Foods, feeding, & growth of the medusa are considered in this section, while the POLYP is considered in another section.

Hydromedusa, possibly Aglantha digitale, with its
manubrium stuffed with copepod prey 2X

A study of feeding behaviour of several species of hydromedusae at Friday Harbor Laboratories, Washington suggests that these particular species do not actively catch their food as has been described for other species; rather, they seem to depend on chance contact with their prey.  Contact with food elicits grasping action in the tentacles, followed by a bending of the bell margin towards the manubrium, which likewise bends to make contact with the manubrium.  Foods generally include small crustaceans and worms.  By use of carmine dye added to suitable food material such as fishes or mussels, the author shows that within 30min food matter has moved through the entire gastrovascular system and is being ingested by the gastrodermal cells (see drawing).  Some species, such as Sarsia mirabilis, have an expanded region near the manubrial tip sometimes referred to as a stomach. In most other medusae the stomach is an expanded region in the summit of the bell.  Hyman 1940 Biol Bull 79: 282; for a general review of hydromedusae as predators on a global scale (but including some west-coast species) see Mills1995 J Mar Sci 52: 575.

NOTE  species used are Aequorea aequorea, Halistaura cellularia, Clytia gregaria (Phialidium gregarium), Somotoca atra, and Sarsia mirabilis

Collections of hydromedusae in Yaquina Bay, Oregon during April-November yield 11 species, with Clytia gregaria (Phialidium gregarium), Sarsia eximia, and Polyorchis penicillatus being among the most common.  Most abundant prey items in the gastrovascular cavities of these 3 species are several species of copepods, other unidentified crustaceans, eggs, and various invertebrate larvae.  The predominant copepod eaten is Arcartia clausi, which is the most abundant species in the plankton of the Bay at the time of the collections. In general, prey eaten by the medusae correspond with components in the plankton, indicating little or no selective feeding.  McCormick 1969 Northwest Science 43: 215.

Medusae of the 2-tentacled hydroid Proboscidactyla flavicirrata are locally abundant in coastal waters of the San Juan Archipelago, Washington from May-September.  Laboratory and field observations at the Friday Harbor Laboratories, Washington reveal the following behaviours.  When feeding, medusae sink slowly downwards with tentacles spread, a behaviour known as “sink-fishing”.  Prey items are caught up by single tentacles or groups of tentacles and transferred to the manubrium.  If the prey is small the manubrial lips engulf and strip the tentacles off the prey; if large, the prey is passed directly to the mouth.  Prey includes small adult copepods and cladocerans, fish and invertebrate eggs, and various larvae such as zoeae, trochophores, and nauplii. Occasionally when swimming a medusa will “crumple”, a defensive response involving mass contraction of contractile tissues that infolds the umbrella margin and tentacles towards the manubrium.  If all radial muscles contract symmetrically, then the medusa takes on a square outline when seen from above (see drawings).  When crumpled the medusa is motionless and sinks slowly downward in the water column.
Spencer 1975 Biol Bull 149: 236.

Studies of the life cycle of the hydroid Clytia gregaria (Phialidium gregarium) in Saanich Inlet, British Columbia reveal significant and interesting community interactions based on intensive feeding by the medusae.  Plankton blooms in springtime (April-May) provide food for growth of overwintering polyps of Clytia and lead to blooms of medusae.  The increased predation by medusae on the herbivorous zooplankters reduces numbers of the latter and leads to a second bloom of phytoplankton in late spring/early summer (Jun-Jul).  However, with the decrease in zooplanktonic food leading up to this second phytoplankton bloom, medusae numbers decline.  With more food and fewer predators, zooplankton numbers rebound.  Huntley & Hobson 1978 J Fish Res Bd Can 35: 257.

Laboratory observation of feeding in the hydrozoan medusae Stomotoca atra, Eutonina indicans, and Aequorea victoria at the Pacific Biological Station, British Columbia reveals selective predation on each other and on different species of hydrozoans, scyphozoans (Aurelia aurita), and ctenophores (Pleurorbachia pileus).  Based on their observations the authors construct a preliminary food web for the genera involved.  Arai & Jacobs 1980 Can J Fish Aquat Sci 37: 120.

Observation of fishing behaviour of the hydromedusa Aglantha digitale in a 2-m tank at Friday Harbor Laboratories, Washington reveals regular upward and downward cycles.  The upward phase involves active swimming with tentacles contracted, while the downward phase involves inverted floating with tentacles extended in fishing behaviour.  Balance is provided by 8 statocysts arranged symmetrically around the margin of the bell.  If these are removed, the animal is unable to perform righting movements during swimming, and typical fishing cycles are absent.  The author reports that this is the first definite evidence regarding statocyst function in a hydromedusa.  Mackie 1980 Can J Fish Aquat Sci 37: 1550.

As indicated in Research Study 1 above, an old idea was that hydromedusae feed passively through random contact of prey with their tentacles. If this were so, feeding efficiency of medusae would be equivalent to combined tentacle length, or the sum of tentacle number times length, and volume of seawater sampled. However, observations of hydromedusan feeding at Friday Harbor Laboratories, San Juan Island, Washington, show that many other factors influence feeding efficiencies.  The researcher employs a 2-m tall plastic tank, which has sufficient volume to permit a greater range of movements than might have been available in earlier studies.  These other factors include: 1) tentacle posture, 2) velocity of tentacles moving throught the water, 3) swimming pattern of medusa, 4) streamlining effects of medusa bell on water flow, 5) diameter of prey, and 6) swimming pattern and velocity of prey.  The hydromedusae studied utilise these factors in different combinations, as shown in the accompanying photos.  In addition to the postural variations described here, the author notes that tentacle length is rarely static, and changes with proximity to the sea surface and degree of food-deprivation. The author remarks that selection of these particular species for study was governed by their availability, and that study of other species will likely reveal even more patterns of spatial exploitation in feeding by hydromedusae.  Mills 1981 Mar Biol 64: 185.

NOTE  species used are Proboscidactyla flavicirrata, Stomotoca atra, Clytia gregaria (Phialidium gregarium), and Polyorchis penicillatus

Laboratory observations at the Pacific Biological Station, British Columbia document the consumption of larval herring Clupea harengus pallasi by several species of hydromedusae.  Plankton tows indicate that the distributions of herring larvae and the hydromedusans Aequoria victoria and Sarsia tubulosa overlap in the field (see graph), thus confirming the potential for predation to occur on the larvae.  The authors also directly observe feeding by Sarsia tubulosa and Aequorea victoria on herring larvae in the field during night-light sessions.  Although medusae have been thought to prey on herring larvae in British Columbian waters, prior to this report no detailed studies have been available.  Arai & Hay 1982 Can J Fish Aquat Sci 39: 1537.

NOTE  these are Bougainvillia multitentaculata, Sarsia tubulosa, Aequorea victoria, Eutonina indicans and, to a lesser extent, Stomatoca atra and Clytia gregaria (Phialidium gregarium). Obelia spp. and Proboscidactyla flavicirrata do not appear to eat larval herring

In Barkley Sound, British Columbia medusae of Polyorchis penicillatus undertake upward vertical migrations of several meters at dusk and sink downwards at dawn.  Use of bottom-mounted “catch-bottle”-type traps shows that this migration occurs concomitantly with the emergence from the sediments into the water column of several major taxa of demersal zooplankton at dusk.  Dissections of the gastrovascular cavity of Polyorchis at various times show that many of these zooplankters are eaten.

Hunting behaviour of Polyorchis mostly involves “sink-fishing”, a behaviour in which the tentacles are held outward at their bases with the more distal portions dropping downward, enabling prey to be captured both from the sides and from below as the medusa sinks slowly down (see drawings).  During fishing the medusae swim slowly (3-15 pulses ^. min^-1), maintaining a more-or-less constant depth, but this is interspersed with more active swimming as shown in the drawings, with the major arc of movement being about 25cm in diameter.  In its sink-fishing posture the prey-encounter distance of the medusa is represented as the volume of a cylinder.  A large individual of 4cm bell height extends its tentacles outward to about 10cm diameter and downward about 15cm.  Thus, all prey are effectively captured in a cylindrical volume greater than 1000cm³.  These diel shifts in water-column position allow Polyorchis access to high densities of demersal plankters both at night near the surface and during daytime at the bottom.  Arkett 1984 Can J Fish Aquat Sci 41: 1837. Photograph courtesy Arkett 1985 Biol Bull 169: 297.

NOTE  living on or near the sea bottom, in this case, including copepods, amhipods, cumaceans, ostracods, polychaetes, rotifers, and several types of invertebrate larvae

In Saanich Inlet, British Columbia the springtime diets of hydromedusae Clytia gregaria (Phialidium gregarium) and C. lomae are dominated by eggs and larvae of euphausids Euphausia pacifica, representing 97-100% of biomass eaten. Minor constituents of the diets include larvae of barnacles, copepods, and decapods, and other incidental prey.  Larson 1987 J Plankt Res 9: 811; see also Larson 1986 Mar Ecol Progr Ser 33: 89 for information on seasonal changes in standing stocks of several hydromedusae species and other gelatinous species in Saanich Inlet, British Columbia.

Related studies by the same author on the hydromedusae Clytia gregaria (Phialidium gregarium) and C. lomae confirm their importance as predators on euphausid eggs and larvae in Saanich Inlet, British Columbia, and provide information on daily ration requirements.  Digestion times range from 2-6h at 13^oC, depending upon the type of prey, and considerably more energy, expressed as % of carbon mass, is taken in over and above that required for daily maintenance, at least for juvenile medusae.  At larger sizes, equivalent on the graph to >1g carbon mass, daily maintenance needs of the medusae appear to be more or less matched by amounts of carbon assimilated.  These data help explain the fast growths of hydromedusae and other gelatinous predators in the plankton, but it is not clear where the energy required for reproduction would come from in the larger, older individuals. Larson 1987 Neth J Sea Res 21: 35.

NOTE  respiration data used for calculation of maintenance requirements are apparently provided in a later paper.  However, as it is difficult to assess energy expenditures of an organism from oxygen-consumption data obtained in respirometry flasks where natural behaviours may be constrained, such imbalances as noted above for reproductive needs are perhaps not unexpected

An interesting approach to the study of dietary specialisation in hydromedusae is to look for correlations between nematocyst types in the predators and the kinds of prey that they eat.  The authors collect a wide array of hydromedusae from areas around Friday Harbor Laboratories, Washington and southern Vancouver Island, British Columbia, and categorise their nematocyst complements in relation to prey eaten. 

Results show that species that eat mostly crustaceans, such as Sarsia tubulosa, have adhesive rhopaloneme nematocysts and other similar ones such as desmonemes that are sticky and adhere to prey surfaces, as well as stenoteles that penetrate the prey exoskeletons (see photographs).  In contrast, hydromedusan species that eat soft-bodied prey tend to lack adhesive-type nematocysts and have 1-2 nematocyst types that penetrate soft-bodied organisms. The authors note that while medusae species feeding on soft-bodied prey may have penetrant-type nematocysts, they seem generally to lack the adhesive-type nematocysts that enable capture of crustaceans. Purcell & Mills 1988 p 463 In, The biology of nematocysts (Hessinger & Lenhoff, eds) Academic Press, NY.

NOTE  data on prey eaten by the medusae come mainly from other publications by the authors, which make for a good literature review.  Common prey include copepods, nauplii, zoeae, cladocerans, veligers, larvaceans, invertebrate eggs, and others

Hydromedusae are often locally abundant and, owing to their good appetites, may have significant impact on fish populations, especially in spawning grounds where eggs and larvae are most susceptible. This is true in Kulleet Bay, Vancouver Island, where during Mar-Jun the eggs and larvae of herrings Clupea herengus pallasi are eaten by hydromedusae Aequorea victoria in significantly greater proportion than their prescence in the environment.  During the peak hatching time of the herring, each A. victoria in the Bay has over 20 larvae in its gut at any time of day or night, representing almost 50% of all ingested prey for that individual.  Other prey items eaten by A. victoria are larvaceans, euphausids, copepods, medusae, and invertebrate larvae.  In comparison, other soft-bodied predators present including ctenophores, chaetognaths, siphonophores, and other hydromedusae, contain few fish larvae.  Purcell 1989 Can J Fish Aquat Sci 46: 1415. Photograph courtesy Chris Gunn, North Island Explorer, Campbell River, British Columbia.

NOTE  embryos added as prey are dyed with Neutral Red vital stain to distinguish them from eggs/embryos produced by females during an experiment

Examination of pelagic predators, including 15 species of hydromedusans, and one species each of ctenophore, chaetognath, and siphonophore, found around 28 herring Clupea harengus pallasi spawning grounds in British Columbia reveals 8 species that prey on herring larvae. Of the hydromedusans, Aequorea Victoria is the most important predator.  Total hydromedusan densities exceed 50 ^. m^-3 at 6 of the 28 sites, with Clytia gregaria, Sarsia tubulosa, and Rathkea octopunctata being the most numerous.  Purcell 1990 Can J Fish Aquat Sci 47: 505.

The hydromedusa Aequorea victoria in areas of Vancouver Island, British Columbia eats a variety of planktonic prey including copepods, larvaceans, other small invertebrate and fish larvae, and various gelatinous zooplankton.  In fact, about 10% of its diet by number of prey items is made up of 10 other species of hydromedusae.  These other species have a relatively high dietary overlap with A. victoria, suggesting a potential for food competition. In fact, the highest predation by A. victoria is on the 2 species with which it has the highest dietary overlap, namely, Clytia gregaria and Rathkea octopunctata.  Aequorea victoria, therefore, reduces potential competition by directly eating its own competitors. One species eating other species that use similar, often limiting, resources, and are thus potentially competing with the first species, is termed intraguild predation.  The species in question thus benefit nutritionally from intraguild prey as well as from reduced competition.  Apparently, intraguild predation of this type is common among both scypho- and hydromedusae, and usually the dominant predator is larger than other members of the guild, as is the case for A. victoria. Purcell 1991 Mar Ecol Progr Ser 72: 255.

NOTE  a total of 801 A. victoria are examined over a 5-yr period for general dietary preferences, and 424 specifically in a single year for gelatinous prey

A comparison of feeding ecology of 2 non-native hydromedusae Maeotias marginata and Moerisia sp. in Suisun Marsh in the San Francisco Estuary shows that both species preferentially eat calanoid copepods (65 and 87% of total gut contents, respectively).  The authors find high dietary overlap between both species of hydromedusae and shad Dorosoma petenense, suggesting the potential for food competition between them.  In contrast, dietary overlap is low between both hydromedusae and bass Morone saxatilis. Other major prey items in M. marginata’s diet include cyclopoid copepods (14%) and barnacle nauplii (14%), and, in Morone’s diet, barnacle nauplii (8%).  Overall, M. marginata seems the more fussy of the 2 species, eating both calanoid copepods and corophiid amphipods in proportions significantly greater than represented in the environment during both years of a 2yr study period.  The authors credit this selectivity in Maeotias to a dense packing of tentacles around the bell (up to 600),  effective exoskeleton-penetrating nematocysts both on the bell and on the manubrium, and a behaviour of flipping over, sinking to the bottom, and twitching its tentacles in an apparent luring behaviour.  This last feature may explain this species’ selectivity for bottom-dwelling corophiid amphipods, but none explains the preference for for open-water crustaceans, and this would be a good future research project. Wintzer et al. 2011 Mar Freshw Res 62: 952. Photograph courtesy the authors and Genomic Variation Lab, University of California, Davis.

NOTE  both species are from the Caspian Sea area

NOTE  other prey for both species include corophiid amphipods,calanoid egg sacs, cumaceans, mysids, zoeae, and fish larvae, all eaten in small quantities