Gagbrn Abstract Caprella californica, C. brevirostris, and C. penantis may occur together on the hydroid Aglaophenia struthionides. Caprella incisa may occur on the bryozoan Tricellaria occidentalis growing in close proximity. Cleaning and feeding behavior of these caprellid species were compared under conditions simulating natural water flow. All species were seen to engage in filter feeding and some in scraping and macrophagous feeding. Quantitative comparison of the main components of feeding behavior in the three species on Aglaophenia shows little obvious evidence of resource partitioning, and the role of motility and positioning are con- sidered. Grogborn Introduction Mayer (1882) and subsequent workers have noted that the vast majority of caprellids are found on living substrates. Numerous species occur on hydroids, and a variety of rela¬ tionships between the caprellid and its hydroid host have been reported. Lockington (1875), Mayer (1882), and Mackay (1945) stated that caprellids eat or parasitize hydroids. Green (1963) concluded that hydroids were utilized solely as a substratum, while McDougal (1943), reported that the major food source of caprellids were sessile protozoans and diatoms growing on the hydroids. Food acquisition by filter feeding was noted by Saunders (1965), Patton (1968). and Keith (1969). Keith concluded that caprellids were opportunists, utilizing the most readily available organic material as food. This appears to consist mainly of floating organic detritus in their natural habitat. Caine (1974) examined the possible relationship between food sources and appendage structure in various sympatric caprellid species, In a subsequent paper (Caine,1977), he reported that most co-occuring, filter-feeding caprellids don't utilize the same size of particle at the same height above the substratum. At Mussel Point, Pacific Grove, California, 3 species of Caprella (C. penantis, C. brevirostris, C. californica) occuring on the hydroid Aglaophenia struthionides (Murray) and one species (C. incisa) associated with the bryozoan Tricellaria occidentalis (Trask) were examined. Since four species of caprellids have been found in the same small area, three of these in the same microhabitat,it seems likely that resource partitioning of some type occurs, Gragborn particularly with respect to food. This possibility was investigated in the present study. Materials and Methods Four species of Caprella were collected from intertidal waters surrounding the northeast shore of Mussel Point, California. C. californica, C. penantis, C. brevirostris, and C. incisa were collected by gathering colonies of Aglaophenia struthionides and nearby substrates. Prionitis was the substrate for A. struthionides in all cases and Tricellaria occidentalis was found growing at the base of A. struthionides in approximately one fifth of the collections. Specimens were gathered within one hour of low tide and transferred to the lab in jars filled with sea water. Only specimens submerged at low tide were collected. Upon return to the lab, caprellids and their substrates were placed in bowls filled with seawater and most of the Prionitis (hydroid substrate) was removed. Fresh seawater at a temperature of 10-15 degrees C was allowed to flow into the bowls. The T. occidentalis was identified by C. Baxter of Hopkins Marine Station. Feeding behaviors were observed for each species with the aid of an event recorder coupled to a strip chart recorder, permitting the recording of 8 different categories of stereo- typed behaviors. A chamber for the observation of caprellids on Aglaophenia consisted of a 15 cm length of glass tubing 2.2 cm in inside diameter, fitted at both ends with single hole rubber stoppers and provided with a constant supply of fresh seawater flowing at the rate of 132 ml per minute. Linear flow past caprellids was approximately i cm per second. This Gragborn system was designed to simulate in some degree the effects of local turbulance and water flow in the natural habitat. Substrates with resident caprellids were pinned to the inlet rubber stopper. Feeding behaviors were observed with a stereo- microscope. Two ten minute feeding observations were made of the same individual caprellid on the same Aglaophenia plume at twelve hour intervals. After two 10 minute observations were completed, milk was injected to study turbulance and particle settling in the region occupied by the caprellid. Caprellids were seen to move in response to contact with the milk. No further behavioral observations were made until all visible traces of milk in the tube had disappeared. All in- dividuals were examined 12 hours after the last experiment in which they were used, and results are presented only for animals still alive at that time. Distribution information was obtained in the lab within 2 hours of collecting A. struthionides in the field. Hydroid colonies were collected submerged in separate glass jars, and placed in glass bowls filled with seawater. With a bowl under the dissecting scope, the number of adult and juvenile caprellids on the proximal and distal half of each frond were recorded under minimal illumination. Species identification of juveniles was not feasible. After caprellid size, species, and distribution was recorded, the frond under consideration was removed from the colony with forceps and measured. Once removed, fronds and accompanying caprellids were not returned to the bowl. Line drawings of feeding appendages of the four caprellid species encountered were made with the aid of a camera lucida Gragborn and a compound microscope. Results Distribution Distribution of adult and juvenile caprellids on Aglaophenia half fronds is presented in figure i. More juveniles than adults were present on Aglaophenia fronds of all size classes. Species of juveniles was not determined. The distinction between adults and juveniles was based on size and to some extent, color. Transparent individuals smaller than 2.2 mm were classified as juveniles. Transparent individuals larger than 2.2 mm and caprellids with coloration in regions other than the gut (generally larger than 2.0 mm) were considered as adults. Distribution of adult species is summarized in table 1. Only one species (C. penantis) is common on A. struthionides, It therefore seems likely that most juveniles are C. penantis. Other species occur only sporadically, and their main dis- tribution may be elsewhere. The skewed distribution of species suggests that resource partitioning may not be important in this microhabitat. Feeding Behavior Five types of feeding mechanisms have been described in adult caprellids, including browsing, filter feeding, predation, scavenging, and scraping (Caine, 1977). In the present study, predation and scavenging were not observed. The lack of predation may be due to rem val of prey items by filter feeders present in the laboratory seawater pipes. The current maintained in the observation chamber may have washed away organisms which weren't actively clinging to the substrate, Grag botn thus offering a possible reason why scavenging wasn't noted. Of the remaining three types of feeding, filter-feeding was frequently seen; no distinction was made between browsing and scraping, and these are considered together. Caprellids were observed to situate and orient themselves such that they were out of the main current and exposed to flow rates generally less than 1 cm per second. The adults of all 4 species investigated often stood virtually motionless, taking advantage of eddy currents created by the substrates. Positioning was such that particulate matter present in the water column drifted past or fell slowly on the caprellid. Browsing and scraping seem to be related to the side of the frond where a caprellid is found. Observation of both sides of Aglaophenia fronds showed a much greater frequency of encrusting organisms on the abaxial side. In the flow tube, C. californica was found on the abaxial side of the Aglaophenia frond 78% of the times observed, while C. penantis and C. brevirostris were found in this orientation 70% and 65% of the time respectively. Filter-feeding and related behaviors, as well as scraping, were subdivided into eight classes of stereotyped behavior, defined below. A summary of feeding behaviors of four species of caprellids is presented in figure 2. Behavior 1 represents the bending of the first antenna over the cephalon. Gnathopod 1 on the same side of the body is then closed around the base of Al, which is drawn through Gi, presumably removing trapped particles (figs. 3 & 4). Behavior 2 is similar to behavior 1, except that antenna 2 is being cleaned instead of antenna 1 (figs. 3 & 4). Grabot Behavior 3 represents simultaneous bending of antennae 1 and 2 over the cephalon. Gnathopod 1 on the same side of the body is then closed around the bases of Al and A2, which are drawn through Gl, removing trapped particles (figs. 3 & 4). Behavior 4 signifies bodily cleaning from the gills to the distal pereopod. Gnathopod 1 picks and/or scrapes material from the body (figs. 3 & 4). Behavior 5 represents the removal of material from G2 by Gl (figs. 3 & 4). Behavior 6. This involves pulling or scraping material from the substrate by Gl (figs 3 & 4). Behavior 7 represents active gathering of food. It includes removal of detritus directly from the water column with both first gnathopods, where it is then held in place. Larger detritus may be similarly captured by all four gnath¬ opods and held in place (figs. 3 & 4). In C. incisa, waving of both first gnathopods through the water, presumably filtering food from the water, has been observed. This feeding mechanism is also included in tabulated results. Behavior 8 considers movement of Gl to the caprellid's mouth, where the maxillipeds contact it (figs. 3 & 4). Statistical analysis using the Student-Newman-Keuls test indicated significant differences (p=0.05) between C. incisa and C. californica in frequency of behaviors 4 (cleaning of body with Gl from gills to distal pereopod) and 7 (active gathering of food, including catching of large detritus with Gl and G2, and combing of water with G1); between C. incisa and C. brevirostris in behaviors 4 and 5 (cleaning of G2 with G1), and in behavior 5 for C. incisa Gragb and C. penantis. Some of the behaviors found in all four species occured in typical coordinated sequences. Cleaning of antennae, G2, and the body with Gl, and scraping were often followed by clean¬ ing of Gl with the maxillipeds. Walking was carried out in the inchworm fashion typical of the caprellid. Variance for data related to walking was high, so this information is presented in two ways. First, data were pooled and the average number of walks per individual per 10 minute trial was determined for each of the four species examined (table 2). Next, the number of 10 minute trials in which walking occured and the number of trials in which it was absent are compared for four caprellid species (table 3). The four species of caprellids taken as a group were found to differ significantly (p=0.05) in regard to the number of trials in which walking was or was not present. Miscellaneous natural history Caprellids can be kept alive for periods of 2 weeks and perhaps indefinitely at room temperature in glass bowls using 1 mm mesh nylon screen as the sole substrate. If held down with glass rods, caprellids could easily walk on it, or cling to one location. Nauplius larvae from the goose barnacle Pollicipes polymerus were used as a food source and provided every 3-4 days. Larvae were stained with toluidine blue and the guts of caprellids began to assume this color soon after larvae were introduced. A stirring bar with a plastic blade at the tip, suspended from a motor above the bowl, was used to prevent the nauplii from sinking to the bottom where they were unavailable to the caprellids. In cases where hydroids, Grayborn algae, or bryozoans were left in the bowl,the water was fouled within a few days. Cryptic coloration was obvious in various species of caprellids. All juveniles observed were transparent, making them difficult to see. On adults of C. incisa found on T. occidentalis, short bars of deep red against a light background made this species inconspicuous on its substrate, This caprellid was not found further than 1.5 cm from T. occidentalis. C. penantis adults had a mottled orange brown col- oration matching well the background presented by Aglaophenia C. californica adults possessed red bars running across their bodies and gnathopods. Adults of C. brevirostris were transparent except for their eyes, as well as their guts, which tended to assume the color of food material within. Discussion The need for resource partitioning among co-occuring species of caprellids was addressed by Caine (1977). He found that co-occuring, filter-feeding caprellids tend to utilize particles of different sizes, collected at different heights above the substratum. He noted that niche diversity in caprellids may involve morphological, behavioral, or physiological features of the species. Few differences in feeding behavior and morphology of feeding appendages were found between the species present on A. struthionides. Due to the great abundance of C. penantis relative to other species, resource partitioning may not occur. Competition may well be limited and other species (such as C. incisa) may have other preferred substrates. In addition, it may be that experimental designcombined with the opportunistic nature of 10 Graborn feeding has obscured certain secondary feeding behaviors. The caprellids investigated differ in the amount of walking done, and in the positions and locations they assumed (tables 2 & 3 ). Characteristically in Aglaophenia a group of fronds radiate from a compact holdfast at the base, such that different fronds receive different amounts of exposure, some being shielded by surrounding fronds. Shielding may influence the location a caprellid uses for feeding. This is evidenced by caprellid positioning in slow currents to take advantage of particle settling. Positioning on the axial or abaxial side of a frond seems to be related to the availability of encrusting organisms which can be utilized as food, since caprellids have been observed more frequently on the abaxial side. Caprellids found on Aglaophenia were sometimes observed to scrape a certain region and then move. Walking may be a response to depleted local supplies of encrusting organisms available as food. In addition, a wa king caprellid may be "seeking" a site suitable for filter-feeding in terms of currents and food. Acknowledgements I would like to thank Dr. D. P. Abbott, whose comments and criticisms aided and added to this paper, and without whose help this paper would not have been possible. Gragburn Literature cited Caine,E.A., 1974. Comparative functional morphology of feeding in three species of caprellids (Crustacea: Amphipoda) from the northwestern Florida Gulf Coast, J. exp. mar. Biol. Ecol. 15, 81-96. Caine,E.A., 1977. Feeding mechanisms and possible resource partitioning of the Caprellidae (Crustacea: Amphipoda) from Puget Sound, USA. Mar. Biol. 42, 331-336. Costa,S., 1960. Note preliminaire sur l'ethologie alimentaire de deux Caprellides de la rade de Villefranche-sur-Mer. Trav. Sta. zool. Villefranche, 19 (20): 103-105. Green, J., 1963. A biology of Crustacea. H.F. & G. Withersby, Ltd., London, 180 pp. Keith, D. E. 1969. Aspects of feeding in Caprella californica Stimpson and Caprella equilibra Say (Amphipoda). Crustaceana, vol. 16, no. 2, pp. 119-124. Lockington, W.N. 1875. Observations on the genus Caprella, and description of a new species. Proc. Calif. Acad. Sci., vol. 5, pp. 405-406. Mackay, D.C.G., 1945. Notes on the aggregating marine invertebrates of Hawaii. Ecology, vol. 26, pp. 205-207. Gragbot Mayer, P. 1882. Die Caprelliden des Golfes von Neapel. Fauna and Flora des Golfes von Neapel, vol. 6, pp. 1-201. MoDougal,K.D., 1943. Sessile marine invertebrates of Beaufort, North Carolina. Ecol. Monogr. vol. 13, pp. 321-374. Patton, W.K., 1968. Feeding habits, behavior and host specificity of Caprella grahami, an amphipod commensal with the starfish Asterias forbesi. Biol. Bull. mar. biol. Lab, Woods Hole, 134 (1), 148-153. Saunders, C.G., 1965. Dietary analysis of caprellids (Amphipoda). Crustaceana, Vol. 10, pp. 314-316. 2 Graborn mtvaa- oooooo oooo oo oo ooo oooo stataa- 8 9 Grgootn Species C. californica C. penantis. C. brevirostris C. incisa — Average no. walks/ 10 min. trial 0.92 0.12 1.52 0.06 No. trials 12 21 17 Jable 2 0 Species C. californica C. penantis C. incisa C. brevirostris Jahle 3 + trials walking observed — f trials walking absent 16 16 11 Grayborn Gragborn 17 f o 93 KEY Juveniles Adults 114 48 34 97 32 55 . . . . . 1-5 6-10 11-20 21-30 31-40 41-50 51-80 Length of A. struthionides frond, mm Gragborn 15 15 8 Caprella penantis - 16) C. californica Ss E (12) C. incisa E (18) C. brevirostris H e (21) L 2 6 5 4 3 1 Behavior Type 2 Graborn Ro 22 X A2 4 A N a 61 im F2 A2 D V VA Grayborn 20 Foe 4 G2 S 32 7 G2 O.1 mm A K ITE B G1. 4 + ppt D G2 G1 Graybirn Table captions Table 1. Summary of the distribution of adult caprellids on Aglaophenia struthionides. Table 2. Average number of times spent walking per ten minute trial at constant flow rate. Table 3. Comparison of number of 10 minute trials in which walking was seen and in which it was absent. . Gragborn Figure captions Figure 1. Average number of caprellids per A. struthionides half frond versus length of A. struthionides frond. Total number of caprellids are indicated above each bar. Figure 2. Frequency versus feeding behavior type for caprellids associated with the A. struthionides community. Number of 10 minute observations is indicated to the right of each graph. 1: cleaning of Al with G1;2: cleaning of A2 with Gi ;3: cleaning of Al & A2 simultaneously wiyh Gl; 4: cleaning of body with Gl from gills to distal pereopod; 5: cleaning of G2 with G1;6: scraping of substrate with Gl or contact of maxillipeds with substrate; 7: active gathering of food, including catching of large detritus with Gl & G2, and combing of water with G1;8: cleaning of Gl with maxillipeds, Figure 3. A. Caprella californica: female ; antenna 1, antenna 2; B.C. penantis: male ; antenna 1, antenna 2; C. C. brevirostris: female ; antenna 1, antenna 2; D. C. incisa: female; antenna 1, antenna 2. Figure 4. A. Caprella californica: female ; gnathopod 1, gnathopod 2; B. C. penantis: male; gnathopod 1, gnathopod 2; C. C. brevirostris: female; gnathopod 1, gnathopod 2; D. C. incisa: female; gnathopod 1, gnathopod 2. 22