Field behavior of the nudibranch mollusc Tritonia diomedea.
The goal of neuroethology to link behavior to the underlying neural control requires an understanding of relevant environmental cues and of behavioral responses to those cues. Information on behaviors and the sensory cues that influence them guides the study of sensory systems, central processing, and motor systems, the elements that constitute the neural control of behavior. The nudibranch mollusc Tritonia diomedea Bergh has, over the last 40 years, become a neuroethological model system. In T. diomedea, a relatively small number of large, re-identifiable neurons organized in a relatively simple nervous system control behaviors that are robust under the conditions of neurophysiological experimentation. These features have allowed discovery of many aspects of how the nervous system controls behaviors in the laboratory. Work on T. diomedea has focused on locomotion (e.g., Audesirk, 1975; Willows et al., 1997; Popescu and Frost, 2002; Wang et al., 2003, 2004; Cain et al., 2005), feeding (e.g., Longley, 1976; Willows, 1978; Audesirk and Audesirk, 1979; Beck, 2000), and escape responses from predators (e.g., Willows and Hoyle, 1969; Getting, 1981; Katz et al., 1994), as well as various sensory systems (e.g., Lohmann et al., 1991; Murray et al., 1992; Wang et al., 2003). However, only two focused experimental studies have been conducted on behaviors in the field (Willows, 1999; Murray, unpubl. data). There is no comprehensive description of behavior in this species. Here, we attempt to describe the behaviors of T. diomedea in its natural habitat, as well as the sensory context of those behaviors.
T. diomedea adults live, feed, mate, and spawn in the soft coral beds of the northeast Pacific (Thompson, 1971; Murray, unpubl. data). They prey on the pennatulaceans Ptilosarcus gurneyi, Stylatula elongata, Virgularia sp., and Acanthoptilum sp. (Thompson, 1971; Birkeland, 1974), and are themselves prey to the sunflower star Pycnopodia helianthoides and other sea stars (Mauzey et al., 1968; Birkeland, 1974). Therefore, the sensory systems and behaviors of the adult slugs are presumably adapted to interactions with these organisms and the physical environment found in this marine community.
What sensory modalities does T. diomedea possess, and therefore what possible cues should be considered in the natural habitat? Mechanosensation of water flows (Field and Macmillan, 1973; Murray et al., 1992; Murray and Willows, 1996; Murray, unpubl. data), chemosensation of food odor (Willows, 1978), and magnetoreception (Lohmann and Willows, 1987; Lohmann et al., 1991; Popescu and Willows, 1999; Wang et al., 2003, 2004) have all been studied in the laboratory. In addition, photoreception (light levels only; Chase, 1974) and balance information from statocysts (see Harrigan et al., 1986) may also be important. Thus, water flow, odors, touch, magnetic fields, body orientation, and light levels are the major candidate cues for control of behavior in this species.
The primary goal of this paper is to describe the behaviors of T. diomedea and candidate sensory cues in situ. Our goal is not to construct a detailed ethogram (Tinbergen, 1963; Leonard and Lukowiak, 1984), but rather to demonstrate relationships between behaviors and sensory cues in the natural habitat. We use time-lapse underwater video of the slugs in the field to observe behaviors in their natural context. We then seek to answer the following questions for T. diomedea: What are the primary behaviors and associated sensory cues in the natural habitat? Do natural behaviors differ from behaviors observed in past experimental conditions? What sensory cues should be considered when designing further behavioral and neuroethological experimentation?
In particular, we focus on navigation. Recent work on T. diomedea has targeted both sensory systems (Murray et al., 1992; Murray and Willows, 1996; Wang et al., 2004) and locomotory control (Beck et al., 2000; Popescu and Frost, 2002; Wang et al., 2003; Redondo and Murray, 2005). Progress combining the two into a study of navigation would be advanced by a description of navigation in the native habitat. Accordingly, our goal is to record T. diomedea movement patterns relative to environmental guidance cues.
We report several previously unobserved behaviors and document and extend observations on behaviors known from the laboratory. Descriptions of the natural sensory cues associated with those behaviors provide a basis for future neuroethological work on T. diomedea. In addition, we report evidence for navigation relative to conspecifics, prey, and predators that relies on direction of water flow, and we generate several hypotheses that can be tested by quantitative analyses of navigational cues and behaviors (Wyeth et al., 2006).
Materials and Methods
We used two methods to observe behaviors of Tritonia diomedea and potential cues in its natural habitat: time-lapse video ("behavior videos") and observation by scuba divers ("diver observation").
We collected data during five field trips (29 working days) to two sites. MacIntosh Rocks (49[degrees]12.60'N, 125[degrees]57.45'W), North of Vargas Island, British Columbia, Canada, was visited in October 2000, October 2001, and July 2003. Dash Point (47[degrees]19.28'N, 122[degrees]25.22'W), in southern Puget Sound, Washington, USA, was visited in April 2001 and October 2002. Both sites have perennially dense beds of the sea pen Ptilosarcus gurneyi and, usually, high populations of T. diomedea (pers. obs.). No T. diomedea individuals were found on the third trip to MacIntosh Rocks, despite an apparently otherwise unchanged P. gurneyi bed. Consequently, we imported 48 slugs from Yellow Bank (48[degrees]14.07'N, 125[degrees]55.16'W, 3.2 km distant) which, despite its high density of T. diomedea, was too exposed to wind to be suitable for our behavior camera system.
Behavior camera setup. At both field sites, our research vessel was anchored in 6 m of water above the P. gurneyi bed. We used three submersible black-and-white video cameras (model CVC-320WP, Speco Technologies, Amityville, NY), each attached to a 3-m pole driven perpendicularly into the soft sediment (Fig. 1A). The poles formed a 3-m equilateral triangle (Fig. 1B). At the center of the triangle, we placed a small marker on the substratum to indicate the direction of magnetic north. Each camera's height was determined by seawater clarity, and its field of view was angled down and to the right of the north marker. Thus, each camera covered a largely unique field of view of 4.7 [+ or -] 1.6 [m.sup.2] (mean [+ or -] st. dev.), except for a slight overlap at the center marker.
Video recording. Power and video from all cameras were cabled to the research vessel. In October 2000, video from the three cameras was recorded onto a single VCR tape using a video sequencer (cat no. 49-2511, RadioShack, Fort Worth, TX) and a time-lapse VCR (model no. STR-24N, Samsung, Ridgefield Park, NJ). The video was then digitized to computer hard disk using a VideoBus II (model no. F5U208, Belkin Corp, Compton, CA). For all subsequent field trips, video was written directly to hard disk at 2.5 frames per second at a resolution of 320 X 240 pixels using a PC computer digital video recording system (Novex 2000, Novex (Canada) Ltd., North York, ON, Canada). This frame rate and resolution ensured that no behaviors were distorted by either temporal aliasing or excessive pixilation. Power was provided by a portable generator (model no. EU1000i, Honda, Alpharetta, GA) buffered by an uninterruptible power source (model no. F6C650 USB, Belkin Corp, Compton, CA).
Current measurements. We used three methods to determine currents around individuals of T. diomedea. First, natural consequences of flow in the videos, including the motion of drift algae and the orientation and posture of P. gurneyi (Best, 1985), were used to estimate relative flow speeds (high, moderate, and low) and absolute flow direction. Second, we deployed fluorescein sources that created dye plumes, allowing us to visualize flow in the behavior cameras. Fluorescein sources consisted of a 1:4:2 (by volume) mixture of fluorescein powder, plaster of Paris powder, and fresh water, hardened overnight. Finally, we supplemented these with flow direction data collected by "current camera" videos of suspended particle movements from two cameras mounted near the substratum (Wyeth and Willows, unpubl. data).
Daily method. We deployed the cameras by scuba and oriented the recorded videos by holding a compass under each camera. We then collected nearby individuals of T. diomedea and placed them under the behavior cameras. A maximum of 10 slugs (except on 2 days) kept densities comparable to local natural densities (1-3 [m.sup.-2]; pers. obs.). We video recorded undisturbed slugs when possible (n = 44, ~25% of observed slugs), and found little effect of collection. Collected slugs crawled, fed, and mated similarly to undisturbed slugs. After the slugs had settled and all fluorescein sources had been deployed, the divers surfaced to eliminate observer interference. Thereafter, slugs and fluorescein sources were replenished as needed (about every 3-4 h). On 6 days, we introduced one or two individuals of the predatory sea star Pycnopodia helianthoides, either free to crawl or caged under plastic mesh stapled into the sediment. We caged the sea stars because handling them caused them to move quickly beyond the camera fields of view, preventing elicitation and interpretation of slug responses. At the end of each day, we compass-oriented the video again. On most days, we also provided scale by extending a rope marked at 50-cm intervals across the bottom at various orientations under the cameras. Recording was then stopped, and all underwater gear was dismantled and stowed on board the vessel.
Behavior videos. All videos were reviewed to track T. diomedea individuals within and between behavior cameras. For each slug, we recorded a behavior sequence based on the timing of the following behaviors: Stationary: Inactivity. Crawling: Forward locomotion without large changes in direction. Turning: Forward locomotion with a distinct bending of the body and a consequent change in direction. Mating: Contact between two slugs, followed by paired circling, cessation of movement with the right sides of the slugs in close contact, or both. Egg-laying: Behaviors occurring while an egg string was laid. Feeding: Contact with a sea pen, P. gurneyi, which jerks quickly towards the slug and then retracts into its sediment burrow. Swimming: Series of dorsal and ventral flexions. Other: Activity not included in any of the above.
Behavior sequences were used to determine behavior durations and frequencies. We also noted flow directions; locations of predators, prey, and conspecifics; and any other potential sensory cues. Behavior videos from Dash Point were further analyzed by measuring slug positions, headings, and behaviors every 30 s in all cameras (see Wyeth et al., 2006, for details). These data were used to calculate behavior frequencies, durations, crawling speeds, and turn rates.
Calibration. Each behavior camera video from each day was calibrated for both geomagnetic orientation and distance. For orientation, slug and flow headings were calibrated against the compass image recorded in the same video. For distances, pixel distances for known 50-cm segments in the images of the scale rope were averaged, and the ratio of mean pixel distance to actual distance was used to scale all distances in the same video.
Figures from behavior videos. We used two methods to improve the clarity and depiction of motion in figures based on video sequences. First, we used averages of 3 or 5 frames to sharpen features on the substratum. Second, we enhanced motion contrast by subtracting the mean image of a video sequence from each individual frame (Fig. 2). By varying the mean image interval and sequence frame rate, motion of different speeds was enhanced. We used "motion trails," projections of the mean subtracted video frame stacks, to show the movement of T. diomedea, P. gurneyi, fluorescein dye plumes, and other moving features as lighter paths on a dark background. Video versions of some figures are available online at http://www.biolbull.org/supplemental/.
Digital video and photography
For more detailed observation of specific behaviors, scuba divers recorded digital video and still photographs using a Sony PC100 camera (New York, NY) in an underwater housing (Mako, Light & Motion, Monterey, CA).
Video data were reviewed in Premiere 6.0 (Adobe Systems Inc., San Jose, CA). All analyses were performed in Excel 11.6 (Microsoft, Redmond, WA) and Matlab 6.5 and 7.0 (The Mathworks Inc., Natick, MA).
Summary of behavior data
We collected video data on 29 days; observations included all tidal states: 21 floods (incoming), 27 slacks (changing), and 20 ebbs (outgoing). The total duration of all behavior videos was 510 h (170 h in each camera). We observed 273 slugs in the videos; however, we estimate that about 10% of the animals left the field of view and then returned within a single day and about 25% were video recorded on a second day, suggesting that about 180 different animals were observed.
Behavior videos showed Tritonia diomedea individuals behaving in their natural habitat. Individual slugs were continuously visible, both when moving and when stationary. We observed many crawling behaviors, including navigation towards mates and prey, as well as predator avoidance. We also observed multiple instances of feeding, mating, egg-laying, and several other specialized behaviors. Other inhabitants of the Ptilosarcus gurneyi beds were also visible: P. gurneyi (when extended above the sediment), nudibranchs (Armina californica and Tritonia festiva), a common large cancrid crab (Cancer magister), and several sea star species (e.g., Pycnopodia helianthoides, Mediaster aequalis, and Solaster stimpsoni).
Slugs at Dash Point spent about half of daylight hours active (Table 1). T. diomedea individuals were visible for 2.0 [+ or -] 1.9 h (mean [+ or -] st. dev.) when tracked across cameras, and 1.6 [+ or -] 1.2 h in any one camera. The mean field of view for all three cameras combined on each day was 14.0 [+ or -] 4.9 [m.sup.2], which corresponds to a diagonal bottom-left to top-right distance for each camera of 3.1 [+ or -] 0.6 m (max = 4.4 m). The mean crawling rate for animals at Dash Point was 10 cm/min (Table 2). Thus, a continuously crawling slug would spend at most 45 min in the field of view of any one camera. However, the slugs rarely crawled through a field of view without some other behavior being observed. They interspersed relatively short periods of crawling and turns with longer matings or stationary inactivity. Feeding was relatively infrequent (they fed 1/3 as often as they mated) and occupied only a small proportion of their time (Table 1). Egg-laying was even less frequent, though it was the longest recorded behavior, lasting 5 h on average. Swim behaviors to escape predators were rare (n = 2, at Dash Point) and short in duration.
Behaviors observed in the videos could often be associated with the environment also visible in the videos. Interactions between the slugs and environmental features could then be used to explain both the control and consequences of behaviors.
Interactions with the substratum
Crawling and turning. Active slugs spent much of their time crawling across the soft sediment of P. gurneyi beds (Fig. 3A; Table 1). When animals were crawling, their locomotion was always directed forward, never backward. Crawling speeds varied widely (Table 2) but not consistently relative to either preceding or subsequent behaviors. For example, crawling speeds were not consistently higher prior to mating. In a crawling animal, the oral veil was normally extended with the tips repeatedly brushing the sediment, often accompanied by slight left and right head movements. This "sniffing" behavior was particularly obvious when approaches toward mates or prey were viewed in time-lapse mode.
Turns seen in the behavior videos were achieved by bending the body; turning without bending the body was also observed by divers. Turning rates varied widely (Table 2). Sharp turns often shortly preceded mating events: 15 animals turned before mating compared to just 6 after mating.
Berming. Divers observed slugs actively modifying the substratum during "berming" behaviors (Fig. 3B). As the animal crawled slowly forward, its oral veil bulldozed a berm of sand in front of it. Typically the berm was 1 cm higher than the surrounding sediment. Some animals were observed with their dorsum flush with the sediment surface, having moved forward about a body length, raising the berm several centimeters and spilling it along their sides. Although berming was not detectable in the behavior videos, it likely occurred as some small proportion of the slug's stationary time (Table 1).
Interactions with currents
Currents in P. gurneyi beds were highly variable. Direction and speed varied erratically over the course of minutes and hours, were not consistent across days, and did not appear to strongly correlate with tidal cycle state (Wyeth et al., 2006). Yet slugs' movements often correlated with both current direction and speed.
Low and moderate currents: upstream navigation. Slugs interacted both actively and passively with water flows, depending on current velocity. At slower speeds, the animals appeared to use currents to guide their crawling. They were frequently observed crawling directly upstream (positive rheotaxis). In particular, this seemed to occur when a slug approached another prior to mating (Fig. 4). In 24 of 25 cases with adequately recorded approach behaviors, mating was initiated by one slug crawling toward another and touching its tail or sides (the one exception was a head-to-head initiation). Eleven of these approaches could be qualitatively designated as upstream crawling. Only three counterexamples were observed, in which slug movement did not oppose flow. The remaining 10 approaches occurred with either ambiguous current observations (n = 5) or after divers placed slugs close together (n = 5). Furthermore, 6 of the 11 instances of positive rheotaxis prior to mating were observed with either a current change or slug movement that led to a new spatial arrangement, with one slug downstream of another slug (Fig. 5). The downstream slug turned as necessary to crawl upstream until it encountered the upstream slug.
High current: turns, slides, dislodgement, and righting behavior. At higher current velocities slugs risked being dislodged by the currents. We observed passive slides, when the foot remained in contact with the substratum (n = 9), and dislodgement, when animals were swept off the bottom by either bulk flow or oscillating wave motion (n = 20). Eleven (55%) of the dislodged animals were transported downstream out of the behavior camera's field of view. As current speeds increased, the slugs tended to stop crawling and remain stationary. A few examples of slugs turning to face upstream while stationary were observed in higher flows (n = 6). Slugs were observed to respond to both brief current "gusts" and some passive slides by flattening the body and spreading out laterally, beyond the typical width during crawling.
After a slug was swept off the bottom, the distance it traveled varied from a few centimeters to several meters or more, though we have no direct observation beyond the behavior cameras' field of view. Slugs transported in the currents tumbled downstream, after which they needed to regain a foothold on the substratum. If the animal came to rest on its ventral surface (n = 5), then the foot was spread out laterally, similar to the current gust response. If the animal settled on its dorsum or lateral surfaces (n = 4), then righting behavior was observed. The body was twisted along the anterior-posterior axis, rotating the anterior by about 180 degrees relative to the posterior. With the anterior portion of the foot in contact with the substratum, forward crawling began. The body was progressively untwisted as more of the foot came in contact with the sediment, until the animal resumed its normal posture with the entire foot in contact with the surface.
We observed behaviors subsequent to 9 slides and 8 dislodgements. Four of the 17 slugs crawled a short distance before stopping and remained inactive; 2 of these were facing upstream. Two animals stopped mating after being separated by the slide, and remained inactive. One slug stopped egg-laying, turned, and crawled upstream to find a mate. The remaining 8 slugs returned to prior activities: inactivity, crawling, mating, crawling on top of a previously mating pair, and egg-laying.
Interactions with conspecifics
Mating. Slugs mated frequently (Table 1). Mating pairs were observed at all times of day and within minutes of placing the slugs under the cameras. In addition, 10 serial matings were observed, in which a single animal mated consecutively with two or three slugs (either immediately or over the course of the day).
Mating in T. diomedea can be separated into three stages: initiation, alignment, and copulation (Fig. 6). Initiation involved one slug (the initiator) touching the other (the initiate) with its oral veil. The initiate (if mating was to occur) performed a right turn. The initiator also proceeded to turn right. The magnitude of these turns varied from 0[degrees] (or rarely a very slight left turn) to more than 180[degrees], depending on the relative orientations of the two slugs (Table 3). The paired right turns led to a clockwise circling behavior that brought the right sides of the animals together. This alignment stage continued with the animals circling clockwise for varying amounts of time. While the animals were circling, their oral veils swept the partner's lateral posterior body wall, and the gill tufts on the right side were usually withdrawn. Circling usually then slowed or stopped, and the animals remained stationary for about an hour. Observations in the field and the laboratory suggest that circling behavior is used to align the copulatory organs, and the animals either slow or stop circling during copulation.
Mating usually ended when one (n = 11, 50% of 22 mating pairs with adequate data after mating) or both (n = 8, 36%) of the slugs crawled away. Only three instances (14%) were observed where the animals finished copulation, separated by no more than a few centimeters, and then both remained stationary. Initiators, by definition crawling before mating, rarely remained stationary following copulation (2 of 16 with adequate data from both before and after mating). Initiates, rarely crawling before mating (2 of 16), were more evenly split after mating: 9 crawled and 7 remained stationary.
Not all contacts between slugs resulted in mating. Some were apparently aborted attempts to mate on the part of either one or both slugs. In other cases, slugs showed no reaction despite touching each other. No consistent pattern emerged to suggest why certain slug-to-slug contacts led to mating and others did not.
Mating pairs were also attractive to downstream slugs (Fig. 7). A third slug was often (n = 12) seen crawling upstream to find a pair in copulo, in similar fashion to mate approaches. The third slug would then repeatedly brush the mating pair with its oral veil, crawling around or even overtop of them. In three cases, this behavior was followed by the departure of one of the previously mating pair, and mating between the third slug and the remaining slug.
Egg-laying. Five examples of egg-laying were observed. The egg string was extruded out of the reproductive openings on the right side of the animal and down onto the substratum. Simultaneously, the animals circled slowly to the left (1.6 [+ or -] 1.0 [degrees]/min; cf. regular turning speeds in Table 2), crawling overtop of the egg string already laid down. The result was a loose counterclockwise coil that divers observed to be attached to algae or other debris.
Interactions with prey
Bite-strikes. We observed 13 predation events (bite-strikes, Fig. 8) in the behavior videos. Similar to our observations of slugs approaching conspecifics (see above), bite-strikes were usually preceded by upstream crawling (Fig. 9). Since P. gurneyi orients to flow (Best, 1985), it was possible to assign a binary upstream or downstream orientation to each bite-strike, based on the location of the strike on the sea pen (Figs. 8 and 9). In 11 of 13 cases, the slug struck the downstream side, 1 strike was equivocal, and 1 struck from upstream. Whenever (n = 8) several sea pens were visible in a behavior camera and consistently oriented as a group (indicating consistent flow direction), bite-strikes were from downstream. In only two cases did T. diomedea individuals approach their prey from another direction and then use a sharp turn to strike from downstream.
Diver observations of feeding showed that as a slug neared its prey, locomotion slowed and then ceased prior to the strike. The anterior foot margin was lifted, the oral veil raised and extended towards the sea pen, and the lips protruded (Fig. 8). Only very light contact appeared to be made with the prey before the radula and jaws lunged forward to grasp and cut away a single pinnea ("leaf") from the sea pen. After a bite-strike, most P. gurneyi individuals retracted rapidly into the sediment (53 [+ or -] 40 s, mean [+ or -] st. dev; n = 10 for complete retraction). Small individuals (<~20 cm when extended) retracted when touched by a slug, while larger ones appeared to respond only after being struck. After a bite-strike, the slug invariably continued crawling.
Interactions with predators
We observed both escape swims and avoidance crawling in response to the predatory sea star Pycnopodia helianthoides.
Escape swimming. A single escape swim occurred in the behavior videos after a slug was touched by P. helianthoides (Fig. 10). Several apparently spontaneous swims were observed, possibly due to small sea stars obscured from view. All swims resulted in the slugs being transported downstream, in some cases out of the camera's field of view. Once, when the flow was negligible, the swimming slug remained close to the swim initiation point.
Avoidance crawling. While focusing on responses to a caged individual of P. helianthoides, we observed one slug that turned and crawled away, and another that swam while still a meter downstream from the sea star (Fig. 11). Due to poor visibility with the behavior cameras, we used diver observations to further examine T. diomedea responses to a distant predator. We coaxed a sea star to crawl cross-stream, between 5 and 50 cm upstream of 10 slugs (tested individually). In 6 trials, the slugs responded by turning and crawling away. In each case, the slug reared its head upwards and lifted the anterior half of the foot off the substrate. The body was bent laterally, 90[degrees] or more. The anterior half was then lowered onto the substrate and crawling commenced, distancing the slug from the sea star. In 3 trials, the slugs swam without being touched by the sea star. Initially, the slugs responded with the same rearing turn, but instead of replacing the anterior half of the foot on the substratum, they began an escape swim with a ventral flexion. The remaining slug showed no reaction to the distant sea star.
Interactions with other organisms
T. diomedea's interactions with other organisms seemed to be limited to treating them as obstacles while crawling. Encounters with drift algae were the most frequent, and slugs were observed crawling around or overtop of algae, which appeared to promote dislodgement. Cancer crabs crawled over or into slugs, which responded with temporary gill and oral veil retraction (with the exception of one escape swim). We observed no responses to the other common sea slugs Armina californica and Tritonia festiva found in P. gurneyi beds.
We observed 180 slugs in four years, three seasons, all tidal states, and a wide range of flow rates, in an effort to describe much of the adult behavior of Tritonia diomedea. We observed most previously known behaviors for this species, extended observations on several behaviors, and document several behaviors not previously described. In addition, we describe navigation by T. diomedea in its natural habitat.
We observed T. diomedea behaviors within their sensory context and can outline possible relationships between different sensory cues and particular behaviors. On average, we observed the activity of a slug for longer than twice the duration of any of the observed behaviors (except for egg-laying; Table 1). We therefore conclude that a single camera field of view is adequate for recording complete behaviors. Similarly, the field of view was also large enough to capture sensory cues (chiefly predators, prey, mates, and flow). Thus, we are able to link the majority of behaviors observed to cues also visible in the videos.
Interactions with currents: navigational hypotheses
We observed several T. diomedea behaviors that involved interactions between water flow and upstream organisms. Previous studies considering flow direction as the only cue reported positive rheotaxis in both the laboratory (Murray and Willows, 1996) and the field (Murray, unpubl. data). Here, slug movement patterns in relation to water flow and the locations of conspecifics, prey, and predators suggest three hypotheses about how T. diomedea navigates with respect to flow. Crawling prior to most matings was directed upstream (Fig. 4), and this frequently did not occur until after a change in currents or slug position that resulted in a newly detectable upstream conspecific (Fig. 5). This suggests an active choice to crawl upstream towards an upstream conspecific. Prey were also predominately approached by crawling upstream (Fig. 9). In contrast, distant upstream predators elicited downstream crawling (Fig. 11). Thus, we hypothesize that T. diomedea uses flow cues to navigate (1) upstream to find potential mates, (2) upstream toward prey, and (3) downstream away from predators. We have now confirmed these hypotheses (Wyeth et al., 2006) and further concluded that orientation to flow depends on odors from upstream mates, prey, and predators (Wyeth and Willows, 2006).
High currents: avoidance, dislodgment, and righting
T. diomedea responds to high flow by reducing drag. When faced with strong current in the field or laboratory (Murray and Willows, 1996), slugs flatten the body laterally, turn upstream, or both. These responses correlate with reduced probability of being dislodged in a laboratory flume (Murray, unpubl. data). Berming may also have similar adaptive value (Willows, 2001). Our observations confirm that in the field there is a risk of dislodgement by either bulk tidal flow or oscillatory wave-induced flow. Once dislodged, the slugs may not settle again for many meters, with consequences similar to those of the escape-swim behavior (Willows, 2001)--that is, separation from mates or prey.
If a dislodged slug failed to land on its ventral surface, righting behavior occurred. Subsequent behaviors were inconsistent, including those characteristic of high current situations (inactivity, turning to face upstream, or lateral flattening) as well as those that appear to risk further dislodgment (for example, resuming crawling on top of a previously mating pair). Apparently, T. diomedea has no fixed response to being moved by the currents and responses are likely to be context-dependent.
Interactions with conspecifics: mating and egg-laying
Mating was frequently observed (Table 1), and despite its similarly frequent occurrence in the laboratory (pers. obs.), we provide the first description of this behavior for T. diomedea. As in some other nudibranchs (Costello, 1938), mating is a long-lasting affair in T. diomedea, with a mean duration of more than an hour. Initiation rarely occurred head-to-head, and regardless of initial orientations, the two slugs turned clockwise to achieve head-to-tail orientation (Fig. 6). Head-to-head contact is the norm for other nudibranchs, followed by further movement forward that brings the right sides in contact in a head-to-tail orientation (Costello, 1938; Rutowski, 1983; Longley and Longley, 1984). We have no data to suggest why T. diomedea should be different in this regard. Continued clockwise turning during the alignment stage by T. diomedea is also unusual; other species stop moving as the copulatory organs are first aligned. Thereafter, copulation in T. diomedea is similar to that in other nudibranchs, with little or no further movement until copulation ends. We observed serial matings between egg-laying events, which is typical of many hermaphroditic gastropods (Baur, 1998). Egg-laying in T. diomedea is also similar to that in other nudibranchs (Costello, 1938; Hurst, 1967).
Interactions with prey: bite-strikes
Predation by T. diomedea was rare (Table 1), as in previous observations (Murray, unpubl. data). Our observations of bite-strikes on the sea pen Ptilosarcus gurneyi are similar to reports of feeding on Virgularia sp. (Willows, 1978; Audesirk and Audesirk, 1979). Differences are presumably due to the differing prey sizes: individual pinnea (leaf-like structure with polyps) are browsed from the larger P. gurneyi, whereas a segment of the rachis (stem) is bitten from the smaller Virgularia sp.
Interactions with predators: avoidance crawling and escape swims
We have hypothesized that T. diomedea uses avoidance crawling in the presence of a distant upstream predator, the sea star Pycnopodia helianthoides. Although this behavior has not previously been described, it is expected given the potential costs of predation. Escape swims, despite their success in avoiding predation (Mauzey et al., 1968), are presumably also costly, chiefly through separation from mates and prey (Willows, 2001). We find support for these costs in the rarity of swims (Table 1), and the current-dependent downstream displacement of swimming slugs (Figs. 10 and 11). Avoidance turns may therefore also reduce the frequency of costly swims.
The behavioral components of swims conformed well to prior laboratory analyses (Willows et al., 1973; Hume et al., 1982). Although most laboratory studies of this behavior have considered only the dorsal and ventral flexions of the swim (except Popescu and Frost, 2002), our observations suggest that parts of the swim behavior are coordinated with other behaviors. The rearing turn used in avoidance crawling also preceded some distant swim responses. This turn is conspicuously similar to the initial dorsal flexion of induced laboratory swims. After an escape swim, the slug crawls at high speed (Audesirk, 1978; Popescu and Frost, 2002), provided it has a grip on the substratum. However, righting behavior is sometimes necessary before the animal can begin crawling. Thus, swims can be coordinated with turns beforehand and both crawling and righting behaviors afterward.
The ability to trigger escape swims at a distance (Fig. 11) suggests that in addition to touch stimuli (Willows et al., 1973; Hume et al., 1982), distance chemosensation can initiate swimming. Swim behaviors are also inhibited by certain types of dorsal stimulation in the laboratory (Mongeluzi et al., 1998). Thus, the escape swim is part of an integrated response to a potential predator, coordinated with other behaviors and affected by several sensory modalities.
Missing cues and behaviors
Magnetic fields (Lohmann and Willows, 1987), light levels (Chase, 1974), and presumably gravity are all detectable by T. diomedea, but without observable variation in these features, we are unable to interpret behavioral responses to them. We continue to find no evidence for the use of conspecific mucus trails as a navigational cue (Willows, 1978). Long-term crawling patterns (over weeks or months) and nocturnal behaviors are two potentially important groups of behaviors for which we also have no data. Understanding the full extent of navigational cues and behaviors will require further work to document overnight activity. Furthermore, as neuroethology of navigation or other behaviors progresses, observation should return to the natural habitat for more detailed description of the behavior in question.
A practical conclusion from these behavioral observations is that slug behavior must be observed with appropriate sampling over time. Viewing T. diomedea behavior in 10X to 100X time-lapse reveals details difficult for humans to observe in real time. Future behavioral and neuroetho-logical experimentation in the laboratory should take into account the appropriate spatial and temporal scales over which the behaviors occur naturally.
Many of the behaviors recorded here have been described previously from observations in the laboratory. Crawling speeds (Audesirk, 1978), upstream crawling and turns (Murray and Willows, 1996), bite-strikes (Willows, 1978; Audesirk and Audesirk, 1979), and swims (Willows et al., 1973; Hume et al., 1982) are examples. However, several have been missed in the laboratory (navigation relative to conspecifics and predators; Willows, 1978), or have received little attention despite being easily observable (mating and righting behavior). In addition, field observations tend to add details to the context in which the behaviors observed in the laboratory are performed. Thus, bite-strikes are preceded by upstream orientation, and swims can be preceded by avoidance turns and followed by righting behaviors.
Our results suggest new neuroethological experiments. For example, several neurons in the pedal ganglion are hypothesized to be motor neurons that play a role in positive rheotactic turning in T. diomedea (Murray et al., 1992; Redondo and Murray, 2005). Recognition that egg-laying and mating behaviors use, respectively, left and right turns provides alternative stimuli for neuroethological tests of turning. Conversely, turning motor neurons are clearly one starting place for neuroethological exploration of reproductive behaviors in T. diomedea, an area with little prior work, yet a dominant part of the slug's field behavior. A second example: neurons involved in the central control of the escape swim have already been linked with control of post-swim crawling (Popescu and Frost, 2002). Do the same neurons also control righting behavior if the animal is inverted?
The recent focus on T. diomedea sensory systems (Lohmann et al., 1991; Murray et al., 1992; Murray and Willows, 1996) and the control of locomotion (Popescu and Frost, 2002; Wang et al., 2003; Redondo and Murray, 2005; Cain et al., 2006) has raised the possibility of exploring the neuroethology of navigation in this species. Four other species of opisthobranchs (Aplysia californica, A. fasciata, Navanax inermis, and Bursatella leachii) have similar behavioral descriptions or ethograms (Kupfermann and Carew, 1974; Susswein et al., 1984; Leonard and Lukowiak, 1984, 1986; Ramos et al., 1995). Yet for none of these other slugs are detailed, continuous descriptions of navigation available. In T. diomedea, we can combine easily recorded navigational behaviors with the neuroethological advantages inherent to gastropod molluscs to consider the problem of how animals navigate. The first step towards understanding navigation in T. diomedea has been to describe navigation in the field. The next step is to use quantitative analysis of independent crawling and current data (Wyeth and Willows, unpubl. data) to confirm the importance of water-flow direction in navigation relative to predators, prey, and conspecifics (Wyeth et al., 2006).
We thank O.M. Woodward for many tireless hours of scuba work. We are grateful to S.D. Cain, O.M. Woodward, W. Moody, R.R. Strathmann, G. VanBlaricom, J.A. Murray, D. Duggins, R. Redondo, M. Baltzley, S. Hardy, the staff of Friday Harbor Laboratories, and two anonymous reviewers for help or suggestions that contributed to this work. Support was provided by the Packard Foundation. R.C.W. thanks C.P. Holmes, "harem j" of Friday Harbor Laboratories, and support from the National Sciences and Engineering Research Council (Canada) and the Conchologists of America.
Audesirk, G. 1978. Central neuronal control of cilia in Tritonia diomedia. Nature 272: 541-543.
Audesirk, T. E. 1975. Chemoreception in Aplysia californica. I. Behavioral localization of distance chemoreceptors used in food-finding. Behav. Biol. 15: 45-55.
Audesirk, T. E., and G. J. Audesirk. 1979. Oral mechanoreceptors in Tritonia diomedea. II. Role in feeding. J. Comp. Physiol. 130: 79-86.
Baur, B. 1998. Sperm competition in molluscs. Pp. 255-305 in Sperm Competition and Sexual Selection, T. R. Birkehead and A. P. Moller, eds. Academic Press, San Diego, CA.
Beck, J. C. 2000. Localization and activity of pedal peptide within the central nervous system of the gastropod mollusc Tritonia diomedea. Ph.D. dissertation, University of Washington, Seattle, WA.
Beck, J. C., M. S. Cooper, and A. O. D. Willows. 2000. Immunocytochemical localization of pedal peptide in the central nervous system of the gastropod mollusc Tritonia diomedea. J. Comp. Neurol. 425: 1-9.
Best, B. A. 1985. An integrative analysis of passive suspension feeding: the sea pen Ptilosarcus gurneyi as a model organism. Ph.D. dissertation, Duke University, Durham, NC.
Birkeland, C. 1974. Interactions between a sea pen and seven of its predators. Ecol. Monogr. 44: 211-232.
Cain, S. D., J. H. Wang, and K. J. Lohmann. 2006. Immunochemical and electrophysiological analyses of magnetically responsive neurons in the mollusc Tritonia diomedea. J. Comp. Physiol. A 192: 235-245.
Chase, R. 1974. The electrophysiology of photoreceptors in the nudibranch mollusc, Tritonia diomedea. J. Exp. Biol. 60: 707-719.
Costello, D. P. 1938. Notes on the breeding habits of the nudibranchs of Monterey Bay and vicinity. J. Morphol. 63: 319-343.
Field, L. H., and D. L. Macmillan. 1973. An electrophysiological and behavioural study of sensory responses in Tritonia (Gastropoda, Nudibranchia). Mar. Behav. Physiol. 2: 171-185.
Getting, P. A. 1981. Mechanisms of pattern generation underlying swimming in Tritonia. I. Neuronal network formed by monosynaptic connections. J. Neurophysiol. 46: 65-79.
Harrigan, J. F., T. J. Crow, A. M. Kuzirian, and D. L. Alkon. 1986. Behavioral, electrophysiological, and morphological investigations of statocyst function in the nudibranch mollusc Hermissenda crassicornis. Biol. Bull. 170: 305-320.
Hume, R. I., P. A. Getting, and M. A. del Beccaro. 1982. Motor organization of Tritonia swimming. I. Quantitative analysis of swim behavior and flexion neuron firing patterns. J. Neurophysiol. 47: 60-74.
Hurst, A. 1967. The egg masses and veligers of thirty Northeast Pacific opisthobranchs. Veliger 9: 255-288.
Katz, P. S., P. A. Getting, and W. N. Frost. 1994. Dynamic neuromodulation of synaptic strength intrinsic to a central pattern generator circuit. Nature 367: 729-731.
Kupfermann, I., and T. J. Carew. 1974. Behavior patterns of Aplysia californica in its natural environment. Behav. Biol. 12: 317-337.
Leonard, J. L., and K. Lukowiak. 1984. An ethogram of the sea slug, Navanax inermis (Gastropoda, Opisthobranchia). Z. Tierpsychol. 65: 327-345.
Leonard, J. L., and K. Lukowiak. 1986. The behavior of Aplysia californica Cooper (Gastropoda; Opisthobranchia): I. Ethogram. Behaviour 98: 320-360.
Lohmann, K. J., and A. O. D. Willows. 1987. Lunar-modulated geomagnetic orientation by a marine mollusk. Science 235: 331-334.
Lohmann, K. J., A. O. D. Willows, and R. B. Pinter. 1991. An identifiable molluscan neuron responds to changes in earth-strength magnetic fields. J. Exp. Biol. 161: 1-24.
Longley, A. J., and R. D. Longley. 1984. Mating in the gastropod mollusk Aeolidia papillosa: behavior and anatomy. Can. J. Zool. 62: 8-14.
Longley, R. 1976. Analysis of the feeding mechanism in the mollusc Tritonia. Ph.D. dissertation, University of Oregon, Eugene, OR.
Mauzey, K. P., C. Birkeland, and P. K. Dayton. 1968. Feeding behaviour of asteroids and escape responses of their prey in the Puget Sound region. Ecology 49: 603-619.
Mongeluzi, D. L., T. A. Hoppe, and W. N. Frost. 1998. Prepulse inhibition of the Tritonia escape swim. J. Neurosci. 18: 8467-8472.
Murray, J. A., and A. O. D. Willows. 1996. Function of identified nerves in orientation to water flow in Tritonia diomedea. J. Comp. Physiol. A 178: 201-209.
Murray, J. A., R. S. Hewes, and A. O. D. Willows. 1992. Water-flow sensitive pedal neurons in Tritonia: role in rheotaxis. J. Comp. Physiol. A 171: 373-385.
Popescu, I. R., and T. M. Frost. 2002. Highly dissimilar behaviors mediated by a multifunctional network in the marine mollusk Tritonia diomedea. J. Neurosci. 22: 1985-1993.
Popescu, I. R., and A. O. D. Willows. 1999. Sources of magnetic sensory input to identified neurons active during crawling in the marine mollusc Tritonia diomedea. J. Exp. Biol. 202: 3029-3036.
Ramos, L. J., J. L. L. Rocafort, and M. W. Miller. 1995. Behavior patterns of the aplysiid gastropod Bursatella leachii in its natural habitat and in the laboratory. Neurobiol. Learn. Mem. 63: 246-259.
Redondo, R. L., and J. A. Murray. 2005. Pedal neuron 3 serves a significant role in effecting turning during crawling by the marine slug Tritonia diomedea. J. Comp. Physiol. A 191: 435-444.
Rutowski, R. L. 1983. Mating and egg mass production in the aeolid nudibranch Hermissenda crassicornis (Gastropoda: Opisthobranchia). Biol. Bull. 165: 276-285.
Susswein, A. J., S. Gev, Y. Achituv, and S. Markovich. 1984. Behavioral patterns of Aplysia fasciata along the Mediterranean coast of Israel. Behav. Neural Biol. 41: 7-22.
Thompson, T. 1971. Tritoniidae from the North American Pacific coast. Veliger 13: 333-338.
Tinbergen, N. 1963. On aims and methods of ethology. Z. Tierpsychol. 20: 410-433.
Wang, J. H., S. D. Cain, and K. J. Lohmann. 2003. Identification of magnetically responsive neurons in the marine mollusc Tritonia diomedea. J. Exp. Biol. 206: 381-388.
Wang, J. H., S. D. Cain, and K. J. Lohmann. 2004. Identifiable neurons inhibited by Earth-strength magnetic stimuli in the mollusc Tritonia diomedea. J. Exp. Biol. 207: 1043-1049.
Willows, A. O. D. 1978. Physiology of feeding in Tritonia. I. Behaviour and mechanics. Mar. Behav. Physiol. 5: 115-135.
Willows, A. O. D. 1999. Shoreward orientation involving geomagnetic cues in the nudibranch mollusc Tritonia diomedea. Mar. Freshw. Behav. Physiol. 32: 181-192.
Willows, A. O. D. 2001. Costs and benefits of opisthobranch swimming and neurobehavioral mechanisms. Am. Zool. 41: 943-951.
Willows, A. O. D., and G. Hoyle. 1969. Neuronal network triggering a fixed action pattern. Science 166: 1549-1551.
Willows, A. O. D., D. A. Dorsett, and G. Hoyle. 1973. The neuronal basis of behavior in Tritonia. III. Neuronal mechanism of a fixed action pattern. J. Neurobiol. 4: 255-285.
Willows, A. O. D., G. A. Pavlova, and N. E. Phillips. 1997. Modulation of ciliary beat frequency by neuropeptides from identified molluscan neurons. J. Exp. Biol. 200: 1433-1439.
Wyeth, R. C., and A. O. D. Willows. 2006. Odours detected by rhinophores mediate orientation to flow in the nudibranch mollusc, Tritonia diomedea. J. Exp. Biol. 209: 1441-1453.
Wyeth, R. C., O. M. Woodward, and A. O. D. Willows. 2006. Orientation and navigation relative to water flow, prey, conspecifics and predators by the nudibranch mollusc, Tritonia diomedea. Biol. Bull. 210: 97-108.
RUSSELL C. WYETH* AND A. O. DENNIS WILLOWS
Department of Biology, University of Washington, Seattle, Washington 98195-1800; and Friday Harbor Laboratories, 620 University Rd., Friday Harbor, Washington 98250
Received 8 June 2005; accepted 24 December 2005.
* To whom correspondence should be addressed, at Department of Physiology and Biophysics, Dalhousie University, Halifax, NS, B3H 1X5, Canada. E-mail: Russell.Wyeth@dal.ca
Table 1 Duration of major activities performed by Tritonia diomedea in the behavior videos from Dash Point Duration % of Mean St. Dev. Min Max Total grand Behavior n (h:min) (h:min) (h:min) (h:min) (h) total Stationary 97 0:52 1:00 <0:01 6:14 84.6 45.3 Mating 45 1:07 0:36 0:07 2:13 50.6 27.1 Crawling 200 0:08 0:05 <0:01 0:24 27.6 14.8 Egg-laying 3 5:09 0:28 4:38 5:34 15.5 8.3 Turning 111 0:04 0:04 <0:01 0:32 8.0 4.3 Feeding 13 0:02 0:02 <0:01 0:07 0.5 <1.0 Escape Swimming 2 <0:01 <0:01 <0:01 0:01 <0.01 <1.0 The behavior of each visible slug was coded every 30 s during 55 h of observation over 11 days. The total duration of all coded behaviors for all slugs was 186.8 h. Table 2 Crawling and turning speeds for Tritonia diomedea locomotion in the behavior videos from Dash Point Behavior Parameter Unit n Mean St. Dev. Min. Max. Crawling Speed cm [min.sup.-1] 127 10.2 3.0 4.0 20.7 Angular [degrees] 199 5.1 5.8 0.0 43.6 Speed [min.sup.-1] Turning Speed cm [min.sup.-1] 74 8.2 3.6 2.4 17.1 Angular [degrees] 110 27.3 20.9 0.24 148.8 Speed [min.sup.-1] Both behaviors involved forward locomotion. Turns, due to bending the body, had a much higher mean angular velocity than crawling. Speed calculations were limited to 6 days with adequate calibration data. Table 3 Duration and rotation of mating stages in Tritonia diomedea Stage Duration (min) Rotation ([degrees]) Initiation 4 [+ or -] 2 Initiator: 42 [+ or -] 57 Initiate: 147 [+ or -] 95 Alignment 9 [+ or -] 12 76 [+ or -] 130 Copulation 69 [+ or -] 30 18 [+ or -] 29 Initiation stage: when one slug (the initiator) touched another (the initiate), and they first turned right to create a head-to-tail alignment. Alignment stage: when both slugs continued turning to the right. Copulation stage: when turning slowed or stopped. Rotation measured as the angle swept by the slug's head as it turned. Data are means [+ or -] st. dev. n = 19.
|Printer friendly Cite/link Email Feedback|
|Author:||Wyeth, Russell C.; Willows, A.O. Dennis|
|Publication:||The Biological Bulletin|
|Date:||Apr 1, 2006|
|Previous Article:||Mechanisms of gravitaxis in Chlamydomonas.|
|Next Article:||Orientation and navigation relative to water flow, prey, conspecifics, and predators by the nudibranch mollusc Tritonia diomedea.|