LIGHT-INDUCED MOTILE RESPONSES OF THE ESTUARINE BENTHIC DIATOMS NAVICULA PERMINUTA AND CYLINDROTHECA CLOSTERIUM (BACILLARIOPHYCEAE)¹

Microphytobenthic algal biofilms are widely distributed in freshwater, estuarine, and marine habitats, wherever sufficient light reaches the sediment surface (Miller et al. 1996, Underwood and Kromkamp 1999). In clear-water environments, microphytobenthos (MPB) can occur down to 30 m depth, suggesting that they play a major role in coastal biogeochemical processes (Underwood and Kromkamp 1999, Thornton et al. 2007). A key property exhibited by the majority of MPB, and particularly epipelic diatoms (mainly mono- and biraphid genera), is rapid motility in response to a range of stimuli (Cohn and Disparti 1994, Waring et al. 2007). Cell motility within the sediment can be sufficiently synchronized to produce “vertical migrations” of whole populations, such that the surface of intertidal mudflats undergoes diurnal changes in surface biomass, commonly identified by the appearance or disappearance of the golden brown color (Aleem 1950, Perkins 1960, Hopkins 1963, Round and Palmer 1966, Happey-Wood and Jones 1988). The involvement of tidal cycles and endogenous rhythms of motility in producing biomass changes was recently reviewed (Consalvey et al. 2004), but MPB biofilms will always respond to light changes at any point in the tidal cycle (Aleem 1950, Perkins et al. 2001, 2002, Sauer et al. 2002); thus, light may be an overriding stimulus. Motility in response to light can endow physiological advantages to organisms, such as maximizing photosynthetic efficiency and rates of cell reproduction (Senger and Schmidt 1994, Underwood et al. 2005), and many types of motile microalgae have been observed to accumulate in areas irradiated by light of moderate fluence rates, favorable for photosynthesis, but disperse from areas of high fluence rates that could cause photodamage (Nultsch 1991, Paterson et al. 1998, Kingston 1999, Underwood 2002, Underwood et al. 2005, Waring et al. 2007).

Diatom movements involve a series of forward and reverse gliding motions, with progression of the diatom across the substratum resulting when the glide in one direction lasts longer than the other (Round 1971, Apoya-Horton et al. 2006). If no external factors act unilaterally, the direction of movement is random (Nultsch 1974). The two most common motile responses to light in microorganisms are phototaxis (oriented locomotion with respect to the light source) and photokinesis (speed of locomotion is affected by the photon fluence rate). Experiments examining reversal rhythms in individual diatom cells have led to conclusions that phototaxis in diatoms is the result of a series of photophobic responses whereby the time between direction changes is prolonged if the cell is moving in a beneficial direction, and curtailed if the cell is moving in a unfavorable direction (Nultsch and Wenderoth 1973, Cohn et al. 2004). This type of phototaxis, characterized by changes in the autonomous rhythm of reversal, contrasts sharply with the active steering exhibited by many flagellates (Nultsch 1985). In photokinesis experiments, Nultsch (1971) found that Nitzschia communis was motile in the dark and that its speed increased in white light, with a maximum speed at 500 lux (∼6–9 μmol · m⁻² · s⁻¹). Further increases in light intensity caused decreases in speed. Other types of movement, such as rotating around the cell center and pivoting around the cell pole, have been described (Nultsch 1956), and some nongliding movements have been linked to environmental stresses (Harper 1977, Apoya-Horton et al. 2006).

Plants and algae contain sophisticated networks of photoreceptors and sensory pathways to exploit light as a sensory stimulus (Hegemann et al. 2001). Three main types of photoreceptors have been identified in higher plants. The phytochromes are primarily responsible for the detection of far-red light and red light, and there are two types of blue-light receptors—cryptochromes and phototropins. In addition, rhodopsins have been identified in algae but not plants. The sequenced genome of the centric diatom Thalassiosira pseudonana has revealed putative homologues of cryptochromes and phytochrome (Armbrust et al. 2004), but no obvious matches to phototropins or rhodopsins were identified.

Previous work on light responses and motility in benthic diatoms has been on intact biofilms or isolated single cells. It has not been shown whether the individual responses in vitro are responsible for population-level responses similar to those seen in the natural environment. The present study seeks to define the light regimes that lead to accumulation and dispersion of diatom populations, the degree to which wavelength and fluence rate contribute to motility, and the predominant forms of motility that occur in response to different light conditions.

Materials and methods

Results

Photoaccumulation and photodispersal. N. perminuta showed significant accumulation in white light in the range 88–302 μmol · m⁻² · s⁻¹ (P < 0.05) and significant dispersal from areas of 3–41 μmol · m⁻² · s⁻¹ (P < 0.05). C. closterium showed significant accumulation in the range 43–367 μmol · m⁻² · s⁻¹ and significant dispersal from an area with PFDs of 15 μmol · m⁻² · s⁻¹ (P < 0.05; Fig. 1). The midpoints of the accumulation ranges for both species were similar (205 μmol · m⁻² · s⁻¹ for N. perminuta and 213 μmol · m⁻² · s⁻¹ for C. closterium). The magnitudes of the photoaccumulation were significantly different for the two species (P < 0.05), with N. perminuta showing a near 4-fold increase in the range 100–200 μmol · m⁻² · s⁻¹, whereas C. closterium showed a near 2-fold increase.

Photokinesis. When compared with movement under background illumination alone, actively moving N. perminuta cells exhibited significantly higher average speed at PFDs lower than 310 μmol · m⁻² · s⁻¹ (P < 0.05; Fig. 2). At PFDs of 310 μmol · m⁻² · s⁻¹ and above, average population speed was not significantly different from that under the background illumination. C. closterium exhibited significantly higher average cell speed at 15 and 56 μmol · m⁻² · s⁻¹ than under background illumination alone (P < 0.05). At all other PFDs, speed was not significantly different from that under background illumination (Fig. 2). At all PFDs, N. perminuta exhibited an average speed that was significantly faster than C. closterium (P < 0.05 for all).

Phototaxis. N. perminuta exhibited strong, highly reproducible directed movement in unidirectional light that corresponded well with the patterns of photoaccumulation and photodispersal. At low PFDs (≤56 μmol · m⁻² · s⁻¹), the majority of the N. perminuta population moved toward the light, and at high PFDs (≥310 μmol · m⁻² · s⁻¹) the majority of the population moved away from the light (3, 4). The maximum population movement toward the light occurred at the lowest irradiance used (15 μmol · m⁻² · s⁻¹), with maximum population movement away from the light at 620 μmol · m⁻² · s⁻¹. A chi-squared goodness-of-fit test, against a random distribution of 50% of the population showing net movement toward the light, showed significant departure from this pattern at all PFDs (P < 0.05) except 112 and 186 μmol · m⁻² · s⁻¹. These latter photon irradiance conditions correspond to those at which N. perminuta showed maximum accumulation of biomass (Fig. 1). C. closterium did not appear to show directed movement at all (3, 4), verified by the chi-squared test, which showed no significant difference from a 50/50 distribution at any PFD (P > 0.05 for all).

Closer examination of all actively moving cells from three replicate N. perminuta populations showed that, at 15 μmol · m⁻² · s⁻¹, the populations all had the same mean angle of travel (P > 0.05) and that this was not significantly different from 0° (V-Test, P < 0.05); that is, cells were moving toward the light (Fig. 4). Similarly, the replicates at 620 μmol · m⁻² · s⁻¹ also all had the same mean angle of travel (P > 0.05), and this was not significantly different from 180° (P < 0.05); that is, cells were moving away from the light.

Three populations of C. closterium showed no significant difference between their distributions (direction of motility; P > 0.05), and their mean angle of travel did not correspond to 0° (P > 0.05) under 15 μmol · m⁻² · s⁻¹ illumination. At 620 μmol · m⁻² · s⁻¹, there was no significant difference between the circular distributions (P > 0.05), and the mean angle of travel did not correspond to 180° (P > 0.05; Fig. 4).

The phototactic behavior of N. perminuta was further examined by correlating the net distance traveled over 30 s against the angle of travel for each N. perminuta cell at both PFDs (P < 0.05 for both). Cells moving away from the light traveled significantly farther than cells moving toward the light at 620 μmol · m⁻² · s⁻¹, with the opposite true at 15 μmol · m⁻² · s⁻¹ (Fig. 5, A and C). At 15 μmol · m⁻² · s⁻¹, cells traveling at 0, 45, and 315° showed no significant difference in the net distance traveled, giving a mean distance covered in 30 s of 161.7 μm ± 0.6 for cells moving toward the light. At 620 μmol · m⁻² · s⁻¹, cells traveling at 180°, 135°, and 225° showed no significant difference in the net distance traveled, giving a mean distance covered in 30 s of 90.8 μm ± 3.7 for cells moving away from the light. There was no correlation between net distance traveled and angle of travel for C. closterium cells at either PFD (P > 0.05; Fig. 5, B and D).

Action spectra of responses. Over the range of fluence rates and wavelengths tested, C. closterium showed the highest average cell speed at 10 μmol · m⁻² · s⁻¹ and 686 nm (2.6 μm · s⁻¹). Average speed of N. perminuta cells was also the greatest at λ of 686 nm, where 20 μmol · m⁻² · s⁻¹ produced speeds of 6.7 μm · s⁻¹ (data not shown). Action spectra of the kinetic responses showed that λ of 645 nm also induced photokinetic responses in both species and that wavelengths between 500 and 550 nm may play a role in the photokinetic response of N. perminuta (Fig. 6, A and B).

N. perminuta showed significant movement toward the light source at 430, 466, 490, and 510 nm (chi-squared goodness-of-fit against a 50/50 distribution, P < 0.05; data not shown). Wavelengths above 555 nm did not produce a directional response at any of the PFDs tested (Fig. 6C). The greatest percentage movement toward the light source occurred with light of wavelength 430 nm and fluence rate 6.5 μmol · m⁻² · s⁻¹ (73%). C. closterium showed no significant movement toward or away from the light source at any of the wavelengths and intensities tested (not shown).

Discussion

N. perminuta and C. closterium both exhibited photoaccumulation at ∼200 μmol · m⁻² · s⁻¹ and photodispersal from PFDs of ∼15 μmol · m⁻² · s⁻¹. Photokinesis contributed toward photodispersal for both species, and red light was most effective at producing the high speeds that were associated with this photodispersal. Only N. perminuta was able to actively move up and down a light gradient, and only wavelengths of light from 430 to 510 nm elicited this directional response. The fluence rates that produced maximum accumulation in N. perminuta and C. closterium cultures correspond well with those reported to produce maximum accumulation of natural assemblages in the field (Perkins et al. 2002, Serôdio et al. 2006), despite the fact that these cells had been maintained in laboratory culture for several years.

Both species exhibited negative photokinesis—a significant reduction in speed as PFD increased (in the range 15–112 μmol · m⁻² · s⁻¹), which would allow dispersal from lower light intensities. There was no evidence that motility increased again at higher PFDs in either species; therefore, photokinesis was not used as a mechanism for avoiding deleteriously high fluence rates. It is worth noting that horizontal movement, as seen here, can be an order of magnitude faster than the predominately vertical movements seen in natural sediments (Hay et al. 1993).

Observations that estuarine diatoms always disappear when darkened (Aleem 1950, Sauer et al. 2002, Consalvey et al. 2004) may be explained by negative photokinesis (increasing speed as fluence rates decrease) and suggest that light stimuli can override circadian rhythms. The only previous detailed work on photokinesis in a diatom (N. communis) observed maximum speeds at similar fluence rates, and that red wavelengths (670 nm max.) were highly effective at stimulating speed increases (Nultsch 1971). A later study largely reported no relationship between speed and wavelength, but this could be due to differences in experimental design (Cohn and Weitzell 1996). We measured speed after a set period of acclimation to the new light regime (40 min), which we determined was necessary for the full response to be elicited. In the present work, red light was much more effective than any other wavelength at altering cell speed. Although this enhancement by red light is consistent with the recent identification of phytochrome in the diatoms T. pseudonana and Phaeodactylum tricornutum (Armbrust et al. 2004, Montsant et al. 2005), current data do not allow us to distinguish between a possible role for phytochrome or enhancement of photosynthesis in this response.

N. perminuta exhibited significant phototaxis, both positive and negative, that contributed to the observed patterns of accumulation and dispersal. Populations of N. perminuta moving in a light gradient traveled significantly farther (in a 30 s period, Fig. 5) when moving toward the 100–200 μmol · m⁻² · s⁻¹ range, irrespective of whether they were moving up or down the light gradient. This “distance traveled” bias could result from cells prolonging their travel times when moving in beneficial directions and curtailing them when moving in detrimental conditions, as has been previously reported at the single cell level (Nultsch and Wenderoth 1973, Cohn et al. 2004). Mean distance traveled when moving in a beneficial direction was not significantly different whether the direction was directly in line with the gradient or at a 45° angle and so equates phototaxis in N. perminuta with the biased random walk seen in bacterial chemotaxis rather than the steering mechanisms exhibited by some flagellate algae (Nultsch 1985). This finding implies that the diatoms sensed and responded to the light gradient and did not orientate with respect to the light source.

Only blue light between 430 and 510 nm elicited directed movement in N. perminuta populations (phototaxis), consistent with previous demonstrations of blue light involvement in directed movement in single diatom cells (Cohn and Weitzell 1996, Cohn et al. 1999) and positive taxis in diatom populations (Nultsch 1971). In the current work, 430 nm was most effective at producing positive taxis, while all wavelengths >412 nm to <555 nm induced some activity. Negative movements were also observed when the overall light intensity was increased, suggesting that the same or similar photoreceptors were being used for positive and negative taxis. While it is only feasible to speculate on the photoreceptor(s) involved in the phototactic response, it is noteworthy that centric diatoms have recently been shown to possess a cryptochrome homolog (Armbrust et al. 2004).

Surprisingly, no phototactic response was shown by C. closterium at any light intensity. It is not known how widespread the ability to move in a directed manner is in the natural environment; however, these experiments have shown that phototaxis is not an essential requirement for photoaccumulation in benthic diatoms. It is worth also noting that, while extremely common in muddy and sandy sediment biofilms, C. closterium is not exclusively benthic (often referred to as tychoplanktonic) and frequently occurs in the plankton as well. It could be that only “true benthic” diatoms have the ability to respond to gradients and that actively moving, but not exclusively benthic, diatoms may only possess the ability to respond kinetically to their environment.

This work sought to investigate in vitro the motile responses of unialgal diatom populations to light stimuli in an effort to identify the contribution of photomotile responses to the observed rhythms of vertical migration. If the sediment surface is exposed by the tide before dawn, then the number of diatoms on the sediment surface will gradually increase as the light intensity increases (Aleem 1950, Perkins 1960, Hopkins 1963, Round and Palmer 1966, Happey-Wood and Jones 1988, Hay et al. 1993). To understand the light cues that a diatom population at depth in the sediment may be receiving, the spectral composition as well as the absolute light intensity needs to be considered. Light of 495 nm has a much lower extinction coefficient than any other wavelength when traveling through clean sediment (50% mud/silt, 50% fine sand; Perkins 1963). We have shown here that this wavelength of light can be used for phototactic responses by N. perminuta. Red wavelengths beyond 600 nm had the next lowest extinction coefficients, and these wavelengths can cause an increase in average population speed in both species. These increases in speed could allow both species to move from within the sediment, toward the surface, by photokinesis, but the response of N. perminuta would also be augmented by a blue-light-mediated phototactic response.

There have also been observations of diatoms moving down into the sediment as the light intensity decreased at the end of the day (Aleem 1950, Perkins 1960, Hopkins 1963, Round and Palmer 1966, Happey-Wood and Jones 1988, Hay et al. 1993). This finding gave rise to suggestions that cells undergo daily rhythms in their sensitivity to a variety of stimuli (Palmer and Round 1965, Round and Palmer 1966, Happey-Wood and Jones 1988, Sauer et al. 2002), or that either chemotaxis (Happey-Wood and Jones 1988) or geotaxis (Round and Happey 1965) may provide the stimulus to initiate downward migration. However, our observations on diatom motility in response to light can explain much of this observed phenomenon. Light extinction coefficients increase with decreasing illumination, and at low solar altitude, (<30°) albedo is high. Perkins (1963) proposed that the tendency to sink into sediment before sunset can be explained in terms of the sediment reflection and light attenuation, and that the effect of albedo alone would ensure that diatoms submerged a short period before sunset. As solar elevation decreases at sunset, not only does the light intensity decrease, but also there is a change in spectral composition. The pathlength through the atmosphere increases, the effects of scattering and absorption are enhanced, shorter wavelengths are preferentially attenuated, and atmospheric refraction will cause longer wavelengths to be enhanced (Smith and Morgan 1981). Therefore, a diatom population on the sediment surface is subject to low-intensity, red-enriched light. Both N. perminuta and C. closterium exhibited the highest speeds at low light intensity and with red light. Hence the light climate at dusk is such that diatoms on the sediment surface would show increased motility as the light decreases. For a species such as C. closterium, which exhibits only photokinesis, this change in light regime would be enough to cause photodispersal, which at a macroscopic level would appear to be a decrease in surface biomass. It may take longer for a species such as N. perminuta, which exhibits a tactic response to blue light, to show an appreciable decrease in surface biomass. The 430–510 nm component of the light would have to decrease to such a level that the cells could no longer detect the light gradient and therefore are no longer held at the surface. Perkins (1963) reported that at low solar altitude, the albedo of the sediments was high. This effect means there would be poor penetration of light into the sediment; hence, cells that had left the surface would have minimal light cues to induce them to migrate upward. Interestingly, Underwood et al. (2005) have recently identified not only migrational responses to diel light changes but also photophysiological ones. These responses were species-specific, and it was suggested that they could be a form of niche separation. However, alterations in physiology could also affect behavioral responses and may give some support to the idea of rhythms in sensitivity. It is also unclear what effects light scattering from mineral grains and absorption by other phototrophs (Jørgensen and Des Marais 1986, 1988, Lassen et al. 1992) would have on the extrapolation of results from horizontal laboratory experiments to intact biofilms.

Observations that estuarine diatoms always disappear when darkened (Aleem 1950, Sauer et al. 2002, Consalvey et al. 2004, Jesus et al. 2006) suggest that endogenous rhythms play only a minor part in vertical migrations, and that the driving force is still the prevailing light climate. However, these rhythms are present, persist for many days in the laboratory (Round and Palmer 1966, Happey-Wood and Jones 1988), and may be responsible for observations of “anticipation” of tidal inundation (Palmer and Round 1965, Serôdio and Catarino 2000). As tidal inundation can occur once or twice during daylight, or not at all, and timings vary on a daily basis, there would need to be entrainment—the synchronization of the circadian clock with the external environment. The experiments conducted here on N. perminuta and C. closterium implicated blue- and red-light photoreceptors in photomotile responses. This finding suggests that cryptochrome and phytochrome may be involved, and both these receptors have been shown to affect circadian rhythms in Arabidopsis (Salomé and McClung 2005) and so are possible contenders for entraining circadian rhythms in diatoms. However, this aspect of the cellular control of diatom behavior remains largely unresearched.