Bow, 1981 INTRODUCTION Information on predation of ascidians is scant and can be found either in studies on the biology of an individual predator (McDonald and Nybakken, 1980; Perron, 1975; Sellers, 1977; Yarnall, 1972) or in anedotal observations of pre¬ dation ontunicates (Ching, 1977; Gulliksen and Skjeveland, 1973; Lambert, 1980). Abbott and Newberry (1980) offers the most current collection of tunicate predator information, Here the sea stars Patiria miniata, Dermasterias imbricata, Pteraster tesselatus, Asterias rubens, Mediaster aequalis, Orthasterias koehleri and Solaster stimpsoni are cited as predators of particular ascidians (not neccessarily in Monterey Bay) along with the nudibranch Hermissenda crassicornis. Millar (1971) lists fish; crabs; the prosobranchs Erato voluta, Lamellariaceangand Cypraeceans; opisthobranchs Pleurobranchus membranaceus and various nudibranchs; turbellarian flatworms and seals as observed predators other than man. My goal was to obtain a survey of possible ascidian predators here in the Monterey Bay area, specifically Pacific Grove. At the same time I hoped to gain general information about predator-ascidian relationships like feeding behavior, tunicate response and damage. Bow, METHODS The survey began with field observations in the inter- tidal of Mussel Point, Pacific Grove, California. Signs of predation such as digested tunicate, missing zooids or actual feeding behavior by predator suspects were noted. Predator suspect is defined as a specimen of a recorded predator, a relative of a recorded predator or any animal associated with and suspected of eating tunicates. Field observations, noting predator species, type of tunicate and source of association (on or eating tunicate). continued throughout the survey. Observations were made at various times and tides in different areas throughout the intertidal and subtidal of Hopkins Marine Station (HMS), Eighteen species of predator suspects were brought into the laboratory for feeding observations. All specimens were collected from the intertidal or subtidal of HMS. All animals except Megathura crenulata were kept in wire mesh cages or plastic tubs with fresh running sea water. The animals were maintained under natural day and night light only. Artificial light was eliminated as a preliminary class study showed that some animals would not feed in bright arti¬ ficial light. Some species are noted to be inhibited by light altogether (Lambert, 1980). A red light created by using a dive light covered with a red plastic sheet provided light to aid in night observations. Specimens of Megathura were kept in an outside tank with running sea water and, thus, were exposed to natural day and night light also. Animals Bow, 1981 were starved for five days and allowed to get accustomed to lab conditions before presenting tunicates, Tunicate specimens used in feeding observations were collected from the intertidal and subtidal at HMS, and from the Monterey Marina floats (Diplosoma macdonaldi, Botryllus spp. Botrylloides spp., and Ascidia ceratodes). Colonies or pieces of colonies, approximately 20 cm2 in surface area, were tied to the bottom of the cages or tied to glass slides and placed on cage bottoms to prevent predators from turning over tunicates and eating them from the potentially injured bottom of the colony. A total of nineteen tunicate species were fed to predator suspects. During an observation run one or more individuals of a suspected predator species would be exposed to equal amounts of six or seven species of ascidians for two to three days. Observation runs were made until all suspects had been exposed twice to all ninteen ascidians. During the first 24 hours, observations were made every one to two hours and every three to six hours thereafter, Every observation was scored for successful feeding, feeding activity without observed success, and association of predator on tunicate species with no observable feeding activity. After the two to three days of feeding, all tunicates were removed from the cages and checked for any damage or effects due to predation such as missing zooids or digested tunic. During the observations, individual feeding behavior and tunicate responses were noted. Bow, 1981 RESULT The laboratory feeding observations are summarized in tables 1-5. The observations of feeding on each tunicate species are summed up for all individual specimens of each predator species. Therefore, the feeding patterns of indi¬ vidual predator specimens are not included. Some predators show a strong association with one prey species. Aplidium californicum was the only ascidian species that Trivia solandri had been observed eating and Eurylepta californica ate only Clavelina huntsmani. Although other tunicate species were eaten, Patiria miniata ate mostly Ascidia ceratodes and Clavelina. Three fissurellid prosobranchs were found to eat a variety of ascidians. Megatebennus bimaculatus was found stronly associated with tunicate prey in both lab and field obser¬ vations. Megathura spent the majority of observations on Archidistoma psammion and ate the greatest diversity of ascidians. Some predator species exhibited behavior difficult to interpretsuch as Hermissenda crassicornis who foraged un¬ successfully on A. californicum for a large portion of the observations, yet spent less than half the number of observations successfully feeding on Clavelina. Calliostoma ligatum was never seen successfully eating A. psammion, but seemed to spend most of its time trying to eat it. Both specimens of Lamellaria diegoensis ate Cystodytes lobatus in the laboratory. Both were found on Cystodytes Bow, 198. colonies of matching color (white specimen on white colony and orange specimen on orange colony). Gut analysis of Calliostoma specimens collected sub¬ tidally from the kelp beds of HMS revealed sand, sponge spicules and detritus, but lacked identifiable tunicate remains. Gut analysis of Megathura, taken subtidally from Lover's Point, Pacific Grove, California, yielded whole orange zooids, a partially digested lobe of Aplidium arenatum and whole Perophora annectens zooids amongst sponge spicules, spirorbis, algal fragments, bryozoans, diatoms, detritus and sand. Feeding Observations: With the exception of a polyclad flatworm, Eurylepta californica, and the bat star Patiria miniata, the other predators found were eight species of gastropods, three of which are keyhole limpets. These three are Diodora aspera, Megatebennus bimaculatus, and Megathura crenulata and they eat both the zooids and the tunic. Diodora uses the outer lips of the mouth to form a suction that pulls the prey up to open jaws. The radula makes one or two strokes, tearing off tissue. Then the jaws close, cutting off any threads hanging from the radula. Megathura and Megatebennus probably eat in a somewhat similar fashion. The mantle of these two limpets covers the head region of the animal, making it difficult to follow Bow, 1981 the exact mechanism of feeding. Gouging bite marks left in tunicates fed to Megathura and Megatebennus suggests that the mouth is applied to the tunicate surface, jaws agape, while the radula rasps up masses of tunicate tissue, While the exact mechanism used by Calliostoma ligatum was not directly observed, it seems feasible that the mechanism would be similar to the feeding behavior of Calliostoma annulatum observed by Perron (1975). C. ligatum crawls across the tunicate surface, tasting the surface with its mouth and tentacles. It raises its head with the mouth open and in a sudden jabbing motion, lunges at its prey, Its mouth attaches to the prey and the head rocks up and down as the tunicate zooid is drawn into the mouth. The feeding mechanism of Hermissenda crassicornis is well described by Yarnall (1972). The rocking action of the buccal sac as it pumps, sucking up the zooids, is accompanied by a jerking movement of the ceratae from a relaxed, ex¬ tended position to a position lieing flat against the body, The feeding method of Lamellaria diegoensis is described by Fretter and Graham (1962) and Lambert (1980). The mantle covering the shell of Lamellaria spreads out over the periphery of the animal, concealing the actions of the mouth. In order to determine feeding, the animal had to be removed and the tunicate inspected for signs of predation (hole in tunic). The feeding behavior of prosobranchs Trivia solandri and Erato vitellina, and the bat star Patiria miniata is Bow, 1981 described in Abbott and Haderlie (1980). Eurvlepta californica is often found wrapped over the siphons of Clavelina. When it eats, the mouth and pharynx extend down through one of the siphons and sucks in the zooid. As the flatworm's mouth and pharynx extend deeper into the tunic, the central part of the body folds into the tunic through the siphon. The pharynx and gut of the zooid are ingested by the flatworm. Bow, DISC SSION To try and get the closest correlation of tunicate predators with their preferred prey species, a point system was employed which included not only successful feeding observations, but also the attempts at feeding where success could not be determined, along with observations of association of the predator with tunicate species. By giving each type of observation a weighted numerical value based on the assumption that a predator eats more of, or associates more with, its preferred food, each tunicate species receives a number that represents the predator's preference for that species. (figure 1) All tunicates that were ranked most preferred had been successfully preyed upon by their prospective predators. whereas only half of the second and third most preferred tunicates were successfully eaten. The result that all of the most preferred tunicates were successfully eaten may be taken as a positive indication of the appropriateness of the point system in ranking preferences. The quick drop in positive correlation for the second and third choices may be explained by: 1) The occurance of specialized preda¬ tors that eat only one or two tunicate species. 2) Sometimes a predator specimen spent a significant amount of observation time sitting on a tunicate, using it as a substrate and not trying to eat it. 3) In some cases active feeding behavior was repeatedly observed on the same prey species without Bow, 1981 observed success. This may be due to a zooid-escape response or the predator may be grazing material on the tunic surface, Predators like Hermissenda and Calliostoma who eat only the zooids may be unsuccessful in feeding attempts because the zooids contract down into the protection of the tunic when disturbed by predator foraging. This same con¬ traction-escape response can be elicited by poking or rubbing the tunic surface with a dissecting probe, Predators such as Megathura, Megatebennus and Patiria that eat both the zooid and tunic were observed either eating or not eating which resulted in scored feeding rates of 100% success. (tables 2 and 5). Even though the zooids are withdrawn, the tunic-plus-zooid-eaters are usually success¬ ful as they eat right through the tunic until the zooids are available for feeding. Sand encrusted tunicates such as Synoicum parfustis, Euherdmania claviformis and sandy lobes, excape successful predation by zooid-only-eaters. The sand provides an extra barrier of protection for the zooids. These tunicates were successfully eaten only by Megathura and Patiria, both voracious zooid-plus-tunic-eaters. Overall, A. californicum is the most preferred tunicate, Clavelina ranks second and A. psammion is one of the third most preferred species. (figure 2) This preference may correlate directly with the intertidal abundance of species as these top three choices simultaneously rank the same in abundance intertidally (Pulliam, 1981). The predators may Bow, 1981 11 have already been conditioned to eating one of the top three species in the field, so when brought into lab, they continued to feed on the same favored species. Many of the Megatebennus and Diodora specimens were collected either on or near A. californicum, and did, in fact, favor A. californicum during lab feedings. There appears to be no definite coorelation between body form (solitary, social or compound) and the degree of predation. (tables 1-5) and figures 1 and 2) Both A. californicum and Clavelina, the top two choices, are non-ascidic and contain low concentrations of vanadium, whereas Euherdmania and Distaplia occidentalis, the two least predated species, contain high concentrations of vanadium (Abbott and Newberry, 1980). However, A. psammion also contains high vanadium concentrations and has a pH of 1-2. yet ranks number three. Though there may exist chemical defenses against predation, they certainly do not offer a complete means of protection. Feeding studies of individual predators would be use¬ ful in determining the importance of tunicates in predator diets. The rarer species Trivia, Lamellaria and Erato tend to be specialists, eating only one or two tunicate species (figure 3). These would be interesting to study, but their rarity may make it difficult to accumulate data. The two nost common predators, Hermissenda and Patiria, ate a larger variety of ascidian species (seven and eight, respectively). These certainly would lend themselves to Bow, 1981 12 more investigation especially in the field. In lab Megathura was found to eat the greatest number of tunicates, a total of 14 species. Despite the lack of feeding observations, they do eat ascidians in the field as the gut analysis shows. These large animals are known to occur in abundance and may prove to be a profitable predator for further study. Megatebennus was primarily found in areas of high surge where water was well ciculated or was brought in to lab with subtidal tunicate specimens. It was found mostly on A. californicum and in addition, often resembled the color of the tunicate with which it was associated. All the ascidians which it ate in lab occur in intertidal areas of high surge. Many of the animals in sea water tables became sick unless a hose was positioned with its excurrent flowing directly through the cage. This suggests that Megatebennus prefers a habitat with a high turnover of water. Perhaps this in¬ formation will be useful in locating specimens in later studies. It would be best to obtain data on tunicate predators from actual field observations; however, it was difficult to observe many instances of feeding in the field. (figure 4) In collecting data in the laboratory, 184 hours were spent observing suspected predators and potential prey. This yielded 147 observations of actual feeding. In contrast, 80 hours were spent looking for predation in the field and only 15 feeding observations were made. Perhaps predators have their most significant impact during larval stages or on newly settled and young adult stages. may be wort ffects yle we inve considered in this study 2) 4) Bow, 1981 14 SUMMAR Based on lab and field observations, the following ten 1) animals were observed feeding on ascidians: fissurellid gastropods; Diodora aspera, Megatebennus bimaculatus, and Megathura crenulata; other prosobranchs; Calliostoma ligatum, Erato vitellina, Lamellaria diegoensis, and Trivia solandri; the opisthobranch Hermissenda crassicornis; the bat star Patiria miniata; and the polyclad flatworm Eurylepta californica. A point system was applied to feeding observations to obtain a correlation between tunicate species and predator preferences. The three most abundant species of tunicates in the field, Aplidium californicum, Clavelina huntsmani, and Archidistoma psammion, were found to be the three most predated tunicates in lab. The less common predator species Erato, Lamellaria and Trivia appear to specialize on one or two tunicate species, Hermissenda, Patiria and Megathura are more general in their choices, eating a variety of seven or more ascidian species. The fissurellid gastropods, Lamellaria and the bat star eat both tunic and zooids. The other prosobranch gastro- pods, opisthobranch and polyclad flatworm eat only the zooids. C Bow, 1981 Zooid contraction and a sand embedded tunic were observed 6) defenses against those predators that eat only the zooids. Bow, 1981 ACKNOWLEDGEMENTS I want to thank my fellow students in Bio 175H, Spring 1981 for their moral support and aid in collecting predator specimens. I am grateful to Jim Watanabe for providing the tank space used in feeding observations and to Dr. Jon Roughgarden for the desk space. I also thank Dr. Donald Abbott for his inspiring guidance and I give special thanks to Chuck Baxter for'time and thoughtful suggestions provided during throughout the project and'the writing of this paper, Bow, 1981 LITERATURE CITED Abbott, D.P. and A.T. Newberry. 1980. Urochordata: The Tunicates, pp. 177-226, in Intertidal Invertebrates of California, Stanford University Press, Stanford, California. 690 pp.; 200p1. Ching, Hilda Lei. 1977. Redescription of Eurylepta leoparda Freeman, 1933 (Turbellaria:Polycladida), a predator of the ascidian Corella willmeriana Herdman, 1898. Canad. J. Zool. 55: 338-42. Fretter, Vera and Alastair Graham. 1962. British prosobranch molluscs. Ray Soc., London. 755 pp. Gulliksen, B. and S. H. Skjaeveland. 1973. The sea star, Asterias rubens L., as predator on the ascidian, Ciona intestinalis (L), in Borgenfjorden, North-Trondelag, Norway. Sarsia. 52: 15-20. Lambert, Gretchen. 1980. Predation by the prosobranch mollusk Lamellaria diegoensis on Cystodytes lobatus, a colonial ascidian. Veliger 22; No. 4: 340-344. McDonald, Gary R. and James W. Nybakken. 1980. Guide to the Nudibranchs of California. American Malacologists, Inc., Melbourne, Florida. 72 pp. Millar, R. H. 1971. The biology of ascidians. Adv. Mar. Biol. 9: 1-100. 17 NEIL ALLEN t:.O. min t: 5 min t: 1O min t: 1 5 min t: 20 min FIG1 ASCIDIA CERATODES LARVAE S 55 n D PAO Seman Mn — CO SP W S Imm 18 Bow, 1981 Tables 1-5 Summary of data from lab observations. All observations were scored in the following manner: + -- Predator was observed feeding successfully on tunicate species. Success was determined by actually seeing the zooid or tunicate material pass into the predator or by strong evidence such as missing zooids or digested tunic. o/+ -- Observed feeding behavior, but found no evidence for success. o -- Predator observed sitting or crawling on tunicate, but no feeding activity was observed. blank space -- Predator was not observed on the tunicate species. The number of scored observations for each tunicate species was summed for each predator species. The success rate of feeding was determined by number of observations % success number of + observations plus -number of o/+ observations Hermissenda crassicornis was the only predator to have a success rate other than 100% or 0%. Therefore, the success rate column (%) is shown for this predator only in table 1. 19 Bow, 1981 PR -- preference ranking For each + observation 3 points were given; every o/+ observation received 2 points; observations scored with o were given 1 point; and blank spaces received no points. type of The number of each'observation was multiplied by the appropriate number of points and all points were added for each tunicate. This gave a number, which when com¬ pared to the number given to the other tunicate species, particula indicated the predator's preference for the 'species. The tunicate species were then ranked accordingly; the species with the highest number of points re¬ ceived the highest preference ranking of 1. The next highest number was ranked 2 and so on. Tunicate species are listed according to body form; a -- solitary b -- social c -- non-acidic colonial d -- acidic colonial 19a + +++ + +++++ + +++++ ++ ++ ++ + ++ ++ + ++ Bow, 1981 PRÉDATOR: TUNICATE Ascidia ceratodes Clavelina huntsmani Euherdmania claviformis Perophora annectens Aplidium californicum Aplidium solidum Archidistoma diaphanes Archidistoma itter Botryllus spp. Botrylloides spp. Polyclinum planum Synoicum parfustis Sandy lobes Trididemnum opacum Archidistoma molle Archidistoma psammion Cystodytes lobatus Diplosoma macdonaldi Distaplia occidentalis ttte + +E +++ Table Hermissenda crassicornis 0/4 46 12 100 10 18 11 50 100 13 13 11 ++++ t ++++ ++ + Diodora asper 0/+ +++ +++ t H + ++ +++ Bow, 1981 Table 2 —. —— Megatebennus PREDATOR: ulatus TUNICA O/ PR — Ascidia ceratodes Clavelina 11 huntsmani — —— Euherdmania 10 VifOERIS Perophora 12 pne k s - ——.— Aplidium lifornicum — —- Aplidium solidum Archidistoma anes — toma hid 32 tte Botryllus DE. — lioides Bot spp. — clinum planum Synoicum parfusti I Sandy lobes — Trididemnum 24 oacum 1 Archidistoma molle Archidistoma psammion Cystodytes 13 lobatus Diplosoma macdonaldi — Distaplig iden 21 Megathura anulata — PR O/+ — — — — — 2 5 3 15 — 7 Bow, 1981 22 Table 3 — EDATOR: Erato Calliostoma ligatum Vitellina TUNICAT o/ PR PR O/+ g — ASC ceratodes. k ka Ciavelina huntsmani — 5 2 — Euberdmania Glavifomis Perophor nnet ens Aplidium Ca ifornicum L —— Aplidiun solidun — Archidistoma diaphanes idistoma Bo Vllus spp. —. Botrylloides spp. 1 Polyclinum planum — — P Synoicum parfustis Sandy lobes — Trididemnum O cum — Archidistoma molle — Archidis oma psammion Cystodyte 5 lobatus 3 Diplosoma Macdonaldi staplia occide Bow, 1981 23 Table 4 —.—— Lamellaria Trivia PREDATOR: liegoensis solandri — TUNICAT PR PR O/+ /+ — Ascidia ceratodes ——— Ciavelina buntsmani — Euherdmania Claviformis Perophora ens — Aplidium californicum — — Aplidium solidun Archidistoma diaphanes. 5 hidistoma ritteri Botryllus spp. s Botrylloides — Polyclinum planum — Synoicum parfustis Sandy lobes Trididemnum pagun T Archidistoma molle — Archidistoma psammion Cystodytes lobatus Diplosoma macdonaldi Distaplia occydentalis Bow, 1981 Table 5 Eurylepta PREDATOR: californica TUNICAT O/+ PR scidia ceratodes — Clavelina nitsmani2 Euherdmania Claviformis — Perophora ectens — Aplioium californicum Aplidium solidum Archidistoma diaphanes Archidistoma ritter Botryllus . Botrylloides sp Polyclinum planum parfusis ynoicun Sandy lobes Trididemnum opacum Archidistoma molle Archidistoma psammion Cystodytes lobatus — Diplosoma macdonaldi Distaplia ogcidentalis 24 — Patiri miniata PR 0/+ 12 2 — + —— 1 5 — ——— — 2 8 — Bow, 198. Figure 1 Preferences and Predation based on lab observations. The numbers indicate each predator's preference for a tunicate species; number 1 being the most preferred and larger numbers for the less preferred species. The same scoring symbols used in tables 1-5 are shown to indicate each predators observed reaction to each tunicate species. ++++ + Bow, 1981 ++ 1 + ++ FIGURE 1 ++ PREDATOR! + TUNICATES: +++ + Ascidia ceratodes Clavelina huntsmani + Euherdmania claviformis Perophora + annectens Aplidium californicum Aplidium solidum ++ Archidistoma ++++ diaphanes + ++ + Archidistoma ritteri + ++++ Botryllus spp. Botrylloides spp. 0/+ Polyclinum planum Synoicum parfustis +++ Sandy lobes Trididemnum opacum 0/+ Archidistoma molle Archidistoma psammion Cystodytes lobatus Diplosoma macdonaldi +— Distaplia occidentalis ++ +++++ ++ + ++++ + + + 11 10 12 13 + 0 0 3 0/4 0/+ O/ 0/+ ++ + + ++— + 00 — 00 1 O/+ 0/+ 11 11 o/4 10 O/+ 0/4 0/4 0-- 9. 0/+ 0/4 + Du O— 0/4 + +++ +++ 41 + + O Bow, 981 Figure 2 Number of Predators per Tunicate Species. This indicates the number of predators found success¬ fully feeding on each tunicate species. 27 Bow, 1981 +++ — TUNICATES ++ LEAST PREDATED Euherdmania claviformis +++ Distaplia occidentalis + Trididemnum opacum Archidistoma solidum Archidistoma molle Archidistoma diaphanes Synoicum parfustis Sandy lobes Polyclinum planum Perophora annectens Botryllus spp. Ascidia ceratodes +4 +++ Botrylloides spp. + ++ + Archidistoma ritteri — ++ Diplosoma macdonaldi ++ +++ + Cystodytes lobatus ++ Archidistoma psammion +++ ++ Clavelina huntsmani Aplidium californicum 01 MOST PREDATED ++ ++ —++ +1 + + +++++ ++++++ + ++ ++ ++ + ++ + ++ + t + +++ I F++ + ++ ++++ +++ +++ + + FIGURE 14 + + E 14 L + 3 4 | 5 6 NUMBER OF PREDATOR + ++ ++ ++ +++ + ++ 10 + ++ Bow, igure 3 Jumber of Tunicate Species eaten pe edat This shows the total numbe of tunicate s each predator was observed eating successfully + + ++ + +++ 1114 +++++ + — — 1 ++ + ++ ++ + +++ + + —— o LO ++ - 1 ++ a +4 Hi + + + + + ++ + ++++ t + +++++ Bow, 1981 30 FIGURE ++ ++ + t ++ 4. Bow, 198 Figure 4 Predation based on Field Observations. The same scoring symbols used in tables 1-5 were employed to score observations made in the field. Double symbols, o, +, +, indicate observations that were seen in the field but not in the lab. + -- observed successful feeding qy+ -- observed feeding behavior but success undetermined undetermined o -- predator observed on tunicate but no feeding activity exhibited Blank spaces mean that no interaction was observed between the predator and tunicate species. Bow, 1981 FIGURE 4 PREDATORS TUNICATES: Ascidia ceratodes Clavelina huntsmani Aplidium californicum Aplidium solidum Archidistoma diaphanes Archidistoma ritteri Synoicum parfustis Sandy lobes Trididemnum opacum Archidistoma molle Archidistoma psammion Cystodytes lobatus Distaplia occidentalis 52 8 6 - 0 + D —- 5 — — O/+ 0 0 + 00 O/+ C --— -+- 0 H 0-— — — 0