Abstract Benthocodon pedunculata is a deep-sea benthopelagic medusa found abundantly in Monterey Bay and Carmel Bay,which has been virtually ignored since its original description (Bigelow, 1913). Recent observations from a remotely operated vehicle (ROV), working in the Monterey Submarine Canyon, reveal that the medusae live primarily near the deep-sea floor and may be potential predators of benthic or benthopelagic crustaceans. This study investigates the gross morphology of the medusae, tentacle types, and nematocyst types. Phase contrast microscopy and scanning electron microscopy were used to detect morphological differences between the inner and outer tentacles and also revealed two nematocyst types previously undescribed in the literature, the modified-aspirotele nematocyst and the modified-stenotele nematocyst. The inner tentacles are annulated, with rings of both nematocyst types, and contain digestive inclusions, which may aid in digestion or storage. The outer tentacles have an abscission point and are readily discarded. They are coverec with both nematocyst types and are possibly used for encountering and capturing prey. Gut analyses of medusae yielded copepods, amphipods, foraminiferans, pycnogonid, and other unidentified organisms which, along with data on distribution and abundance of crustacean zooplankton in Monterey Canyon (Kelly, 1992), show that the medusae are feeding on benthic and benthopelagic organisms. Based on the results of this study, a model of tentacular function in feeding medusae was generated to show that the outer and inner tentacles of Benthocodon pedunculata may entrap prey during different pulsation periods of the umbrella. Further gut analyses and in situ observations of feeding medusae would be helpful in further resolving the foraging and feeding biology of B. pedunculata. Introduction The benthopelagic medusae Crossota pedunculata was originally described by Bigelow (1913) and later renamed as Benthocodon pedunculata (Larson and Harbison, 1990). Dives made with a remotely operated vehicle (ROV) in Monterey Bay (U.S.A.) have typically found aggregations of medusae at depths between 500 m and 900 m, within 5 m of the sea floor. This environment is characterized by an increase in biomass, including medusae (Wishner, 1980a), with some copepods found exclusively in the benthopelagic boundary (Wishner, 1980b). Crustaceans are the dominant arthropods in the deep sea (Gage and Tyler, 1991) and in terms of density in the Monterey Canyon, the most abundant crustacean was shown to be copepods (Kelly, 1992). Therefore, benthopelagic medusae, e.g. Benthocodon pedunculata, may prey on copepods and other abundant deep-sea crustaceans (Larson et al., 1992) and have distinctive morphological features which allow them to live in the deep-sea benthopelagic environment. There are several characteristic features of the order Trachymedusa, to which Benthocodon pedunculata belongs (Mayor, 1910; Marshall and Williams, 1971). Trachymedusae are generally carnivorous and either active or passive predators (Laverack and Dando, 1987). All members are mobile medusae whose tentacles are hollow and stem from the margins of the umbrella. These tentacles are structures that are specialized for prey capture and defense. The gonads are developed from the radial canals. The body wall has 3 basic layers: epidermis, mesoglea (usually well-developed in subumbrella), and gastrodermis (Laverack and Dando, 1987; Thomas and Edwards, 1991). The tentacular epidermis contains nematocysts, complex organelles each of which contains a shaft and an inverted, spined tube coiled within a capsule (Westfall, 1970; Harrison and Westfall, 1991), and interstitial cells that give rise to these stinging cells (Marshall and Williams, 1971). Different species of cnidarians have up to 7 types of nematocysts, which affect the prey type captured. A nematocyst is only used once (Marsical, 1974; Barnes, 1987). Several types of nematocysts have been classified and described (Marsical, 1974). The nematocysts may adhere to, wrap around, penetrate, or paralyze prey, but the actual mechanism of discharge is poorly understood (Laverack and Dando, 1987). It has been proposed that discharge is dependent on chemical and/or mechanical stimuli, such as triggers extending from the nematocyst known as cnidocils. Recent observations from a remotely operated vehicle (ROV), working in the Monterey Submarine Canyon, reveal that the medusae live primarily near the deep-sea floor and may be potential predators of benthic or benthopelagic crustaceans. Because so little is known about these specialized structures along the tentacles that are crucial to the survival of B. pedunculata , the morphology of the nematocysts are interesting to study, particularly if nematocyst type affects the type of prey captured. This study investigates the gross morphology of the medusae, tentacle types, and nematocyst types using phase contrast and scanning electron microscopy. Prey type will be determined by gut analysis of collected specimens. It is hypothesized that because Benthocodon pedunculata are typically found on or near the bottom of the ocean floor, they would have tentacles and nematocysts that would allow them to feed on benthopelagic crustaceans and other benthic organisms. Materials and Methods Location Eighteen specimens of Benthocodon spp. were collected on five dives using the Monterey Bay Aquarium Research Institute's (MBARI) ROV Ventana from the RV Point Lobos. Five specimens were collected using the Johnson Sea Link II manned submersible in the Atlantic Ocean. (Table 1) Specimen Collection Individual medusae were captured in a detritus sampler [transparent acrylic cylinders with lids; designed by Harbor Branch Oceanographic Institute, Inc (HBOI)). Additional medusae were collected using a variable speed suction sampler (HBOI) with five sample jars. Once the samplers were removed from the ROV, the medusae were transferred to round tupperware containers and stored in a cooler maintained at 5°C. Specimens were transferred within two hours of the initial transfer from the samplers to an aquarium and kept under controlled conditions. Eighteen specimens of Benthocodon spp were collected between April 9, 1993 and May 24, 1993. Fourteen medusae were maintained for laboratory and microscope studies. Three specimens that were preserved immediately on board ship and 8 medusae that were maintained in lab were used for gut content analysis. One specimen was identified as Benthocodon hyalinus and was not used for this study. Laboratory studies Four organisms were maintained in an aquarium with 50C running sea water in the deep sea lab under dark conditions; ten specimens under light conditions. Two medusae were kept in a .62 m *42 m *.18m aquarium; the remaining 2 in a 41 m*.2 m *.27 m aquarium. They were fed 24 hours later with Tigriopus californicus and Artemia salina. Ten minutes before feeding, water flow was discontinued. Approximately 20 prey organisms were placed into the aquarium near the outer tentacles of each medusa. Five minute observations of medusa behavior were made every fifteen minutes for 1 hour. Dead specimens were immediately removed from the aquarium and preserved in 3% glutaraldehyde. Gut Contents Eleven medusae from the Soquel Canyon and five specimens collected from the western Atlantic using the Johnson Sea Link II manned submersible on October 29, 1991 (described in Larson et al., 1992). The entire digestive system--mouth, stomach, and peduncle-- was removed from the subumbrella and transferred to filtered sea water for microdissection using a Nikon SM2-U Type 104 dissecting scope with light field/dark field illumination and fiber optic lights (Bausch & Lomb Fiber-Lite). Photographs of some samples were taken using a Nikon 35 mm camera and a Nikon HFX-DX electronic camera system. The grooves of the stomach were examined for prey All items found were identified to the lowest taxonomic level possible and transferred to glass vials filled with 70% ethanol. Microscope studies Eight specimens were fixed in 3% glutaraldehyde and then placed in successive ethanol concentrations up to 100%. The general structure of the medusa was examined under a dissecting microscope and the structure of 22 outer tentacles and six inner tentacles from 6 different specimens were examined with a light compound microscope. Squash preparations of eight preserved tentacle margins from 4 different medusae were made on slides after rinsing the entire specimen in freshwater and replacing the water with 30% ethanol. Nematocyst structures were examined using a Zeiss Axioskop compound microscope under 100x oil immersion phase contrast and photographs were taken using an Olympus OM2S 35mm SLR with Ektar 100 and Ektachrome 160T films. Unpreserved outer and inner tentacles from two specimens were transferred from the aquarium into a vial of distilled water. After five minutes the tentacles were squashed onto slides. Uncollected tentacles were placed into a petri dish filled with sea water and maintained at 50C temperature and analyzed for nematocyst activity. Measurements of the lengths and widths of nematocysts from 22 unpreserved outer tentacles and 3 unpreserved inner tentacles from two specimens were taken using a Dage-MTI digital 81 series high resolution camera with a TECON frame grabber board version 2.0 and the Bioscan OPTIMAS image processing program for Microsoft Windows. Nematocyst lengths were measured from the opening of the nematocyst or base of shaft directly to the other end. Widths were measured from the widest points of the nematocyst. The ultrastructure of nematocysts and tentacles of this trachymedusa was studied using an Hitachi S-450 scanning electron microscope. Tentacles fixed in 5% glutaraldehyde were transferred to 30% ethanol and dehydrated in a graded ethanol series to 100% ethanol, 10 minutes in each concentration. Following dehydration, the samples were critical point dried with CO2 and sputter coated with a 100A layer of gold in a Denton vacuum evaporator. For histological study, five outer and seven inner tentacles were dehydrated in a graded ethanol series: from 50%, 60%, 70%, 80%, 90%, to 95% ethanol in scintillation vials at room temperature for15 minutes at each concentration. Following the dehydration series, 10 mL of 50% LR white resin solution (mixed 1:1 with 100% ethanol) was added to each sample and the vials were rotated at 40( for 2 hours. The supernatant was replaced with 10 mL of a 75% resin solution, followed by rotation 4°C for 2 hours. The procedure was repeated with a 100% solution, then the sample was kept in fresh 100% resin solution overnight. The following day, the samples were placed into a100% resin solution in plastic capsules and heated until the plastic solidified. A Sorvall ultramicrotome was used to make 0.5 u sections along the longitudinal axis of two outer tentacles and four inner tentacles from two medusae. The sections were then fixed onto slides by heating over a bunsen burner. Basic fuchsin and methylene blue dyes were used to stain the samples. Slides were examined using the Zeiss Axioskop compound microscope under 100x oil immersion phase contrast and photographs were taken using an Olympus OM25 35mm SLR. Results Observations of Benthocodon pedunculata in situ (Fig. 1) revealed that these reddish-brown pigmented trachymedusae hover near or sit on the ocean floor. They are generally found in patches in the Soquel Canyon. Characteristic of this species is the presence of three types of tentacles (Fig. 1) that seem to differ in morphology (Fig. 1, 2). There appears to be two types of outer tentacles, although structural differences were difficult to detect with the ROV because the medusae rapidly swam away from view. One group of outer tentacles curved downward and in toward the subumbrella while the other group seemed to extend upward and outward (Fig. 1). The inner tentacles are the third type, those closest to the subumbrella and shorter and thinner than the outer tentacles (Fig. 3, 4). The outer tentacles have an abscission point, which is absent on the inner tentacles (Fig. 5). Even when fully extended in the uncontracted medusa, the tentacles are numerous and closely spaced (Fig. 1). Morphological differences between the two outer tentacles were indistinguishable in the lab and the two types were therefore grouped together as outer tentacles. Laboratory observations of 14 specimens of Benthocodon pedunculata showed that they were sensitive to prolonged exposure to fluorescent light. The two medusae maintained in the deeper tank under dark conditions survived for six days, the ten medusae in the shallow tank under light conditions died in less than two days, and the two specimens kept in the shallow tank under dark conditions were alive for three days. The specimens from the .18 m deep tank were often seen swimming near the surface while the two medusae in the .27 m deep tank were frequently observed to be resting on the bottom. Feeding experiments began within two days of specimen collection. Observations of medusae after exposure to either Tigriopus californicus or Artemia salina showed no visible change in behavior. Dissection of these medusae showed empty digestive tracts. Further post mortem examinations showed that only the inner tentacles were left attached to the edge of the subumbrella. All outer tentacles and a few inner ones were left intact at the bottom of the aquarium. Although the dissections of eight medusae that were maintained in the lab revealed no gut contents, analysis of eight specimens that were immediately fixed on board ship showed recognizable prey in six of the eight (Table 2), five of which were collected during night dives and one during an afternoon dive (Table 1). Copepod and amphipod (Fig. 6) exoskeletons were discovered in all the dissected specimens. One pycnogonid exoskeleton (Fig. 7), 22 Globigerina foraminiferans, and several unidentified exoskeletons were also found. The remaining two did not have any gut contents. Like other trachymedusae, the tentacles of Benthocodon pedunculata stem from the edge of the subumbrella (Fig. 8 ). There are 3 or 4 uneven rows of tentacles (Fig. 9). The inner tentacle is annulated, completely encircled by rings containing nematocysts (Fig. 10a). The rings disappear at the tip of the tentacle, where nematocysts cover the entire surface (Fig. 10b). Nematocysts within these rings are at various orientations and are adjacent to undifferentiated interstitial cells (Fig. 11a, b). Digestive inclusions are found in the gastrodermis of the inner tentacles (Fig. 11a) but not in the outer tentacles. The outer tentacle is more heavily pigmented and nematocysts and interstitial cells are distributed along the entire surface area (Fig. 12a, b). The nematocysts are positioned at various orientations, as shown by the orientation of the shaft and the wound thread (Fig. 12b). The nematocyst types are modified versions of the stenotele and aspirotele nematocysts (Marsical, 1974). Lengths and widths of nematocysts from both the tentacle types were measured and fell into two categories (Fig. 14). Modified-aspirotele nematocysts (MAN) were measured and ranged from 8-13 u in length and 4-6 u in width while modified-stenotele nematocysts (MSN) were measured with a range from 11-15 u in length and 10- 12 u in width. Both types of nematocysts are found on the inner and outer tentacles (Fig. 11, 12) with a MAN: MSN ratio of about 3:1. Ultrastructure of extruded nematocysts were observed with both the phase contrast microscope and scanning electron microscope (Fig. 14, 15, 16). The average lengths of the extruded shaft of MAN and MSN are 7.1841.11 u (n=78) and 10.29+1.58 u (n=54), respectively. Three stylets are visible on the shaft of both nematocyst types, near the base of the thread (Fig. 14a, 15, 16). Also common to both types are spiraled barbs that travel the length of the extended threads (Fig. 15,16). These threads extend up to 261 u in length. Discussion Patches of Benthocodon pedunculata were frequently found hovering over or resting on the ocean floor, suggesting that they may feed on benthopelagic organisms, most of which are crustaceans (Kelly, 1992), and benthic organisms, of which the foraminiferan Globerigina bulloides constitute 40% of collected animals (Petry, 1992). Attempts to observe feeding in the lab by introducing medusae to Tigriopus californicus and Artemia salina were unsuccessful because the gastrovascular cavities of medusae were completely empty. Nevertheless, it is inconclusive whether the medusae are not eating because of the conditions of the aquarium or because Tigriopus and Artemia are not palatable. The specimens may have been damaged during collection and lack of outer tentacles may have prevented the medusae from capturing the prey. Specimens of Benthocodon pedunculata maintained in the aquarium under controlled conditions showed sensitivity to 10 fluorescent light, as measured by differences in survival time. The four specimens maintained under dark conditions survived one to three days longer than those under light conditions. However, the four that survived longer were collected by detritus samplers while the others were collected using a variable speed suction sampler. This apparent sensitivity to light, then, may actually be sensitivity to the method of collection. The detritus sampler is probably less damaging to the specimens than the suction samplers. Of the four medusae collected by detritus samplers and maintained in tanks of different depths under dark conditions, the two that were in the deeper tank (.27 m) survived three days longer than the two in the shallower tank (.18 m). This difference in survival time may be caused by the slight change in pressure between the two tanks or by the difference in nutritional state of the medusae upon collection. The ones that survived the longest may have fed just before being collected. Gut content analysis did reveal recognizable prey in six out of eight specimens that were not used for feeding experiments. Copepod and amphipod exoskeletons were discovered in five medusae and exoskeleton debris was found in all six. Because these crustaceans are non-sessile organisms that inhabit the benthopelagic environment (Kelly, 1992; Wishner, 1980b; Gage and Tyler, 1991), the trachymedusae probably have tentacles that allow them to capture these prey organisms. Sessile benthic organisms that live in the sediment of the ocean floor, e. g. pycnogonid and foraminiferans, were also discovered in the gut contents, indicating that Benthocodon 11 pedunculata probably have mechanisms that allow them to feed while resting on the deep-sea floor. The two nematocyst types found in collected specimens (MAN and MSN) are present on the inner and outer tentacles. The MANs were found to be more abundant than MSNs on both tentacles, with an approximate MAN to MSN ratio of 3:1. The structural similarities between the two nematocyst types are the three spicules on the shaft and the spiraled barbs on the long threads. Observable differences are the sizes of the nematocyst capsule and the shaft, both of which is larger in the MSN. The MSN possibly allows Benthocodon to capture larger prey, although the larger number of MAN found on the tentacles suggest that Benthocodon may feed more often on smaller prey. Both stenoteles (Hufnagel et al., 1985) and aspiroteles have been characterized as probable penetrants (Kramp, 1935), but those found in B. pedunculata have long barbed threads that differentiate Benthocodon nematocysts from the traditional nematocyst description. Microscope studies of tentacular and nematocyst morphology assumed to play a crucial role in prey capture reveal properties that allow Benthocodon pedunculata to capture both small, hard-bodied organisms and larger, gelatinous organisms. In trachymedusa Tesserogastria musculosa, the unbarbed thread of the stenotele was shown to penetrate between two joints of a copepod leg (Hesthagen, 1971). In Benthocodon pedunculata, the barbed thread may both penetrate exoskeletons of crustaceans and wrap around to and adhere to soft-bodied prey, because use of penetration exclusively may tear through the mesoglea and allow gelatinous organisms to 12 escape. Hydrozoans that feed on microcrustaceans often have nematocysts concentrated in batteries and filiform tentacles whereas those possessing nematocysts with extremely long threads and thick tentacles often feed on larger gelatinous organisms (Purcell and Mills, 1988). Nematocysts and tentacles of Benthocodon pedunculata have both properties. The thin, inner tentacles have nematocysts concentrated in batteries except for the very tip of the tentacle, which suggest that they may function in capturing the copepods, amphipods, and other crustaceans discovered in the gut contents. In contrast, nematocysts are spread out over the large, outer tentacles, suggesting that these trachymedusae also feed on gelatinous prey that are probably more difficult to detect than exoskeleton debris in gut dissections. Therefore, these medusae may actually be feeding on other benthopelagic or benthic medusae, even though gelatinous prey was not observed. Prey capture is probably determined by swimming pattern, tentacle posture (Mills, 1981), and nematocyst type (Purcell, 1984; Barnes, 1987; Purcell and Mills, 1988; Harrison and Westfall, 1991). Benthocodon pedunculata were observed to swim rapidly upwards from the deep-sea floor, then drift slowly downwards, extending their tentacles (Larson et al., 1992), as well as actively swimming through the water column, although this may be biased by ROV disturbances in the water current. Prey capture in swimming medusae, therefore, may depend on vortexes produced by bell pulsations and the resulting water flow through the tentacles lining the bell margin and into the subumbrella cavity, similar to the feeding mode of scyphomedusa Aurelia aurita (Costello, 1992). As 13 Benthocodon contracts its bell, fluid is propelled out of the subumbrella cavity and through the inner tentacles on the tentacle margin, where prey is contacted and entangled by nematocyst threads. When the medusa relaxes, water refills the cavity by flowing inward through the outer tentacles where prey is again captured by nematocyst threads. By this mechanism, the close spacing of tentacles is advantageous to B. pedunculata in sieving out small organisms (e. g. copepods and amphipods). During each short upward-swimming and sinking bout, the tentacles are extended to sieve prey from the vertical water column through a cylindrical contact space, characteristic of other trachymedusae (Larson et al., 1989). The inner and outer tentacles that extend downward perform the initial sieving of the water column and the outer tentacles that are more widely spaced and extend above the bell margin may serve to capture organisms that have escaped through the sieve. The feeding on benthic organisms may involve a slightly different feeding strategy. Because Benthocodon medusae are frequently found on the ocean floor, tentacles probing over the sediment surface may be important in detecting sessile organisms such as pycnogonids and foraminiferans. Benthic organisms can also be captured when uncovered and uplifted from the sediment during disturbances caused by the upward-swimming and sinking bouts, like hydromedusae Polyorchis penicillatus (Mills, 1981). These mechanisms assume that these medusae do detect prey are hypotheses to be further tested with in situ and laboratory experiments. 14 Aside from their role in prey capture, histological studies on the tentacles show properties which suggest other functions. Unlike the inner tentacles, the outer tentacles lack digestive inclusions, which may aid in digestion and/or storage, even though the fact that the tentacles extend from the edge of the subumbrella suggests that they are both extensions of the gastrovascular cavity. However, because the outer tentacles appear to be easily shed, as evidenced by the presence of tentacles in the aquarium post mortem and by their presence in the temporary tupperware containers during collection, they would be useless as storage or digestive structures. The inner tentacles, however, remain attached to the subumbrella, so these activities are maintained. In the case of prey capture, the shedding of tentacles attached to vigorously swimming prey may be more beneficial than expending the extra energy needed to subdue the organism. As a predator avoidance mechanism, medusae whose tentacles are entangled in mucoid nets produced by predators such as siphonophores, ctenophores, and other medusae can readily escape by leaving their tentacles. Again, confirmation of these hypotheses depends on further behavioral observations. Feeding behavior, however, has not been observed in situ, because Benthocodon medusae are most likely sensitive to the lights from the ROV, just as they are sensitive to light conditions in the lab. Another hypothesis is Benthocodon may not be feeding in the early afternoon when observations are made. Gut contents were found in all five specimens collected during night dives but in only one collected during dives made in the late mornings and early afternoons. This suggests a rhythmical feeding pattern that may 15 coincide with the downward and upward migrations of zooplankton. Benthocodon would then be expected to feed at night and early morning. Therefore, all the specimens collected later in the day would lack gut contents, as observed in two of the three fixed on board ship. This hypothesis, however, needs to be further tested with dives made at different times of the day. This study was limited by the number of specimens that were examined for both gut contents and tentacle structures and the small amount of time in behavioral observations using the submersible. The optimal conditions for maintaining Benthocodon pedunculata were unknown and the handling of medusae during collection may have stressed the specimens. Morphology revealed by microscope studies may also be biased by preparation of these fragile, gelatinous samples. Regardless of these possible sources of error, this study provides a description of the morphological features of this benthopelagic medusae that were previously unexamined. Postulations about their relationship to function requires further behavioral observations. This research has contributed new information about the structural features of this benthopelagic medusae that are found in high density patches in the Soquel and Carmel Canyons (Larson et al. 1992). The near bottom environment and the organisms that thrive there are virtually unknown. It is fascinating to study the morphological adaptations of benthopelagic medusae to this environment and to explore their interactions with other deep sea organisms, as more information is collected using ROVs. Information about these survival mechanisms lead to an understanding of the 16 interactions between organisms of the deep-sea community, of which so little is known. Figure Legends Figure 1. Benthocodon pedunculata medusa as seen with the ROV. Figure 2. Photo of a medusa of Benthocodon spp showing three different tentacle types. (kindly provided by G. I. Matsumoto) Figure 3. View of the entire subumbrella of a Benthocodon pedunculata medusa. G.C., gastrovascular cavity; L., lips; S., stomach; T., tentacles. (kindly provided by G. I. Matsumoto) Figure 4. Inner and outer tentacles of a medusa of Benthocodon pedunculata E.U., exumbrella; S.U., subumbrella. (kindly provided by G. I. Matsumoto) Figure 5. Outer tentacles of a medusa of Benthocodon pedunculata A.P., abscission point. (kindly provided by G. I. Matsumoto) Figure 6. Gut Content of a specimen of Benthocodon pedunculata (collected by JSL3170 in detritus sampler (DS) 9 at 2779 ft) s howing amphipod exoskeleton. Figure 7. Gut Content of a specimen of Benthocodon pedunculata (collected by JSL3170 in DS-9 at 2779 ft) showing a pycnogonid exoskeleton. Figure 8. Tentacle margin of a medusa of Benthocodon pedunculata E.U., exumbrella; S.U., subumbrella. (kindly provided by G. I. Matsumoto) Figure 9. Scanning electron micrograph of the edge of the subumbrella and tentacle margin of Benthocodon pedunculata showing uneven rows of tentacle bases. S.U., subumbrella; T.S., tentacle stem. (Bar represents 500 u) Figure 10a. Scanning electron micrograph showing the annulated surface of an inner tentacle of Benthocodon pedunculata. N.R., nematocyst rings. (Bar represents 50 u) 17 Figure 1Ob. Scanning electron micrograph of the tip of an inner tentacle of Benthocodon pedunculata . N.R., nematocyst rings. (Bar represents 5 u) Figure 11a. Section along the longitudinal axis of an inner tentacle of Benthocodon pedunculata. (1 u thick section) Stained with methylene blue and basic fuchsin. Oil immersion phase contrast 100x. E.C.T., ectoderm; M.E., mesoglea; E.N., endoderm; M., muscle; D.I., digestive inclusions; N., nematocyst; I., interstitial cells. (Bar represents 10 u) Figure 11b. Section along the longitudinal axis of the distal end of an inner tentacle of Benthocodon pedunculata. (1 u thick section) Stained with methylene blue and basic fuchsin. Oil immersion phase contrast 100x. N., nematocyst; I., interstitial cells; M., muscle; E.N., endoderm. (Bar represents 10 u) Figure 12a. Section along the longitudinal axis of an outer tentacle of Benthocodon pedunculata. (1 u thick section) Stained with methylene blue and basic fuchsin. Oil immersion phase contrast 100x. E.N., endoderm; M.E., mesoglea; N., nematocyst; I., interstitial cells. (Bar represents 10 u) Figure 12b. Section along the longitudinal axis of an outer tentacle of Benthocodon pedunculata. (1 u thick section) Stained with methylene blue. Oil immersion phase contrast 100x. E.N., endoderm; N., nematocyst. (Bar represents 10 u) Figure 13. Graph of the lengths and widths of 391 nematocysts found on the inner and outer tentacles of Benthocodon pedunculata. They fall into 2 categories, showing 2 different nematocysts. Figure 14a. Extruded modified-aspirotele nematocyst on an outer tentacle of Benthocodon pedunculata . Oil immersion phase contrast 100x. M.A., modified-aspirotele nematocyst; S., shaft; T., thread. (Bar represents 10 u) Figure 14b. Extruded modified-stenotele nematocyst on an outer tentacle of Benthocodon pedunculata . Oil immersion phase contrast 100x. M.S., modified-stenotele nematocyst; S., shaft; T., thread. (Bar represents 10 u) 18 Figure 15a. Scanning electron micrograph of fired modified¬ aspirotele nematocyst of an outer tentacle of Benthocodon pedunculata. M.A., modified-aspirotele; S., shaft; T., thread. (Bar represents 5 u) Figure 15b. Scanning electron micrograph of fired modified-stenotele nematocyst of an outer tentacle of Benthocodon pedunculata. M.S., modified-stenotele; S., shaft; T., thread. (Bar represents 5 u) Figure 16a. Scanning electron micrograph of fired modified¬ aspirotele nematocyst of an outer tentacle of Benthocodon pedunculata. M.A., modified-aspirotele. (Bar represents 5 u) Figure 16b. Scanning electron micrograph of fired modified-stenotele nematocyst of an outer tentacle of Benthocodon pedunculata. M.S., modified-stenotele; S.P., spicules; T., thread. (Bar represents 5 u) Table Legends Table 1. Collection Dives. Table 2. Number of recognizable prey in gut content analysis of 13 specimens of Benthocodon spp. Acknowledgements I would like to thank the Point Lobos crew and ROV pilots for collecting the jellies, Chuck Baxter for his helpful advice and for taking me out on collecting trips with the ROV. I am grateful to Chris Patten and Kurt Buck for their microscope expertise, Alan Baldridge and Susan Baxter for helping me find obscure journals and books, and Molly Cummings for teaching me how to be a real marine biology student. Thanks to the many spring students who suffered through seasickness with me on the Point Lobos and Mamp;M fever on Thursdays. Most importantly, I thank George I. Matsumoto for his 19 profound insights to my project, numerous advice, critical readings of my paper, volleyball and walleyball coaching, many encouragements, and for surpassing all my expectations of what a mentor could be. Literature Cited Barnes, R. D. 1987. Invertebrate Zoology. 5 ed. CBS College Publishing, New York. pp 96. Bigelow, H. B. 1913. Medusae and siphonophorae collected by the U.S. Fisheries steamer "Albatross" in the northwestern Pacific, 1906. Proc. U.S. Natn. Mus. 44(1946):1-119. Costello, J. 1992. Foraging mode and energetics of hydrozoan medusae. In Aspects of hydrozoan biology. (J. Bouillon, F. Cicogna, J. M. Gili, and R. G. Hughes, editors) Sci. Mar., 56(2-3): 185-191. Gage, J. D. and P. A. Tyler. 1991. Deep Sea Biology: A Natural History of Organisms at the Deep-Sea Eloor. Cambridge University Press, Cambridge. pp 126. Harrison, F. W. and J. A. Westfall. 1991. Microscopic Anatomy of Invertebrates, Placozoa, Porifera. Cnidaria, and Ctenophora. Vol. 2. Wiley-Liss, Inc., New York. pp 91-183. Hesthagen, I. H. 1971. On the biology of the bottom-dwelling trachymedusaTesserogastria musculosa Beyer. Nor. J. Zool. 19: 1-19. Hufnagel, L. A., G. Kass-Simon and M. K. Lyon. 1985. Functional organization of battery cell complexes in tentacles of Hydra attenuata. J. Morphology. 184: 323-341. Kelly Michael. 1992. Distribution & abundance of crustacean zooplankton in the benthic boundary layer of Monterey Bay. Class paper, Biology 164H. Summer 164H. Hopkins Marine Station. Kramp, P. L. 1935. Polypdyr. I. Ferskvandspolypper og Goplepolypper. Danm. Fauna 41:1-208. 20 Larson, R. J., C. E. Mills, and R. Harbison. 1989. In situ foraging and feeding behavior of Narcomedusae (Cnidaria:Hydrozoa). J. mar. biol. Ass. U. K. 69: 785-794. Larson, R. J. and G. R. Harbison. 1990. Medusae from McMurdo Sound, Ross Sea, including the descriptions of two new species, Leuckartiara brownei and Benthocodon hyalinus. Polar Bio. 11: 19-25. Larson, R. J., G. I. Matsumoto, L. P. Madin and L. M. Lewis. 1992. Deep-sea benthic and benthopelagic medusae: Recent observations from submersibles and a remotely operated vehicle. Bull. of Mar. Sci. 51(3): 277-286. Laverack, M. S. and J. Dando. 1987. Lecture Notes on Invertebrate Zoology. 3 ed. Blackwell Scientific Publications, Palo Alto, CA. PP 30. Marshall, A. J. and W. D. Williams. 1972. Textbook of Zoology. Invertebrates. Macmillan Press Limited, New York. pp118. Marsical, Richard N. 1974. Nematocysts. In Coelenterate Biology. Reviews and Perspectives (Leonard Muscatine and Howard M. Lenhoff,editors) Academic Press, Inc., New York. pp.129-178. Mayor, A. G. 1910. Medusae of the World, The Hydromedusae. Vol.1. Carnegie Institution of Washington, Washington, D. C. pp 12-14. Mills, C. E. 1981. Diversity of swimming behaviors in Hydromedusae as related to feeding and utilization of space. Mar. Bio. 64: 185-189. Petry, Carolin. 1992. Zooplankton of the benthic boundary layer of Monterey Canyon. Class paper, Biology 164H. Summer 164H. Hopkins Marine Station. Purcell, J. E. 1984. The functions of nematocysts in prey capture by epipelagic siphonophores (Coelenterata, Hydrozoa). Bio. Bull. 166: 310-327. Purcell, J. E. and C. E. Mills. 1988. The correlation between nematocyst types and diets in pelagic hydrozoa. In The Biology 21 of Nematocysts. (D. A. Hessinger and H. Lenhoff, editors) Academic Press, New York. Thomas, M. B. and N. C. Edwards. 1991. Cnidaria: Hydrozoa. In Microscopic Anatomy of Invertebrates. Vol2. Placozoa. Porifera, Cnidaria, and Ctenophora (Harrison, F. W. and J. A. Westfall, editors). Wiley-Liss, Inc., New York. pp 91-183. Westfall, J. A. 1970. Ultrastructure of synapses in a primitive coelenterate. J. Ultrastr. Res. 32: 237-246. Wishner, K. F. 1980a. The biomass of the deep-sea benthopelagic plankton. Deep-sea Res. 27(A): 203-216. Wishner, K. F. 1980b. Aspects of the community ecology of the deep sea, benthopelagic plankton, with special attention to gymnopleid copepods. Mar. Bio. 60: 179-187. 22 a K 55.4 EU Hg. 14 + 12 10 + 8 6 O 2 4 Nematocyst Lengths and Widths 6 8 10 12 14 16 length-(u) Fg. H4.1 Ra e