| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
School of Marine Sciences and Technology, Ridley Building, University of Newcastle upon Tyne, NE1 7RU, United Kingdom
* To whom correspondence should be addressed. E-mail: k.s.last{at}ncl.ac.uk
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Photoperiodism in marine animals is less intensively studied than in terrestrial organisms, but "photoperiod-like" phenomena have been documented in several marine species, including the echinoderm Picaster ochraceus (Pearse and Eernisse, 1982), the copepod Labidocera aestiva (Marcus, 1986), and the polychaetes Neanthes limnicola (Fong and Pearce, 1992a, b) and Harmothoe imbicata (Garwood and Olive, 1982). The difference between the environmental and the evolutionary history in marine organisms requires that these processes be investigated to understand environmental signal transduction in the control of life-history events and to determine whether these observations are likely to be the consequence of a common clock mechanism. Confirmation that they are such a consequence would have important implications for our understanding of the evolution of clock-based processes.
To test the suggestion that an endogenous long-term rhythm moderates responses to fixed photoperiodic inputs in N. virens, we used an actographic procedure (Last, 2003) to monitor its emergence (foraging) behavior and examined the spontaneous diel activity patterns of individual animals maintained under fixed LD cycles for several months.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Two time periods were used for the short-term assay: six days in September in the autumn of 1998 (A98) and six days in February in the winter of 1999 (W99). Both A98 and W99 animals had been born from the same broodstock (16 March 1998), and as a consequence, the experimental animals for W99 were 5 months older than those used for A98. To prevent any size- or maturation-related bias in feeding rates, all animals chosen were of about the same weight, 2.5 ± 0.1 g, with all coelomic oocytes having a diameter 
120 µm. All animals (n = 16) were collected from a commercial supplier (Seabait Ltd.) a week prior to the experiments, which took place under controlled conditions in the laboratory. Since birth, animals had been maintained outside under natural photoperiods, so the natural photoperiod at the time of collection was LD 13:11 for A98 and LD 11:13 for W99. After acclimatization, animals were introduced into the actographs (see below) and maintained under one of two photoperiods, LD 16:8 or LD 8:16, that approximated the photoperiodic extremes between midsummer and midwinter at a latitude of 55° north (where the animals were collected).
Data obtained from the actographic recording were analyzed for two factors: (1) variability in the diel activity of individuals both between photoperiodic treatments and seasons, and (2) overall changes in "strength" of rhythm. Variability in diel activity was visually represented using actograms, while the strength of rhythm was characterized with autocorrelation analysis (Dowse and Ringo, 1989; Palmer et al., 1994; Dutilleul, 1995). This statistic provides a measure of the strength to the rhythm that is not affected by the overall level of activity: the less noisy the data between cycles and the higher the signal, the stronger the rhythm. Since the autocorrelation output is normally distributed around the lag time that most closely matches the period of that particular rhythm, robust parametric statistics can be applied.
The concurrent long-term experiment was carried out for 9 months under constant LD 16:8 and LD 8:16 photoperiods, using animals from the same broodstock that were born naturally at Seabait Ltd. on 16 March 1998 and reared continuously outside under natural photoperiods until the start of these investigations. As with experiments A98 and W99, the animals chosen had about the same weight (2.5 ± 0.1 g) and sexual maturity (oocyte diameter 120 µm). All stock animals (n = 100) for this experiment were maintained continuously under photoperiods of LD 16:8 or LD 8:16. Such continuous rectilinear photoperiods have previously been used to detect circannual rhythms (Randall et al., 1998; Nisimura and Numata, 2001) without the possibly detrimental effects of inducing a free-running state by using continuous light or darkness. Each month between September 1998 and June 1999, four animals were selected at random from stock aquaria in the laboratory and placed into the artificial burrows of the actograph. Their activity was then recorded for 6 to 7 days under the two photoperiodic treatments, LD 16:8 and LD 8:16. At other times, the animals were maintained under the respective photoperiodic treatments LD 16:8 and LD 8:16 in stock aquaria. Hourly actograph data were summed into 48-h epochs and partitioned into mean monthly actograms. The mean data points were plotted against maximum overall activity that normalized the amplitude between treatments and made direct comparisons possible.
To prevent transient behavioral activity for both short (A98 and W99) and long-term assays, animals were always acclimatized for 7 days under their new photoregimes.
The actograph consisted of two aquaria housed in light-tight photoperiodically controlled chambers. Artificial burrows consisting of lengths of PVC tubing were inserted through the bottoms of these experimental aquaria and connected to infra-red optocouplers that were broken as soon as any animal emerged. Out-of-burrow activity for all animals was monitored using a data logger. A complete description of the actographic procedures has been published in Last (2003). The aquaria were supplied with recirculating seawater, salinity 34 to 36
, that had been biologically trickle-filtered and sterilized with UV light; temperatures in the aquaria were maintained constant at 
16 °C. Every 2 days during the middle of the photophase (the light portion of the cycle) the experimental animals were fed coarse trout pellets to and above the maximum that they would consume.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
Analysis of locomotor activity of N. virens: Winter (W99)
Actograms.
The mean hourly activity (beam breaks) of all animals over 6 days in the winter when maintained under LD 16:8 and LD 8:16 photoperiods is shown in Figure 2. Note that these animals are from natural photoperiods of LD 11:13. Under both photoperiods, activity was consistently high for each successive scotophase despite the relatively high within-treatment variability between individuals, as shown by the standard deviation bars. There was no significant difference (t test: t = 2.09, P > 0.05) in the number of emergence events for animals under LD 16:8 and animals under LD 8:16 for each consecutive 24-h period. This is in striking contrast to the results recorded 5 months earlier under the same photoperiodic treatments.
|
Comparative analysis in strength of rhythm: Autumn and winter
Autocorrelation analysis of rhythm strength is not affected by overall amplitude of activity, and hence a comparative analysis of activity between treatments (LD 16:8 and LD 8:16) and between seasons (A98 and W99) was possible even though overall activity in A98 was higher than in W99. The strength of rhythm of animals maintained under LD 16:8 in the autumn was significantly (t test: t = 2.54, P < 0.05) stronger than that of animals under LD 8:16 at this time. No significant difference (t test: t = 0.27, P > 0.1) was observed in strength of rhythm in animals maintained under a photoperiod of LD 16:8 in the autumn compared to animals maintained under the same photoperiod in the winter. Similarly, no significant difference (t test: t = 0.65, P > 0.1) was observed in strength of rhythm in animals maintained under LD 16:8 or LD 8:16 photoperiods in the winter. Finally the strength of rhythm of animals maintained under the LD 8:16 photoperiod in the winter was significantly (t test: t = 3.29, P < 0.01) stronger than that of animals under the LD 8:16 photoperiod in the autumn. In the autumn, the most noticeable between-treatment variation was the much reduced overall activity and strength of rhythm under LD 8:16 compared to LD 16:8. In the winter, no such differences in strength were observed since the strength of rhythm was high in all animals irrespective of photoperiod.
Analysis of long-term locomotor activity of N. virens
The results from each experimental treatment (LD 16:8 and LD 8:16) were pooled to plot a three-dimensional (3D) topographical graph of activity (y-axis) over time (x-axis) and month (secondary x-axis) (Fig. 3a, b). Dashed rectangles beneath the x-axis represent the times of the imposed scotophase. The surface topography represents the mean hourly activity of four animals over the duration of 6 to 8 days every month for 9 months. (Note: these results do not include data from the short-term A98 and W99 experiments).
|
Figure 3b shows the pattern of locomotor activity under LD 8:16. Between September and December there was no activity in either the photophase or the scotophase. Nocturnal activity increased dramatically after December, and then continued until termination of the experiment in June. This elevated activity occurred spontaneously and independently of any environmental signal. The major peak of activity was generally seen directly after lights-off, with raised activity during the rest of the scotophase. Activity during the photophase was low. A perturbation was observed in March, when activity dropped significantly for the duration of the month. This is represented by the trough running the length of the graph from left to right.
The results from the topographical data for LD 16:8 and LD 8:16 are summarized schematically in relation to the natural photoperiod in Figure 4. At the start of the experiment in September, the natural photoperiod was LD 13:11. All animals placed under the LD 16:8 photoperiod (Fig. 4a) showed strong, nocturnal activity that continued until March, when the natural photoperiod was once again LD 12:12. At this time, additional activity was noted during the later part of the photophase. After March, activity was once again restricted solely to the scotophase. Animals placed under the LD 8:16 photoperiod, however, displayed a quite different activity pattern (Fig. 4b). In these animals, all activity was relatively low for 3 months, during the time when natural photophases are diminishing in the natural habitat (LD <12:>12). By the middle of December, the time when the natural scotophase is at its longest at this latitude (17 h), a spontaneous recovery of activity during the scotophase was observed. This pattern of elevated nocturnal activity continued in the laboratory at a time when the nights in the natural environment were gradually shortening. At the spring equinox in March, when natural photoperiods were once again LD 12:12, overall activity dropped substantially. From April to June, however, activity was once again elevated.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Our results also show that the activity of this polychaete is predominantly nocturnal under all photoperiods, irrespective of season. The photophase suppressed activity in animals under both photoperiodic treatments, and we believe that spurious beam breaks during the photophase constitute primarily "within-burrow" rather than "out-of-burrow" activity (Last, 2003). We hypothesize that the conserved nature of this nocturnal response reflects a strong selective advantage as a predator-avoidance mechanism, since feeding excursions for N. virens are associated with predation risk from shorebirds and pleuronectenoid fish, which are all well-known diurnal, visual feeders (Thijssen et al., 1974; Carter et al., 1991; Wilson, 1991), and such risk is reduced under cover of darkness.
Activity in the autumn
Under the LD 16:8 photoperiod in the autumn, the actogram showed a high level of nocturnal activity (Fig. 1a), which is synonymous with a high feeding rate (Last, 2003). In contrast, those animals under LD 8:16 displayed relatively reduced overall activity (Fig. 1b) signifying a low feeding rate. These findings are similar to those of Last and Olive (1999) using semiquantitative, manual methods to assess the degree of foraging activity at a similar time of year.
Photoperiodism necessitates the presence of an underlying endogenous clock. The results of experiment A98 support the presence of such an oscillator in N. virens. During the first scotophase, animals under LD 8:16 ceased activity at between 1000 and 1100 GMT (Fig. 1b), 10–11 h after the onset of activity. This was surprising since the lights-on signal for these animals was at 1600 GMT, and inactivity was thus not expected for a further 7–8 h. The animals used in the A98 experiment would have been experiencing natural photoregimes of LD 12:12, suggesting that the observed time of activity cessation was due to the entrainment of past photoperiodic regimes. We hypothesize that the nocturnal activity cessation under LD 8:16 may reflect a proactive response to the anticipated time of sunrise, therefore providing indirect evidence for an endogenous timer. This response is surprising since these animals had been acclimatized to this photoperiod for a week.
Activity in subsequent scotophases rapidly became bimodal (Fig. 1b) with activity peaks in both early and late scotophases. Similar daily changes in activity under various photoperiods have been recorded for the locomotory activity patterns of the onion fly Delia antiqua (Watari and Arai, 1997). When this diurnal fly was subjected to photoperiods in which LD >8:<16, two peaks of activity were always observed. The first occurred in the early photophase and the second in the late photophase. Under short photoperiods (where LD <8:>16), the activity peaks fused as the "window" of diurnal activity became smaller. For this species, it was suggested that the main, late-photophase activity peak was due to a predictive circadian oscillator, whereas the early-photophase peak was solely due to a reactive lights-on response. For the experiment described here using N. virens, the response is interpreted as being an adjustment to the new photoperiodic cue, or Zeitgeber, where the observed inactivity at the expected dawn is evidence for transient behavior prior to attaining a steady state in the following scotophases.
Coupling between the environmental Zeitgeber and the endogenous oscillator occurs through entrainment (Aschoff, 1965). The results illustrate that the process of entrainment of an activity rhythm in the autumn experiment depends on the photoperiodic regime. Subjecting animals to a photoperiod of LD 8:16 considerably reduced the strength of the rhythm. We believe that this represents the initial stages of the autumn feeding cessation, which is triggered when the animals pass through a critical photoperiod LDcrit <12:>12 (Last et al., 1999), a characteristic physiological response at this time of year. This photoperiodically mediated change in activity provides an interesting result. At a time when the nights were becoming longer and hence, for these nocturnal animals, the "potential" for nocturnal feeding was gradually increasing, all activity became much reduced. It has been demonstrated elsewhere (Last and Olive, 1999) that under a photoperiod of LD 8:16, rates of somatic growth, segmentation, and regeneration all become much reduced and energy is channeled into future reproductive growth through changes in vitellogenesis or vitellin incorporation into the developing oocytes (Rees and Olive, 1999). We regard this as the initiation of a switch from somatic to reproductive growth. The photoperiodic transition therefore appears to trigger a change in physiological state (McNamara and Houston, 1996) from one in which essentially somatic growth processes predominate to one in which acquired resources are redeployed to sexual development. It seems at first surprising that this change in physiological state is not restricted to animals that will breed in the following spring, but occurs in all individuals regardless of age or state of maturity. A fitness model based on the life history of N. virens suggests, however, that when either or both net foraging risk and net foraging gain vary seasonally, a substantial fitness benefit accrues when all members of a mixed-age population respond to environmental signals and reduce foraging at times of maximum risk or minimum benefit (Olive et al., 2000, 2001).
Parallels can be drawn with the Salmonidae, in which temperature and daylength alone do not dictate physiological changes. Bimodality in wild fish will separate animals destined to smolt in the following spring from those that are not (Thorpe et al., 1980; Skilbrei, 1991), a mechanism that has long been known to be photoperiodically induced (Villarreal et al., 1988). In S. salar, the advancement of smoltification through photoperiodic manipulation has been shown (Duston and Saunders, 1995) to largely reduce the effects of bimodality and has provided evidence that, in the salmonids at least, a coupling between threshold length, photoperiod (and the circadian clock), and time of year (the seasonal/circannual clock) determines bimodal segregation of individuals and maturational development.
Activity in the winter
The seasonal modulation of activity in N. virens was further exemplified when the same experiments were repeated in the winter. Whereas in the autumn there had been a nearly complete cessation of foraging activity under LD 8:16 (Figs. 1b and 3b), by the winter the amplitude of activity under LD 8:16 (Fig. 2b) had dramatically increased and was high compared to activity in the autumn or under LD 16:8 (Fig. 2a). In addition, the results of autocorrelation analysis showed no significant between-treatment difference in mean strength of rhythm between winter LD 8:16 and LD 16:8. We postulate that, for animals that have not reached some critical state of sexual maturity, the optimum strategy must change as the optimum time for breeding approaches. In these animals, renewed foraging increases the probability that sufficient energy reserves will be accrued for reproduction one year later. Since out-of-burrow foraging by N. virens is inevitably associated with predation risk, increased nocturnal activity at this time will maximize exploitation of the feeding time and minimize the risk of predation from diurnal predators. Animals that have reached a critical state of maturity become susceptible to a feedback response from the maturing gametes (Porchet and Cardon, 1976; Golding and Yuwono, 1994). This seasonally mediated change in activity under LD 8:16 occurred independently of photoperiodic treatment, which is suggestive of an endogenous seasonal (and possibly circannual) rhythm of activity or the operation of some long-term interval timer.
Interestingly, the mean peak activity during the first two nights occurred around midnight (Fig. 2b), which cannot readily be explained at this stage. We hypothesize that, as in the A98 experiment (Fig. 1b), activity onset may have been due to a "memory" of past photoperiodic regimes. The 7-day acclimatization period imposed on these animals under their new photoregime may not have been long enough to prevent transient activity before a new steady state had been reached. The observed "snapshot" of activity under the autumn and winter photoperiodic treatments may be influenced by the photoperiodic history of the animals. In view of these results, a long-term assay under constant photoperiods was deemed the only suitable method for establishing true endogenous behavior without transient manifestations. Under those conditions, any spontaneous changes in activity would imply the effects of a purely endogenous, rather than exogenous, response to an innate underlying seasonal oscillator or interval timer.
Long-term activity patterns in N. virens
The results shown in Figure 3b demonstrate that a long-term regulator of activity does indeed exist in the polychaete N. virens. This endogenous modulation is revealed under a constant LD 8:16 photoperiod, 3 months after the transition through the critical photoperiod LDcrit <12:>12 previously described by Last et al. (1999). The resurgence of activity under this photoperiod was spontaneous and, irrespective of the experimental photoperiod, occurred around the time of the winter solstice. Subjecting animals to a photoperiod of LD 16:8 will result in the negative masking (Aschoff, 1960) of any long-term endogenous oscillators.
Proactive anticipation of the spring equinox has been suggested previously for other nereid polychaetes (Olive and Garwood, 1983; Fong and Pearse, 1992a, b), and the results presented here provide further evidence for the expression of an overt endogenous seasonal clock in the Nereidae. Perturbations in the activity patterns of N. virens were observed under both photoperiodic treatments in March. Under LD 16:8, a late photophase peak was observed prior to the scotophase (Figs. 3a and 4a), and under LD 8:16 a marked drop in activity was observed in both the photophase and the scotophase (Figs. 3b and 4b). These perturbations occurred when the animals would, under natural photoperiods, pass through the spring critical photoperiod (where LDcrit >12:<12). The significance of these perturbations is not yet clear, but may indicate a change in the sensitivity of the endogenous pacemaker to the external photoperiod around the time of the spring equinox.
Seasonally overt changes in physiological state and behavior of this long-lived polychaete do not occur solely as a direct response to changes in proximate factors such as temperature and, particularly, daylength. We believe that while seasonally changing daylengths will actively drive changes in physiological state through photoperiodism at certain times of the year, at other times, photoperiod acts as a Zeitgeber to entrain an endogenous seasonal rhythm. Such a mechanism has also been implicated in another marine invertebrate, Pisaster ochraceus. In this sea star, gametogenesis can be shifted by maintaining the animals on a seasonally changing photoperiod regime out-of-phase with ambient photoperiods (Pearse et al., 1986). Like our results, this finding is suggestive of an endogenous annual calendar.
To infer that these results are evidence for a circannual rhythm in N. virens would be premature since the observed response could have also been caused by an interval timer measuring a time period of approximately 90 days. Work is in progress to examine the endogenous modulation of activity patterns over the duration of years instead of months. We provide here evidence of a long-term modulation of activity in N. virens under fixed photoperiods. This mechanism modulates the response of individuals to the natural photoperiod according to real time and may be an essential adaptive component of the photoperiodically adjusted seasonal cycle of growth and reproduction.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Literature Cited |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. K. Hotchkiss, R. M. Sternberg, and G. A. LeBlanc Environmental cues trigger seasonal regression of primary and accessory sex organs of the mud snail, Ilyanassa obsoleta J. Mollus. Stud., April 17, 2008; (2008) eyn009v1. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
