Rice and Cotsirilos -la Abstract Fecal pellets of seven species of tunicates (Ascidiacea) were collected from animals inhabiting the same intertidal habitat near Pacific Grove, California in order to assess the kinds material ingested. Two additional species were studied but of their fecal pellets were not examined. Dry weight per pellet was established for all species and 5.0 x 10 5g of fecal matter from each were compared microscopically. Of a variety of ma¬ terials tabulated, diatoms and dinoflagellates were the most abundant. Species that had very high numbers of diatom frustules also had a large number of digested diatoms. Tunicates pro¬ ducing pellets with large numbers of dinoflagellates were of a different species and morphology than those showing diatoms to dominate. Five species of tunicates were selected for laboratory feed¬ ing studies and involved solitary, social, and colonial animals. Graded sizes of two Macrocystis species ( brown algae) and sand grains were offered. Larvae of the barnacle Pollicipes polymerus (Sowerby,1833) were fed to the three largest animals (Styela mon- terevensis Dall, 1872; Ascidia ceratodes Huntsman, 1912; and Perophora annectens Ritter, 1893) and Ascidia tadpoles were introduced to the two solitary species. The acceptance and re¬ jection of these food items was observed in connection with mouth size and tentacular screens. Items with diameters too large to pass the screen without striking a tentacle were found in fecal pellets while smaller ones were occasionally rejected. Our observations indicate that the tentacles show important activity when encountered with potential food particles. -2 Rice and Cotsirilos Introduction In their surveys of size and shape of marine invertebrate fecal pellets, Edge (1934) and Arakawa (1971) demonstrated that pellets from the few species of tunicates they examined differ¬ ed in morphology. Whether these differences were specific to the species, as shown by Arakawa (1971) for the three species of Styela, or different because of food choice was not consider¬ ed by these authors. While mechanisms associated with feeding such as pumping rates, pumping rhythm, filtration rates and filtration efficiency have been studied in ascidians (Fiala-Médioni 1978 a ,b, and e), the fate of particles taken in during these active pro¬ cesses has not been looked into. That many of these particles serve as food is a reasonable surmise, but some such particles may be ingested by accident. A study of size, type and number of particles ingested in field populations of tunicates was undertaken. Laboratory feed¬ ing studies were subsequently performed to determine both poten¬ tial food kind and size and also how a tunicate might actively exclude particular items from entering its digestive tract. -3 Rice and Cotsirilos Methods and Materials The following ascidians: Styela monterevensis (Dall, 1872), Ascidia ceratodes (Huntsman, 1912), Clavelina huntsmani (Van Name, 1931), Perophora annectens (Ritter, 1893), Synoicum parfustis (Ritter & Forsyth, 1917), Polyclinum planum (Ritter & Forsyth, 1917), Archidistoma molle (Ritter, 1900), Archidistoma psammion (Ritter & Forsyth, 1917) and Aplidium solidum (Ritter & Forsyth, 1917) are available in the intertidal off Mussel Point at Hopkins Marine Station, and were chosen for this study. A channel subject to heavy surf action was selected in order to obtain Styela which appeared to prefer the well-exposed, low intertidal zones. Species were collect¬ ed within a 10 m radius. Barnacles, clams, and other encrusting or¬ ganisms were trimmed off the colonial species to prevent mixing of waste materials. Tunic and substratum surfaces were washed free of debris. Specimens were placed in cut polyvinylchloride tubing (15 cm diameter) with Nitex 0.15 mm weave floors which allowed for water exchange but prevented fecal pellet loss. One hundred pellets were collected from each species, except from Styela which produces con¬ siderably larger pellets. Eighteen millimeters of Styela fecal matter was obtained. Pellet weights were obtained in order to con¬ duct a cross species comparison of food content between identical masses of fecal matter. All samples were rinsed twice in fresh distilled water to remove salt residue. Twelve glass vials were baked for 1 h at 70° C. and then vacuum desiccated for 48 h. Crystalline phosphoric acid was the desiccating agent. Dry vial weights were obtained with a Mettler balance accurate to 1.0x 10 ' g. Pellet samples were -4- Rice and Cotsirilos gathered in 2 ml of fresh deionized water and placed in separate vials, leaving five controls. The same baking and drying pro¬ cedures were repeated and dry weight per pellet was established for each species. An average weight reduction of 1.1 x 10 (R = 1.0 x 10 g to 1.2 x 10"4 g) was observed for the five control vials and was added to dry pellet weights in order to correct for them. Water weight as a source of error was elimin¬ ated by obtaining dry weight data and vials were handled with forceps to avoid fingerprint weight. In late May these species were collected within a 10 m radius. In the laboratory, tunicates were rid of extraneous external material, and 1.0 x 10" g of fecal matter was collect¬ two glass slides ed from each species. Pellets were evenly spread between,and a grid was etched to determine area. Half of the total fecal area was examined at 450 X from different sections of the grid in order to observe random parts of the fecal matter. Laboratory feeding studies were performed on Styela, Ascidia, Perophora, Synoicum and A. psammion. Digestive times for each species were obtained in order to predict elimination time of a given ingested particle. Collected animals were placed in the Nitex containers for a minimum of 1 h. A carmine¬ seawater suspension was gradually introduced after verifying that zooids were relaxed and expanded. Due to its larger size, Styela was placed on a non-toxic clay base in a beaker receiving running water. After 10 min of active carmine ingestion, sub¬ sequently voided pellets were rinsed in carmine-free water and -5 Rice and Cotsirilos examined microscopically for remnant carmine. The first appear¬ ance of red particles enveloped within the mucous rope was scored as a laboratory digestive time (table 1). Food material included Macrocystis spp., Pollicipes larvae, Ascidia tadpoles and sand grains. Detrital Macrocystis spp. blades were collected and allowed to decompose for 3 days in stagnant water before grinding in a blender. Algal pieces of 2.20-1.35 mm, 1.35-1.00 mm, 1.00-0.56 mm, 0.56-0.25 mm, 0.25- 0.15 mm, 0.15-0.10 mm, and less than 0.10 mm were obtained by straining through Nitex and geological sieves. Animals were initially fed the smallest algal sizes mixed with carmine to habituate them to the food item. Running water assured continual plant fragment circulation. At 5 min intervals small amounts of the next largest size category were added. Pellets were chron¬ ologically examined and size of largest algal fragments were noted (see figure 3). Pollicipes larvae studies were carried out in still water under the dissecting microscope to allow for behavioral response observations. Species were simultaneously introduced to carmine, larvae and algal fragments of size 0.25-0.15 mm after habituation to the maximum algal fragment proven to be acceptable in the feeding studies. Larvae were present during active feeding on carmine and algae. Fecal pellets were examined for larval remains. Sand grains and Ascidia larvae were fed to Styela and Ascidia. The inert particles proved too heavy for Styela to suck into its curved siphon so the animal was inverted. Styela was -6 Rice and Cotsirilos fed algae of known maximum size acceptance to demonstrate its ability to feed while inverted. Sand grains were then intro¬ duced and responses were noted. Mouth sizes and tentacular screens were measured with an ocular micrometer and sketched to scale (figure 4). Mouth dimen¬ sions and tentacular shape, organization and orientation were measured for Perophora, Synoicum and A. psammion during carmine intake and diagrams are replicas of a feeding zooid. Styela and Ascidia tentacular networks were taken from preserved and anasthe- tized animals respectively and measured from the inside. Tentac¬ ular clearance, a measure of the largest sphere and rectangle that could slip through the tentacles without contact, was estab¬ lished for each species (figure 4). Rice and Cotsirilos -7- Results—Fecal pellet contents Figure 1 shows in descending order, the abundance of dia¬ toms within a 50 ug fecal sample from each of seven species of tunicates. Mean weight per pellet was calculated for each species and is shown in table 1. A large number of empty diatom frustules occurred in identical fecal weights of all the species. A portion of these could have been present in the water column and a portion entering as food. Dinoflagellates, the second most numerous item found in the pellets were within the size ranges of the diatoms, but of different shapes. Table 2 divides the seven species into two groups accord¬ ing to individual or colony shape. Polyclinum and Synoicum are lobed colonies attached to the substratum by stalks. These two were placed in Group L along with Clavelina and Styela which are also lobed with their oral siphons elevated above the substratum. Aplidium solidum, A. psammion and A. molle comprise Group M, which are encrusting mounds growing in close proximity to the substratum. Fecal pellets from species in Group L con¬ tain more dinoflagellates per ug than species in Group M. -8- Rice and Cotsirilos Discussion—Fecal pellet contents It is assumed that all species were exposed to uniform environmental conditions since they were collected from the same site within a 10 m radius. Some of the species, however, had several times as many diatom remains in their fecal pellets than did others. Species which take in the most diatoms leave the highest proportion of diatoms digested. (Digested diatoms are defined as those frustules partially or completely emptied of contents.) This correlation indicates that they are efficiently digesting the food source which is most often seen in their fecal pellets. Fecal pellet architecture was compared between species in light of the data presented in figure 1. No correlation was observed between fecal pellet contents and pellet structure. Any settling of diatoms upon or close to the substratum could potentially create higher densities there and facilitate ingestion by encrusting species such as A. molle and Ap. solidum. If diatoms come to rest upon the tunics of these colonies then they are in an excellent position to be taken into the oral siphons. The greater number of dinoflagellates in pellets from Group L indicate that species with oral siphons located some distance above the substratum are more likely to ingest dino¬ flagellates. Whether this is due to a lower abundance of dino¬ flagellates close to the substratum or to some selective action of the intake mechanism is not clear. Since most of the dino¬ flagellates counted ranged in size from 15-20 um, and tentacular C Rice and Cotsirilos clearances ranged from 80-230 um, it is hard to visualize : mechanism by which tunicates could actively select dinoflagellat Rice and Cotsirilos -10- Results--Feeding The two Macrocystis species were ingested by all five tun¬ icates. Animals with larger mouth sizes ingested greater sized algal fragments (figure 3). Large plant material tended to be rejected unless the individual was introduced and habituated to smaller food items first. When algal clumps entered oral siphons, a violent contraction of the pharyngeal cavity and subsequent expulsion of the plant matter occurred. Algal fragments were larger or comparable to the largest inorganic rocks found during pellet examinations. Dimensions of the largest sand grains found in waste prod¬ ucts of each species exceeded the tentacular clearance dimensions. Largest dimensions of algal patches also exceeded tentacular clearance restrictions in all species except Synoicum. When individually introduced, Pollicipes larvae were ingested and digested by the largest sized animals—Styela, Ascidia, and Perophora. Larvae fed in large quantities, however, were actively rejected by all three tunicates. Larvae dimensions ex¬ ceed or match tentacular clearance estimates but are smaller than all mouth sizes (figure 4). Ascidia larvae averaging 0.85 mmx 0.12 mm were introduced to and digested by Styela and Ascidia. Sand grains of 0.6 mm cross-section were fed to Styela and Ascidia. It was impossible to experimentally introduce smaller grains. Twenty-five grains were individually placed into their oral siphons at one minute intervals and all were rejected. -11 Rice and Cotsirilos Discussion—Feeding Data suggest that algal pieces are the largest particles acceptable to the five tunicates. However, tentacular clear¬ ance measurements indicate that many particles found in fecal pellets must have contacted the screen without eliciting pharyngeal contractions. Among such particles were sand grains larger than the largest algal fragment ingested by Ascidia. On the other hand, animals fed sand grains of 0.6 mm cross-section immediately after ingestion of larger sized algae consistently rejected these grains. The parameter most likely to prevent in¬ gestion of a grain particleappears to be its mass rather than its size. This conclusion contradicts a study made by Hecht on Ascidia atra ( - Ascidia nigra) who claimed "The selection of its food- if mere exclusion may be called selection-- is made on the basis of size, and rejection depends on the mechanical stimulation by the larger particles" (Hecht, 1918). Sand grains dropped through the water column into the oral apertures were immediately expelled; particles placed into the siphon would occasionally sit within the tentacular mesh for several minutes before rejection. Gravity's effect on a sand grain is considerably more pronounced than on an algal fragment of comparable size. The disparity in size acceptance is clearly depicted for Styela in figure 3, which accepted plant fragments larger but less massive than the largest rocky material. Algal fragments descended slowly and often were accelerated by the animal's own intake suction. In the sand particle study which produced 100% rejection, a grain's velocity greatly exceeded the oral intake velocity. Impact force of grains -12- Rice and Cotsirilos against the tentacular screen, especially when dropped through the water column, exceeded that of algal fragments and provoked rejection. Observations on oral tentacular sensitivity were also made in Hecht's study of A. nigra. He performed similar sand grain experiments. Pharyngeal contractions following ingested sand grains led him to discount earlier hypotheses by Roule and Lecaze Duthiers et Delage who both suggested non-innervation of tentacular screens. Our study supports Hecht's theory that oral tentacles are innervated and capable of excluding certain items from entering the digestive tract. Pollicipes larvae were used to determine if live animals were acceptable to tunicates. Their mass was less than the rejected rocks and they were not limited to free-falling motion due to their own mobility. Observations on Synoicum indicated that larvae could actually swim against an incurrent of carmine and were often seen swimming in the siphon opening but not actually striking the tentacles. An errant leg of the larva against the tentacles would elicit a brief contraction, ejecting the barnacle larva from the oral siphon. Entanglement within the tentacular screen and ensuing frenzy of beating appendages by the larva initiated violent contractions by the tunicates, preventing ingestion of barnacle larvae. Styela and Ascidia accepted larvae readily. It is plausible that these largest ascidians, due to their powerful water intake, frequently ingest larvae parts (or particles of equal mass) and are therefore insensitive to them. Larvae may also be moving so rapidly -13 Rice and Cotsirilos that they span the tentacle mesh with little chance of eliciting a rejection response. Barnacles force-fed in large numbers to both species were readily rejected; presumably their additive mass and appendage movements are enough to stimulate an expul¬ sive contraction by the larger ascidians. Perophora repeatedly repulsed larvae but fecal analysis indicated that they were indeed ingested and digested by this tunicate. Careful procedures in larvae selection ensured that only live animals were fed, thus it can be assumed that larvae cleared the tentacular screens, regardless of appendage movements. Perophora only allowed the smallest of larvae to pass into its gut, suggesting that they stimulated the tentacles the least. -14- Rice and Cotsirilos Summary-fecal pellet contents 1. Diatoms and dinoflagellates account for a large percentage of a tunicate's diet. 2. An increase in diatom content in identical fecal weights correlates well with increasing numbers of digested diatoms. 3. No correlation is seen between the gross structure of fecal pellets and their contents. 4. In our sample, species morphology appeared related to dinoflagellate abundance in their fecal pellets. Perhaps the taller individuals (Styela) or stalked colonials are feeding upon a water column containing more dinoflagellates than are those species growing closer to the substratum. -15 Rice and Cotsirilos Summary—Feeding 1. All five species involved in the feeding study consistently accepted algal pieces larger than tentacular clearance dim¬ ensions, indicating that while tentacular contact occurred, ingestion was still allowed. Consistent rejection of sand grains smaller than readily 2. accepted algal patches suggests that a screening mechanism is indeed available to tunicates but does not operate strictly on size of the particle encountered. This rejects Hecht's (1918) appraisal that ascidians select their diet on the basis of size alone which would predict that larger particles are preferentially excluded. The force of contact applied by an ingested particle on the tentacle network is closely correlated to its probability of being rejected. Dropped sand grains, accelerated by gravity, are more likely to be ejected than a grain of com¬ parable size gently rolled into the oral siphon. Hecht (1918) proposed that detection of ingested matter occurred at the tentacle screen, suggesting innervation of these tentacles. We support this conjecture. 4. -16 Rice and Cotsirilos Suggestions for further study 1. An interesting follow-up to this project would be to elucidate the maximum sized particle a species might be able to digest. Perhaps tentacular irritation can be bypassed by anastheti¬ zation, in which case a large organic particle could be pass¬ ed through the tentacles into the branchial basket. Ob¬ servations on the fate of this artificially fed particle could be made after the animal recovers. 2. Mechanism used to construct the species specific fecal architecture and location of these mechanisms within the digestive tract could be studied by examining the consis¬ tency and shape of fecal material at different sites of the tract. c Rice and Cotsirilos Acknowledgement We thank Dr. Isabella A. Abbott and Dr. Donald P. Abbott for their encouraging suggestions and willingness to spend long hours with us. -17 -18 Rice and Cotsirilos Literature cited Arakawa, Kohman Yi 1971. Studies on the faecal pellets of marine invertebrates (excluding molluscs) I. Publ. Seto Mar. Bio. Lab., 19: 231, 239-241. Edge, Elton R. 1934. Faecal pellets of some marine invertebrates, Am. Midl. Nat.. 15 (1): 82-84. Fiala Medioni, Aline 1978a. Filter feeding ethology of benthic invertebrates (ascidians).III. Recording of water current in situ—rate and rhythm of pumping, Mar. Biol. 45: 185-190. Fiala Medioni, Aline 1978b. Filter-feeding ethology of benthic invertebrates (ascidians).IV. Pumping rate, filtration rate, filtration efficiency. Mar. Biol. 48: 243-249. Fiala Medioni, Aline 1978c. Filter-feeding ethology of benthic invertebrates (ascidians).V. Influence of temperature on pumping, filtration and digestion rates and rhythms in Phallusia mamillata, Mar. Biol., 48: 251-259. Hecht,Selig 1918. The physiology of Ascidia atra Lesueur.I. General physiology.II. Sensory physiology. J. Exp. Zool. 25: (1) 266-270. 0 Rice and Cotsirilos Table 1 Although captions for fecal matter are given in terms of micrograms per pellet, Styela data are represented as micrograms per length. Amount of fecal matter is represented as number of individual fecal pellets. -10 Rice and Cotsirilos O C3 C - CO — L ooog PE — - N B ONOINO — — - JNN O ou OR ONDOON 0 3 0 ooaaa L —. -20- 0 Rice and Cotsirilos Table 2 Group L contains lobe shaped species and Group M contains mounded colonies. -21- Rice and Cotsirilos -22- Table 2 Number of dinoflagellates in the Group L 50 microgram sample of fecal pellets Polyclinum planum 148 Clavelina huntsmani 118 Synoicum parfustis 115 Styela montereyensis 114 Group M Aplidium solidum 88 45 Archidistoma psammion 37 Archidistoma molle — e Rice and Cotsirilos -23 The column of numbers on the left represents the Figure 1 percentage of all diatoms found in the fecal pellets of each species which were either partially digested or completely empty frustules. A number has been placed at the tip of the diatom bars to facilitate exact comparisons. Rice and Cotsirilos -24- Percent of all diatoms Figure which are digested k Clavelina huntsmani 388 2 H 90 30 H Archidistoma molle 34 — + 57 94 23 — Aplidium solidum —23 88 S 20 H Polyclnum planum 199 88 Styela montereyensis —17 II I 80 Archidistoma psammion —80 II Ldiatoms 76 10 22 I dinoflagellates Synoicum parfustis 178 animals / misc. s 74 Hother algae L 200 50 100 50 Number of items in 5 x IO gramsof fecal pellets -25 Rice and Cotsirilos Figure 2 In order to show details better, most of the pelle are shown enlarged by the given factors. For in¬ stance, the sketch for Aplidium solidum must be reduced by a factor of six in order to make a scaled comparison with Styela. Arrow indicates separation point between two Styela pellets. The Styela is 8.5 times longer than life size. Rice and Cotsirilos Figure 2 Styela montereyensis 1 Clavelina huntsmani 2X Bol Synoicum parfustis Polyclinum planum 5X Archidistoma molle side view 3X Aplidium solidum 6X 3) cross section 2 -26 -27 Rice and Cotsirilos Figure 3 Horizontal line within a bar indicates the smaller of two dimensions. Vertical line superimposed on mouth size data indicates range of mouth sizes that were measured. Pollicipes larvae were measured along their greatest dimension which involved distances between laterally projecting appendages. Rice and Cotsirilos N E 5 wr . . 1020 azo00 9 11 0 EHDE vaao i r 1 1 0 l — 2092•0 H 5 * * t10 18h0 O O2br L 0 *. t L w4ro 1780 o jj so se talleel 0 o kk T —— aa- 2OX60 l — S -28- e Rice and Cotsirilos Figure 4 Diagram of mouth is depicted in the top row. Circles and rectangles in rows 2 and 3 repre¬ sent the largest area that can fit through the tentacular screen without touching a tentacle. Largest sand grain found in fecal matter of each animal is represented schematically in the bottom row. All diagrams for Styela, Ascidia, and Perophora are enlarged by a factor of 4.3. Diagrams for Synoicum and A. psammion are drawn at 2X relative to the other three species. Rice and Cotsirilos —— I J O O *- L C C -30