readily observable shortly after 06:00 h (Figure 3A and B). The
near-surface fluorescence increased and the mid-column fluorescence decreased
1788
DinoflageDate taxis and migration
(Figure 3C and D) until -14:00 h. Note that the positive geotaxis appeared to
strengthen at the surface (14:00 h) before it strengthened at mid-column. The
near-surface fluorescence then decreased, while the mid-column fluorescence
increased as the negative geotaxis and negative phototaxis returned.
The DVM pattern of G.breve (Figure 4), based on in vivo fluorescence {F)
which was closely related to cell concentration (N in cells ml"1) as F = 0.38 + 1.34
x 10~3iV (r2 = 0.94), is displayed over a 3 day period. The surface aggregation
(here marked by the 3 contour) occurred between 08:00 and 21:00 h on the first
and second days, but appeared earlier in the morning on the third day. Again, the
water column below -20 cm cleared as the aggregation formed. In order to
emphasize the time course relationships between DVM and taxis preferences at
the sample depths, the percent surface in vivo fluorescence values are plotted in
Figure 5 as in Figure 3.
The surface population of G.breve exhibited strong negative geotaxis (Figure
5 A) except at 18:00 h during the first 2 days. The light response was neutral except
at 12:00 h on the last 2 days when the positive phototaxis occurred (Figure 5B).
The mid-column population resembled the surface population in both geotaxis
(Figure 5C) and phototaxis (Figure 5D), except that the strength of the negative
geotaxis tended to be strongest at 06:00 h.
Gymnodinium breve
-120
6
12
18
24
30
36
42
48
54
60
66
72
Fig. 4. Gymnodinium breve. The diel vertical migration displayed as contours of in vivo fluorescence
profiles obtained from the mesocosm at 2 h intervals (except for 04:00 h on day 3) over a 72 h period.
The closed bars along the oc-axis mark the dark period.
1789
D.Kamykowski, EJ-Milligan and R.E.Reed
"00 061218 0006 1218 00 081218 00 0612
'00 061218 0006 1218 00 061218 00 0612
Surface Phototaxis
Mid-Column Phototaxis
100
100
80
80
60
I 60
S. 40
20
°00 06 12 18 0006 12 18 00 0612 18 00 06 12
M T O P I B M I D C3BOT -o-MIG
ho
S. 40
20
"00 061218 0006 1218 00 061218 00 0612
Fig. 5. Gymnodinium breve. Summary plot of the temporal relationship (6 h intervals) between geotaxis (A and C) and phototaxis (B and D) precision strengths (bar chart of cuvette positions represented by a trio for each sampling time where the top third is left, the middle third is center and the
bottom third is right) and in vivo fluorescence (connected open circles) expressed as percents for
mesocosm populations collected from the surface (A and B) and mid-column (C and D) over a 72 h
period. Error bars represent + 1 SD; n = 3 for geotaxis and n = 2 for phototaxis. The numbers above
the highest part of a trio of histogram bars refer to statistical significance codes for that set. For
example, 1 refers to 'TMB', which means the top (T) is significantly different from the middle (M)
and the bottom (B). The other codes are 2 - 'T:M,M:B'; 3 - T:MB,M:B'; 4 - 'T:M*; 5 - 'MB'; 6 'T:B'; and 7 - T:B,M:B'. If there is no number above a histogram trio, the bars are not significantly
different at the 5% level. The closed bars along the x-axis mark the dark period.
Negative geotaxis (Figure 5A and C) and a neutral response to light (Figure
5B and D) corresponded to the ascent of the population toward the surface that
was readily observable shortly after 06:00 h (Figure 5A and B). The near-surface
fluorescence increased and the mid-column fluorescence decreased (Figure 5C
and D) until 12:00 h. Positive phototaxis at 12:00 h (Figure 5B and D) probably
contributed to the surface aggregation. However, the 6 h sampling intervals
limited the resolution of the detailed time course of positive phototaxis. The
weakened negative geotaxis at 18:00 h (Figure 5A and C) corresponded to the
descent from the surface to depth. The return of strong negative geotaxis at
24:00 h was related to a slight tendency toward the surface, but the real ascent
apparently began after lights on. Notice that the strength of the surface aggregation appeared to weaken over the 3 day period as the observed geotaxis and
1790
DinoflageDate taxis and migration
phototaxis strengthened. The mechanical disturbance of the cells inherent to the
taxis measurement may strengthen the ascent response of cells that otherwise
remain more dispersed in the water column. Strong surface re-aggregation visually observed after the local, near-surface disturbance imposed by column profiling supports this hypothesis.
Discussion
The present study has the same goal as that expressed by Eggersdorfer and Hader
(1991a): 'to study daily vertical migrations... and to correlate this behavior with
photo- and graviorientation (geo-orientation) of the organisms'. Although other
environmental factors may influence how taxis preferences translate into cell
trajectories in the field, controlled laboratory-based studies help elucidate the
direct relationships. To provide a connection between data sets that were
collected in different, but conceptually related, ways, the Eggersdorfer and Hader
(1991a,b) data for A.carterae, P.faeroense and P.micans were normalized to signed
ratios scaled from 0 to 1 for positive taxis and from 0 to -1 for negative taxis, and
were plotted with similarly normalized data for H.illdefina and G.breve (Figure
6). For phototaxis, the data were obtained from their figures at 555 umol quanta
m -2 s -i P A R for p.micans at 13 days and for P.faeroense at 23 days as the light
intensity most similar to the original measurements reported here. In their study,
A.carterae did not show phototaxis. Note that Eggersdorfer and Hader (1991b)
reported that phototaxis precision and sign depended on the intensity of the stimulus. For geotaxis, the data were obtained from their figures for P.micans at 19
days, for P.faeroense at 12 days and for A.carterae at 10 days. Although different
days were used to represent the geotaxis and phototaxis of P.micans and
P.faeroense because of the way data were reported in the source papers, a middle
day in the observation set for each species was chosen [see Eggersdorfer and
Hader (1991a,b) for how the taxis preferences varied on different days in an
observation set]. The temporal series of surface aggregations of P.micans and
P.faeroense were measured with a ruler to represent DVM patterns; Axarterae
did not show a DVM.
Table I summarizes the diversity of taxis responses that are involved with a
surface aggregation during DVM. No two species of the five studied exhibit
exactly the same combination of geotaxis/phototaxis signs. This is a surprising
result considering the limited number of possible combinations of geotaxis/phototaxis signs.
Amphidinium carterae (Figure 6A) is a relatively weak swimmer that appears
to maintain a position in the water column that does not vary by more than 1 m
over a 24 h period (Kamykowski and Zentara, 1977; Eggersdorfer and Hader,
1991a). Amphidinium carterae always exhibits relatively strong negative geotaxis,
but no phototaxis. This negative geotaxis may allow this species to remain
suspended in the water column at a given depth. In this case, the upward orientation may compensate for the slow sinking rate of this small cell.
The four other species all form surface aggregations during the day, but each
is characterized by a different temporal sequence of taxis precision in addition to
1791
D.Kamykowski, EJ.Milligan and R.E.Reed
Amphidinium carterae
• DVM
• Photo
•Goo
1.0
0.5
0.0
-0.5 •
-1.0
-
No DVM
No Photo
4
0
4
8
12
16
20
24
28
32
4
0
4
8
12
16
20
24
28
32
4
0
4
8
16
20
24
28
32
16
20
24
28
32
Peridinium faeroense
1.0
0.5
0.0
-0.5
-1.0
-
-
Prorocentrum mkans
Gymnodinium breve
1.0
0.5
0.0
-0.5
-1.0 •
12
Time (h)
Fig. 6. A summary of the temporal relationship between the precision strength and sign of phototaxis
(open triangle) and geotaxis (open inverted triangle) and the surface pattern of biomass (closed
circles) expressed as percents for five species of marine dinoflagellates. (A) Amphidinium carterae;
(B) Peridinium faeroense; (C) Prorocentrum micans; (D) Heterocapsa illdefina; (E) Gymnodinium
breve. On the y-axis, '0 to 1' represents a positive taxis, while '0 to - 1 ' represents a negative taxis. The
bars along the jr-axis mark the dark (closed) and light (open) periods for each data set.
1792
Dinoflagellate taxis and migration
Table I. Summary of taxis responses and DVM for five marine dinoflagellate species. For taxis
responses, 'Dark' refers to night, 'Light' refers to daylight, 'No' refers to a neutral response,'-' refers
to a negative response,'+' refers to a positive response,'+/-' or '-/+' refer to the occurrence of both
types of responses in a time period, and '-/No' or '+/No' refer to a signed response combined with a
neutral response in a time period. For DVM, 'No' refers to the absence of DVM, while 'Yes' refers to
the presence of DVM
Geotaxis
Phototaxis
Amphidinium carterae
Peridinium faeroense
Prorocentrum micans
Heterocapsa (Cachonina)
illdefina
Cymnodinium breve
Dark
Light
No
No
No
DVM
Dark
Light
+
+
+
-/+
No
Yes
Yes
Yes
-/No
Yes
the previously observed differences in taxis signs. Peridinium faeroense (Figure
6B) always exhibits positive geotaxis and positive phototaxis, but the strengths
vary with time of day. Phototaxis is weakest during the night and generally
strengthens over the light period contributing to ascent. Geotaxis weakens during
the first 8 h of the light period contributing to ascent, but is strong the rest of the
time. This species tends to accumulate at the surface ~4 h after lights on and
remains concentrated until -2 h before lights out, based on weakened positive
geotaxis and increasing positive phototaxis. Descent appears to be related to
strengthened positive geotaxis.
Prorocentrum micans (Figure 6C) exhibits negative phototaxis under higher
light intensities, but is always negatively geotactic. Positive phototaxis at lights on
contributes to ascent. However, negative phototaxis is strongest ~6 h after lights
on, gradually weakens over the next 6 h, and then is replaced by positive phototaxis for the remainder of the day. This observation of negative phototaxis may
be exceptional since Levandowsky and Kaneta (1987) state that 'there are no
reports of negative phototaxis in all the studies of dinoflagellate phototaxis', but
Kreimer (1994) stated that negative phototaxis is observed in flagellated algae at
high PAR intensities. The negative phototaxis may support descent at midday,
but the negative geotaxis oscillates in strength with a minimum at lights on and a
maximum at 6-10 h after lights on. This maximum negative geotaxis may compensate for the negative phototaxis over the same time period. This species shows a
gradual ascent toward the surface that maximizes 10 h after lights on in response
to weakened negative phototaxis and strong negative geotaxis. Descent appears
to be related to weakened negative geotaxis.
To summarize the patterns for new species first described in this paper, H. illdefina (Figure 6D) generally exhibits positive geotaxis and negative phototaxis,
except ~4 h after lights on. This second observation of negative phototaxis in a
dinoflagellate species may mean that this response is more common than previously thought. The surface aggregation of this species increases soon after lights
on and is related to the relaxation of strong taxis preferences that could keep the
1793
D.Kamykowski, EJ.MMigan and R.E.Reed
cells away from the surface. Descent appears to be related to strengthened positive geotaxis. Gymnodiniwn breve (Figure 6E) exhibits occasional positive phototaxis and positive geotaxis that varies in strength with time of day. The surface
aggregation of this species strengthens soon after lights on and is related to the
maintenance of strong taxis preferences that bring the cells to the surface.
Descent appears to be related to weakened negative geotaxis.
The reasons behind the diversity of relationships between positive or negative
geotaxis and phototaxis that yield similar surface aggregations are not known.
From a mechanistic point of view, phototaxis may be sensitive to the structural
details in the photoreceptor since the physical structure of the apparatus is known
to vary widely (Levandowsky and Kaneta, 1987; Kreimer, 1994). Geotaxis is not
well studied in flagellates (Levandowsky and Kaneta, 1987), but Lebert and
Hader (1996) recently examined Euglena gracilis which exhibits a spiral swimming motion (Jennings, 1976) like dinoflagellates (Kamykowski etal., 1992). They
suggested that the cell surface deformations caused by the movement of the cell
through the water due to gravitational effects activate stretch-sensitive ion channels. The change in membrane potential resulting from differential stretching of
the cell membrane is considered the primary mechanism for sensing gravity. They
also suggest that the location of these channels on the cell body can vary. These
sensor differences may contribute to the various expressions of geotaxis and
phototaxis in different species.
Accumulating evidence (Cullen, 1985; Maclntyre et al, 1997; Kamykowski et
al., 1998) also suggests that an individual species may change geotaxis and phototaxis precision and sign as cell state changes. For example, Kamykowski et al.
(1998) reported that negative geotaxis precision was inversely correlated with
cellular lipid content, while other biochemical pools like carbohydrate, chlorophyll a, DNA, protein and RNA also changed over the daily cycle, apparently
uncorrelated with measured taxes. Other reported short time scale biochemical
changes include diel variations in nucleic acids (Berdalet et al., 1992), pigment
composition (Latasa et al., 1992) and alkaline phosphatase activity (Rivkin and
Swift, 1979), and more rapid variations in pigment composition (Demers et al.,
1991; Lesser, 1996) that can occur in association with variations in photosynthetic
potential (Prezelin and Matlick, 1980) that modulate carbon flux. Cellular mass
balances may influence mechanical/structural characteristics of a cell and thus
affect how a given sensory array functions. Coupling between cell biochemical
state and sensory activity may be an important link in the mechanism that
supports species-specific growth optimization in dinoflagellates (Kamykowski,
1995).
Although the data are sparse and not contemporaneous, the different taxis
patterns of precision and sign observed in this paper appear related to the timing
of surface aggregation. Appearance after daylight at the surface follows the order:
P.faeroense, G.breve, H.illdefina and P.micans. Disappearance from the surface
follows the order: P.micans, G.breve, H.illdefina and P.faeroense. As previously
reported, continuous positive phototaxis corresponds to P.faeroense's long stay at
the surface and strong negative phototaxis in the morning corresponds to
P.micans' late arrival at the surface. Consider the possible interaction among the
1794
Dinoflagellate taxis and migration
behavioral types that these species represent if they co-exist in the same water
column. Species-specific, taxis-mediated migration patterns can provide a
mechanism by which environmental resources like light and nutrients, as previously discussed by Cullen (1985) and Maclntyre et al. (1997), are partitioned to
reduce competition. If a specific temporal sequence of environmental conditions
favors one of the behavioral types, then that most successful migration pattern,
combined with an effective net growth rate, probably contributes to eventual
bloom formation. Many field studies of dinoflagellates focus on a single species
because such blooms are easier to study and the dominating species may be
harmful. However, an understanding of bloom initiation as related to cell growth
and division may benefit from comparative studies of how different phytoplankton species function in the same water column at the same time and of how
the different behaviors affect the growth potential of each species.
Acknowledgements
Hidekatsu Yamazaki and Atsuko Keiyu Yamazaki contributed to the theory that
stimulated these laboratory experiments. The culture of H.illdefina was obtained
from the Guillard-Provasoli CCMP, while Gary J.Kirkpatrick (Mote Marine
Laboratory) provided the culture of G.breve. This work was supported by NSF
grant OCE-9503253.
References
Berdalet.E., Latasa,M. and Estradajvl. (1992) Variations in biochemical parameters of Heterocapsa
sp. and Olithodiscus luieus grown in 12:12 light:dark cycles. I. Cell cycle and nucleic acid composition. Hydrobiologia, 238,139-147.
Berman.T. and Rodhe.W. (1971) Distribution and migration of Pendinium in Lake Kinneret. Mitt.
Int. Ver. Limnoi, 19, 266-276.
Blasco.D. (1978) Observations on the diel vertical migration of marine dinoflagellates off the Baja
California coast. Mar. Biol., 46,41-47.
Cullen J.J. (198S) Diel vertical migration by dinoflagellates: roles of carbohydrate metabolism and
behavioral flexibility. In Rankin,M.A. (ed.). Migration: Mechanisms and Adaptive Significance.
COM. Mar. ScL, 27,135-152.
Cullen J.]. and Horrigan.S.G. (1981) The effect of nitrate on the diumal vertical migration, carbon to
nitrogen ratio and the photosynthetic capacity of a dinoflagellate, Gymnodinium splendens. Mar.
Biol., 62,81-89.
Demers.S., Roy,S., Gagnon.R. and Vignault,C. (1991) Rapid light induced changes in cell fluorescence
and in xanthophyll-cycle pigments of Alexandrium excavatum (Dinophyceae) and Thalassiosira
pseudonana (Bacillariophyceae): a photoprotection mechanism. Mar. Ecol. Prog. Ser., 76,185-193.
Eggersdorfer3- and Hader,D.-P. (1991a) Phototaxis, gravitaxis and vertical migrations in the marine
dinoflagellates, Peridinium faeroense and Amphidinium caterea. Acta Protozool., 30, 63-71.
Eggersdorfer,B. and Hader,D.-P. (1991b) Phototaxis, gravitaxis and vertical migrations in the marine
dinoflagellate, Prorocentrum micans. FEMS Microb. Ecol., 85, 319-326.
Eppley,R.W., Holm-Hansen.O. and StricklandJ.D.H. (1968) Some observations on the vertical
migration of dinoflagellates. /. PhycoL, 4, 333-340.
Forward,R.B.rIr (1974) Phototaxis by the dinoflagellate Gymnodinium splendens Lebour. J.
Protozool., 21, 312-315.
Forward,R.B.,Jr (1976) Light and diurnal vertical migration: Photobehavior and photophysiology of
plankton. In Smith,K.C. (ed.), Photochemical and Photobiological Reviews Vol. 1. Plenum, New
York, pp. 157-208.
Guillard.R.R.L. and HargravesJ".E. (1993) Stichochrysis immobilis is a diatom, not a chrysophyte.
Phycologia, 32, 234-236.
1795
D.KamykowsIri, EJJVfffligan and R.E-Reed
Hand.W.G. (1970) Phototactic orientation by the marine dinoflagellate Gyrodinium dorsum Kofoid.
I. A mechanism model. /. Exp. Zool, 174,33-38.
Hand.W.G. and SchmidtJ.A. (1975) Phototactic orientation by the marine dinoflagellate Gyrodinium
dorsum Kofoid. II. Flagellar activity and overall response mechanism. / ProtozooL, 22, 494-498.
Hasle,G.R. (1950) Phototactic migration in marine dinoflagellates. Oikos, 2,162-175.
Hasle.G.R. (1954) More on phototactic diurnal vertical migration in marine dinoflagellates. Nytt.
Mag. BoU, 2,139-147.
Heaney.S.I. and Eppley.R.W. (1981) Light, temperature and nitrogen as interacting factors affecting
diel vertical migrations of dinoflagellates in culture. / Plankton Res., 3, 331-344.
Heaney.S.I. and Furnass.T.I. (1980) Laboratory models of diel vertical migration in the dinoflagellate
Ceratium hirundinella. Freshwater Biol, 10,163-170.
Heaney.S.I. and TallingJ.F. (1980) Dynamic aspects of dinoflagellate distribution patterns in a small
lake./ EcoL, 68,75-94.
JenningsJH-S. (1976) Behavior of Lower Organisms. Indiana University Press, Bloomington, IN, 366
pp.
KamykowskiJ). (1980) Subthermocline maximums of the dinoflagellates Gymnodinium simplex
(Lohmann) Kofoid and Swezy and Gonyaulaxpolygramma Stein. Northeast GulfSci., 4, 39-43.
Kamykowski.D. (1981) Laboratory experiments on the diurnal vertical migration of marine dinoflagellates through temperature gradients. Mar. Biol., 62, 57-64.
Kamykowski,D. (1995) Trajectories of autotrophic marine dinoflagellates. /. Phycol., 31,200-208.
Kamykowski.D. and Zentara,S.-J. (1977) The diurnal vertical migration of motile phytoplankton
through temperature gradients. Limnol. Oceanogr., 22,148-151.
KamykowskiJ}., McCollum.S.A. and Kirkpatrick.GJ. (1988) Observations and a model concerning
the translational velocity of a photosynthetic marine dinoflagellate under variable environmental
conditions. Limnol. Oceanogr., 33, 66-78.
Kamykowski.D., Reed,R.E. and Kirkpatrick.GJ. (1992) Comparison of sinking velocity, swimming
velocity, rotation and path characteristics among six marine dinoflagellate species. Mar. Biol, 113,
319-328.
Kamykowski.D., Milligan,EJ. and Reed,R.E. (1998) Biochemical relationships with the orientation
of the autotrophic dinoflagellate Gymnodinium breve under nutrient-saturated conditions. Mar.
EcoL Prog. Ser, in press.
Kreimer.G. (1994) Cell biology of phototaxis in flagellate algae. Int. Rev. Cytol., 148, 229-310.
Latasa.M., Berdalet,E. and Estrada.M. (1992) Variations in biochemical parameters of Heterocapsa
sp. and Olithodiscus luteus grown in 12:12 light:dark cycles. II. Changes in pigment composition.
Hydrobiologia, 238,149-157.
Lebert,M. and Hader,D.-P. (1996) How Euglena tells up from down. Nature, 379, 590.
Lesser,M. (1996) Acclimation of phytoplankton to UV-B radiation: oxidative stress and photoinhibition of photosynthesis are not prevented by UV-absorbing compounds in the dinoflagellate Prorocentrum micans. Mar. EcoL Prog. Ser., 132, 287-297.
Levandowsky,M. and Kaneta,P. (1987) Behavior in dinoflagellates. In Taylor,FJ.R. (ed.), The Biology
of Dinoflagellates. Blackwell Scientific, Oxford, pp. 360-397.
MacIntyreJ.G., Cullen JJ. and Cembella,A.D. (1997) Vertical migration, nutrition and toxicity in the
dinoflagellate Alexandrium tamarense. Mar. EcoL Prog. Ser., 148,201-216.
Prezelin3- and Matlick.H. (1980) Time-course of photoadaptation in the photosynthesis-irradiance
relationship of a dinoflagellate exhibiting photosynthetic periodicity. Mar. Biol., 58, 85-%.
Riley.G.A. (1956) Oceanography of Long Island Sound, 1952-54. II. Physical oceanography. Bull.
Bingham Oceanogr. Coll, 15,15-46.
Rivkin.R. and Swift.E. (1979) Diel and vertical patterns of alkaline phosphatase activity in the oceanic
dinoflagellate Pyrocystis noctiluca. LimnoL Oceanogr, 24,107-116.
Staker.R.D. and Bruno.S.F. (1980) Diurnal vertical migration in marine phytoplankton. Bot. Mar., 23,
167-172.
Weiler.C.S. and Karl,D.M. (1979) Diel changes in phase-dividing cultures of Ceratium furca (Dinophyceae): nucleotide triphosphates, adenylate energy charge, cell carbon and patterns of vertical
migration. /. Phycol., 15, 384-391.
Received on December 16, 1997; accepted on May 13, 1998
1796