Foraging Feeding and Digestion of the Midwater Narcomedusa Solmissus spp. Anne H. Reese Biology 175H, Hopkins Marine Station, 1992 Permanant Address Abstract: Solmissus spp, a narcomedusa found commonly throughout the midwater region from the surface to 1000m may play an integral role in the midwater food web. Foraging feeding and digestive studies in situ and in the laboratory indicate that Solmissus has adapted to living in a low prey density environment by behaviors which increase prey encounter time and area. These behaviors include extending tentacles and swimming while foraging resuming foraging while still in the ingestive process; eating a very large range of prey sizes; and ingesting more than one prey item at a time. Thirteen Solmissus spp. were collected using a Remote Operated Vehicle in the Monterey Bay Submarine Canyon. In the laboratory, these specimens were kept in tanks and fed various shallow and deep water flora and fauna. In the laboratory, Solmissus preyed on Beroe abyssicola and Nanomia biyuga which they digested in 36 hours and 24 hours, respectively. Introduction: Although common in many parts of the world and throughout the water column, medusae have rarely been studied compared to other marine organisms. In fact, many of the fundamental aspects of their behavior and interactions have not been explored because they are fragile, inaccessible, and difficult to keep alive in a laboratory. Undamaged deep-sea medusae, especially, have been inaccessible until recently with the technology of submersibles and Remote Operated Vehicles (ROVS). As early as 1879, Haeckel (1879), Agassiz (1902), Bigelow (1909), and Mayer (1910) described midwater medusae from their trawl nets; however they could record no behavioral information from the dead and damaged specimens. Researchers have obtained very little information in regards to: What is the midwater food web, and how do medusae fit into it? Predator-prey relations represent the major interaction of marine organisms. Subsequently, in accessible areas such as the intertidal, subtidal, and surface ocean habitats, researchers have spent many years understanding the complex food webs of the vertebrates and invertebrates. A growing body of studies has explored basic feeding habits and interactions of shallow water gelatenous creatures. (e.g. Madin, 1988; Purcell, 1983; Biggs 1977, Reeve et al., 1978; Miller and Williams, 1973). However, very few researchers have ventured into the midwater zone extensively enough to gather the same amount of information on the feeding of deep-water gelatinous creatures. Specifically, Larson, Mills, and Harbison (1989) have looked at the general feeding and foraging of narcomedusae; Mills and Goy (1988) have recorded migrations of the Solmissus albescens and some behavior, Purcell and Mills (1988) have identified the correlation between types of hydrozoan nematocysts and prey; and Larson (1979) has described the feeding of coronate medusae. Solmissus, a narcomedusa, has been observed from the surface in upwelled waters to 975m in the Monterey Bay Submarine Canyon. Because Solmissus is in such abundance in the water column, it is natural to assume that they form an integral part of the mid-water food web. While there have been some in situ observations of Solmissus with prey in its guts, in depth information on the feeding habits and digestion of Solmissus does not exist. Solmissus catches prey by firing nematocysts of the apotrichous isorhizas type (Purcell and Mills, 1988) when the prey touches the animal's tentacles. Öften the prey struggles to free itself, but if Solmissus keeps hold of the prey, its tentacles bend along a peronial groove toward the mouth (Larson et al. 1989). The mouth, usually flat along the gut, begins to flare out toward the tentacles in order to engulf the prey. When the lips contact the prey, cilia lining the lips and mouth bring the prey into the mouth and in toward the gut (Larson et al. 1989). Once the prey is inside the mouth, the lips pinch closed, securing the prey inside the mouth and gut. Solmissus moves the prey item towards the edge of the gut to the stomach pouches where the digestion is hypothesized to take place. The gut and mouth are made from a thin layer of tissue. One of the difficulties of feeding studies, then, is that the fragile gut tends to void itself when handled or exposed to rough current (Mills and Goy, 1988). The tentacles also tend to break off with rough handling, a condition which severely limits laboratory feeding experiments. With the advent of submersibles and ROV's capable of capturing specimens without much physical contact and the plankton kriesel tank which can recreate a deep-sea current (Hamner, 1990), researchers can conduct experiments on the midwater gelatinous animals in laboratory environment which resemble the mid-water conditions. The information that exists on the feeding behavior of Solmissus is scattered and comes from several researchers' in situ observations. On ROV dives, Monterey Bay Aquarium Research Institute (MBARI) researchers have seen a Solmissus feeding on a Beroe (P. Ctenophora) and on a Nanomia (P. Cnidaria). Mills and Goy (1988) reported the pteropod Cavolina in the gut of Solmissus albescens, (Larson et al, 1989) saw comb rows of an unidentified ctenophore in the gut of a S incisa, and Mackie and Mackie (1963) saw S. marshalli feed on the hydromedusa Euphysa. However, these observations are rare; the vast majority of observed animals have had empty guts. This study was undertaken to clarify the feeding behavior of the narcomedusa, Solmissus sp. Because the midwater may have few and widely separated potential prey items, I hypothesized that Solmissus must be an opportunistic predator. In order to survive in a habitiat with low prey density, Solmissus must exhibit behaviors which take advantage of as much prey-encounter time as possible. 1 proposed a set of behaviors which represent general adaptions of an opportunistic Solmissus: (1) They would be able to swim while fishing with tentacles (2) They would be able to eat a wide size range of prey items; (3) They would release their tentacles into feeding position as soon as the prey was secure in the mouth; (4) They would resume swimming as soon as the prey was secure; (5) They would feed on a variety of types of prey (6) They would be able to ingest a second prey item while digesting a previous capture. I carried out in situ observations and conducted feeding experiments on Solmissus held in kriesels to determine these behaviors. Methods: The 13 Solmissus specimens studied in the lab were collected by the ROV Ventana on 8 dives in the Monterey Bay Submarine Canyon. The specimens were collected in detritus samplers which were kept and sealed until brought into the laboratory. Four of the ROV dives were at latitude 36 42°22“, longitude 122 03’30“, one was at lat. 36 44·56", long 122 0401", one was at lat 36 47’12“, long. 122 0232", and two at lat 36 42’43", long.122 03115' (see table 1.) The ROV also caught the Nanomia biyuga and Paraphyllina ransoni? The brine shrimp and Beroe abyssicola were provided by Monterey Bay Aquarium, and the Fleurobrachia bachei from labs in Santa Barbara. The Solmissus specimens were kept together in one plankton kriesel tank with a circulating current (Hamner, 1990.) The number of Solmissus in the tank during the period of study varied from 1 to 4. The prey items were kept together in plastic containers of chilled seawater until the time of the experiments. The tank and plastic containers were located at the Moss Landing Marine Operations Station of MBARI (Monterey Bay Aquarium Research Institute) in a dark, temperature regulated room at 5 degrees Celsius. The water that circulated through the kriesel was kept at 6 degrees Celsius, and the salinity was checked every day to keep a constant 345 parts/thousand. The animals were observed under red light and a white flashlight. To collect data, I utilized videos from past ROV dives and their preserved specimens, observed video transmissions from the RÖV on the ship the Point Lobos, and kept tanks with live collected specimens. From the MBARI video library, I analyzed a catalog of Solmissus interactions with other deep sea animals, and watched selected video series. 1 went to sea on 6 cruises in order to make in situ observations and collect Solmissus and potential prey. On each dive, the RÖV was down for approximately 5.5 hours making my total shipboard observations 33 hours. The tanks were checked during at least one period per day for 36 days. My research was a total of 52 days, but the tanks were not checked on days when (1) there were no Solmissus specimens, (2) the specimens had unusable tentacles, or (3) the specimens had no gut. When I conducted the feeding experiments, I checked the tanks at 20 minute intervals lasting from 1-20 minutes depending on the activity level of the animals. Each observation period lasted from 1-36 hours depending on the activity of the animals. In the experiments, I attempted to feed Solmissus brine shrimp, Nanomia biyuga, Beroe abyssicola, a dead Paraphyllina ransoni? small pieces (25 inches long) of Macrocystis blades, and Pleurobrachia bachei. To identify the species of the Solmissus specimens, I used the standard reference materials on medusae by Russell (1953), Bigelow (1909) and Mayor (1910). Results: From the ROV dives, I received a total of 13 Solmissus specimens (see table 1,2.) All were caught using the RÖV detritus samplers, and the physical condition of each was very good with the exception of three specimens which were damaged prior to collection (dives on 4/13/92, 4/23/92, and 5/28/91.) The number of specimens collected at each dive site is in table 1. In the plankton kriesels, Solmissus tended to circulate on the bottom right-hand side of the tank. However, the smaller animals would circulate with the current around the whole tank and often scrape against the filter. When circulating the whole tank, animals would often slide along the bottom of the tank and pick up detritus in their tentacles. Two of the specimens which had bells of 4 inches diameter were too big for the kriesel tank and were preserved during the first few days after capture. Three which were damaged prior to capture and preserved after capture. The eight remaining Solmissus specimens remained in condition adequate for experiments between 6 and 14 days. I determined that animals were adequate for feeding experiments if they had five or more tentacles. As Solmissus became unhealthy, the tentacles broke off, the bell lost its shape becoming flaccid, and the animal lost its ability to swim. However, the animals continued to pulse on the bottom of the tank for several days even when shapeless and unable to swim. Solmissus behaved the same in the red light or under a white flashlight. Observations: Of the 103 sightings of Solmissus in the video catalog 23 either had a prey item in the gut, in the tentacles, or in the mouth. Four of these were preying on Beroe, three on Cydippid, two on Nanomia, one on Pleurobrachia, one on a Euphausiid, nine on unidentified ctenophores, two on unidentified siphonophores, and two on unidentified gelatinous creatures. Two of these video segments included Solmissus actually feeding on Heurobrachia and Nanomia. Of two previously preserved specimens, one had a Pleurobrachia in the gut. On the ship, I observed Solmissus from 200m to 926m. All of the Solmissus observed had empty guts. Approximately one half were swimming with tentacles extended and the other half were immobile with tentacles extended. Most were in feeding postures and did not change from these postures unless the current from the RÖV disturbed the animal. From my observations in the lab, I identified the stereotypical feeding pattern of the Solmissus. Half the time it incorporated swimmin, periodically, stopping and floating with tentacles extended in one of four feeding positions. The other half it swam while extending its tentacles (see figured 1a,b,c,d,e.) Solmissus swims with the bell in front by contracting the velum and trailing the tentacles behind the disk. A further laboratory observation was that when stressed by touch with the rim of a glass jar, handling or change of tank Solmissus bioluminesced a purple-blue color for a few seconds. Two separate animals seemed to leak bioluminescent droplets into the tank when they were stressed. I identified 4 of the Solmissus specimens as S. incisa, but the other 11 specimens did not correspond to the descriptions of previously described Solmissus species in literature. There were two groups of Solmissus in the unidentified 11: Three specimens with a purple ring around the edge of the lappets (here as species 1,) and the other 7 which were colorless (here as species 2) Feeding Experiments: The animals did not attempt to eat Macrocystis kelp blades even when the kelp connected with the tentacles. WhenSolmissus caught a dead Paraphyllina ransoni?, in the tentacles, it began the feeding behavior of puckering the lips to surround the food each of three times when it caught the medusa (see figure 2,) but did not ingest it. Additionally, when Solmissus caught Fleurobrachia bachei, Solmissus began to pucker the lips, but could not hold onto the prey with its damaged tentacles. I was able to feed four Solmissus live, gelatinous food. Two Solmissus, one S. incisa with a bell 2.5 inches in diameter and one species 2 with a bell 1.5 inches in diameter, ate Nanomia biyuga 3.5 inches and 4.5 inches in length, respectively. Each digested the N. biyuga in 36 hours. Two different S. incisa, one 2 inches and one 1 inch in diameter, each ate Beroe abyssicola. The larger S. incisa ate a B. abyssicola of diameter 25 inches and length 5 inches, and the smaller Sincisa ate a B. abyssicola with the dimensions 5 inches and 75 inches. Each digested B. abyssicola in 24 hours (see table 2.) While digesting the B. abyssicola, both of S. incisa caught a N. biyuga and ingested the zooids of the siphonophore (the nectophores and pneumatophore were not ingested). In the beginning of the ingestion period the swimming of Solmissus was inhibited. When Solmissus ingested the N. biyuga prey, it directed the ingested zooids to the edge of the gut to the stomach pouches while the rest of the prey was still being ingested. With B. abyssicola, Solmissus generally positioned the prey slightly off center of the gut after ingestion so that one side of B. abyssicola contacted the stomach pouches. Most of the digestion and decomposition of B. abyssicola took place in the flaccid, distended gut (see figure 3.) During ingestion of both the N. biyuga and the B. abyssicola, the Solmissus ceased its swimming and used between 3 and 5 tentacles to hold the prey and direct the prey toward the lips and mouth. When the prey was partially but securely surrounded by the lips,Solmissus released all of the tentacles except one which remained holding the prey in place and resumed swimming. Frequently, Solmissus would stop and position the released tentacles resume one of the four feeding postures (see figure la, b, c, d.) Less frequently, Solmissus would swim with 2-3 tentacled stretched out above the bell. Finally, when Solmissus ingested the pneumatophore of N. biyuga, it separated the pneumatophore from the rest of the body of N. biyuga once in the gut. During digestion, this gas-sac moved up toward the pinched-shut mouth of the Solmissus (see figure 3.) Slowly, traces of the sac disappeared. In addition, pieces of broken-up comb rows from Beroe seemed to follow the same pattern of moving up to the pinched mouthand also slowly disappeared. Discussion: There are several possibilities as to why the Solmissus specimens stayed healthy enough to do experiments with for less than two weeks. Four main ones could include: (1) damage from the kriesel tank, (2) close proximity with other Solmissus specimens in the kriesel, (3) inadequate pressure, and (4) contaminated water. Öften, while circulating with the current in the kriesel, Solmissus would bump into the tank filter and scrape along the rough surface possibly incurring physical damage to the bell and/ or tentacles. Solmissus sometimes scraped along the bottom the tank picking up detritus which could weigh down the tentacles and break them. Additionally Solmissus frequently swam into the sides and bottom of the tank, again possibly physically damaging itself. These hypotheses would explain the fact that the first signs of an unhealthy specimen was broken tentacles. However, these speculations are made without knowledge of the natural life-span of Solmissus or the age of the captured specimens. Second, because often several Solmissus specimens were in the tank together, they would come in contact with each other's tentacles and sometimes become entangled. Not only could this break off some of the tentacles, but I observed one animal become entangled in another’s tentacles and rip out its own gut while struggling free. (Interestingly, this animal subsequently in 5 days time had fully regenerated its gut. This is consistent with Claudia Mills' report at a meeting at the Monterey Bay Aquarium Jellyfest (4/11/92) report that Solmissus will regenerate damaged parts of its body in situ as long as it can get adequate nourishment.) Third, the kriesels were not pressurized, thus keeping Solmissus under substantially less pressure from the midwater environment. Although Solmissus seemed unaffected when brought up to the surface, perhaps the change in pressure had more long-term effects. For instance, because the bell deflated in older animals, the loss of pressure may have enabled the mesoglea in the bell to leak out. i.e. In situ the pressure may keep the mesoglea in the bell secure within the exoderm. Fourth, the kriesel was part of a circulating sea water system which recycled the water through 3 tanks containing other midwater species and the water quality may have deteriorated in the closed system. It has been hypothesized that because the prey density is very low in the midwater environment, many medusae spread their tentacles and swim in order to increase the area in which it can contact potential prey items (Larson et al. 1989; Madin, 1988). The observed feeding behavior of swimming both in conjunction with and alternating with feeding postures of Solmissus (see figures la, b,c, d, e) is consistent with this hypothesis. In each of the feeding postures, the tentacles are extended wide and increase the volume of water swept while swimming to touch any nearby prey. The active swimming enables Solmissus to forage in different areas in the midwater. It is advantageous for Solmissus to move around because the prey density in the water column is low and also patchy; thus, the probability of encounter in one area is proportionally small. Solmissus additionally decreases the amount of time taken away from fishing for prey by sometimes swimming with tentacles extended above the bell for prey capture with food already in the gut. Solmissus proved to be carnivorous in keeping with studies done on other medusae (Larson, 1979). They did not eat the dead food they came in contact with. The reason they did not eat it could have been that the P ransoni? was not part of their diet; however, because they did pucker their lips and begin moving the mouth out towards the prey, it seems as if they would have consumed the medusa had it stayed in the tentacles. However, each time the contact occured, within 10-20 minutes, the Solmissus released the medusa. Because P ransoni? was dead, the release could not have happened from a struggle as in cases with N biyuga. I did not attempt to feed Solmissus dead Beroe or Nanomia. Madin (1988) suggested that tentaculate predators select their prey by stimulation of the tentacles after the prey is caught. Madin related this in the context of very strong stimulation by the animal touching the tentacles may imply another animal preying on the medusa and trigger an escape response rather than a feeding response. Less strong stimulation would trigger the feeding response. In the case of the dead food, it is possible that in absence of a stimulation or tugging from the prey, the Solmissus aborts the feeding response and releases the dead prey. Because there is much marine snow and detritus in the water column which Solmissus could come in contact with during foraging the animal must have means to distinguish prey from organic waste. It would not be advantageous for Solmissus to fire the nematocysts when it contacted each piece of detritus. The animal would both expend energy in the encounter and waste foraging time and energy producing more nematocysts. The other gelatinous food which Solmissus did not ingest in the kriesel was live Heurobrachia bachei. Although I had seen Feurobrachia in the gut of Solmissus in video and preserved specimens, the Solmissus I had in the tank at the time Pleurobrachia bachei was available was too unhealthy to feed. Most of the tentacles of Solmissus had broken off and the remaining two were not adequate to hold onto the prey while the mouth surrounded the food. Each of the five times that the tentacles of Solmissus contacted P bachei, Solmissus began the feeding behavior of puckering the lips and extending the mouth towards the prey item. However, each time the P bachei escaped by jerking its body away from Solmissus. It has been assumed in the past that digestion takes place in the stomach pouches, (Larson et al., 1989) but in cases of the digestion of the B. abyssicola, there is more of a question of what is the specific function of the stomach pouches as the prey's body decomposed mostly in the gut. For instance, digestive enzymes may come from the stomach pouches into the gut to digest the prey there, or the stomach pouches may be the mode by which the animal absorbs the digested food. Their physiological function is yet unclear. The physiology of excretion in Solmissus is also unclear. No studies have been done regarding how Solmissus discharges undigestible material from its body. The observations about the pneumatophore and the comb rows in the digestion of N. biyuga and B. abyssicola imply that one mode may be pushing the excrement out the pinched mouth during digestion. In addition to the efficient foraging behavior, during the ingestion and digestion studies, Solmissus exhibited the five behaviors which reflect its need to be an opportunistic predator in the midwater environment where prey density is low: (1) They were able to eat extremely large prey relative to their own size; (2) They released their tentacles into feeding position as soon as the prey was secure in the mouth; (3) They resumed swimming as soon as the prey was secure; (4) They extended several foraging tentacles while swimming with prey already in gut, and (5) They were able to ingest a second prey item while digesting a previous capture. In the case of the 1 inch S. incisa ingesting the 75 inch/ 5 inch B. abyssicola, S. incisa consumed a prey item almost as large as itself. The tissue of gut, mouth and lips have adapted to be flexible enough to engulf prey close to the size of Solmissus. By being able to distend the gut so far, the Solmissus increases both the size apectrum and meal size of prey available in relation to the animals it encounters in the midwater. My experiments suggest that in order to increase its chances of opportunistic feeding, the Solmissus returns to its feeding postures and foraging as soon as physically possible after feeding. Even before the prey has been fully engulfed by the mouth and gut, the Solmissus resumes its fishing. Then, if it contacts another prey item in its foraging the body has adapted so that it can continue to digest while ingesting. Perhaps the motion of cilia which brings the prey into the mouth from the tentacles keeps pushing the digesting food farther into the gut so that more food can enter without letting the digesting food escape out the flared mouth. On several occasions in video and in the laboratory while resuming swimming Solmissus would extend 2-3 tentacles above the bell to catch additional prey. These abilities enable Solmissus to take advantage of opportunities to feed at all times except the short periods in which it secures the prey in the mouth and lips. Acknowledgements: I would like to thank: • my advisors Bruce Robison and Chuck Baxter for sharing not only their enthusiasm but their knowledge of and experience with research of the midwater, •George Matsumoto for his infinite time, patience, suggestions, and willingness to be a sounding board for both strange and serious ideas; • Kim Reisenbichler for help in all the technical aspects of the project; ethe Foint Lobos ROV pilots Jim, TC, and Chris for working so hard to and doing such an excellent job of capturing undamaged animals for my studies •jena Hickey, for being there, .at all times •Alex Cronin for support, constructive pushing and critique; •Subaru. The funding for this project came from Anagnostopoulos Inc. (aka thanks Mom and Dino.) References: Agassiz and Mayer, Memoirs of the Museum of Comparative Zoology at Harvard College vol. XXVI, 1902. Cambridge, USA: University Press. Bigelow, "Medusae," in Menoirs of Comparative Zoology at Harvard College vol.XXXVII, 1909. Cambridge, USA: University Press. Biggs, D.C, 1977. Respiration and ammonium excretion by open ocean gelatinous zooplankton. Limnology and Oceanography, 22, 108-116. Hamner, WM, 1990. Design developments in the Planktonkriesel, A Plankton Aquarium for ships at sea. Journal of Plankton Research, 12, 397- 402. Haeckel, Das System der Merdusen, 1879. Jena: Gustav Fisher. Larson, RJ., 1979. Feeding in coronate medusae (Class Scyphozoa, Order Coronatae). Marine Behaviour and Physiology, 6, 123-129. Larson, R.J.; Mills, CE. and Harbison, G.R., 1989. In situ foraging and feeding behaviour of narcomedusae (Cnidaria: Hydrozoa). Biological Assosciation of the United Kingdom, 69, 785-794. Mackie, G.O. and Mackie, G.V., 1963. Systematic and biological notes on living hydromedusae from Puget Sound. Bulletin. National Museum of Canada, no. 199, 63-84. Madin, LP., 1988. Feeding behavior of tentaculate predators: in situ observations and a conceptual model. Bulletin of Marine Science, 43, 413-429. Mayor, A.G., Medusae of the World, the Hydromedsae vol 2, 1910. Washington D.C.: Carnegie Institute of Washington. Miller, R.J. and Williams, R., 1973. Energy requirements and food supplies of ctenophores and jellyfish in the Patuxent River estuary. Chesapeake Science, 13,328-331. Mills, C.E. and Goy, J., 1988. In situ observations of the behavior of mesopelagic Solmissus narcomedusae (Cnidaria: Hydrozoa). Bulletin of Marine Science, 43, 739-751. Purcell, J.E, 1983. Digestion rates and assimilation efficiencies of siphonophores fed zooplankton prey. Marine Biology, 73, 257-261. Purcell, J.E, and Mills, C.E., 1988. The correlation between nematocyst types and diets in pelagic hydrozoa. In The Biology of Nemtocysts (ed. D.A. Hessinger and H. Lenhoff), pp. 463-485. San Diego: Academic Press. Reeve, M.R.; Walter, M.A; and Ikeda, T., 1978. Laboratory studies of ingestion and food utilization in lobate and tentaculate ctenophores. Limnology and Oceanography, 23, 740-751. Russell, F.S., The Medusae of the British Isles, 1953. Cambridge: University Press. Figure Legends: gure 1: Feeding positions of Solmissus a) tentacles stretched to side 1. tentacles, 2. velum, 3. bell. b) tentacles stretched above bell. c) tentacles stretched above and coiled. d) tentacles stretched both to the side and above. e) several tentacles stretched above bell while swimming. Figure 2: Ingestion behavior of Solmissus. 1. tentacles, 2. velum, 3. bell, 4. puckered mouth and lips, 5. Nanomia. Figure 3: Digestion of Solmissus. 1. tentacles, 2. velum, 3. bell, 4. distended gut, 5. stomach pouches, 6. ingested prey. S C V 6 oooooooooooo 8 3- 8 0 8 — NN — — e — OOUNN — — e ON So NN ONN N o S 6 8 — — o- ———— — —— ONNNNN ONNONSSA C oooooC anadaaaa — ooo oNo DNO N 3c oooo 3 3 3 3 3 3 3 3 X — ooooooooooooo 3 - 8 3888 (D 8888 NNN — NN —— —N NNNO oouooopado 1— — 3- 33383335 2 Oo 88 5