INFESTATION OF THE SANDY BEACH AMPHIPOD ORCHESTOIDEA CORNICULATA BY THE MITE GAMMARIDACARUS BREVISTERNALIS (LAELAPTIDAE) by Donna Scurlock Introduction In California sandy beach communities, members of the amphipod genus Orchestoidea are among the most abundant macrofauna. At night these beach hoppers leave their sand burrows and may be found feeding in large numbers on standed seaweed. As early as 1912. small, apparently ectoparasitic mites had been found on these gammarids (Hull, 1912), and in later studies of Orchestoidea both McClurkin (1953) and Bowers (1954) noted their occurrence. In both Oregon and Washington, Canaris (1962) found Orchestoidea californiana Brandt, 1851 infested with males, females, and deutonymphs of a mesostigmatid mite which he described as Gammaridacarus brevisternalis Canaris, 1962. These mites have also been discovered in the sandy beach wrack by both Irwin E. Newell (pers. comm., 1972) and the author. This investigation was undertaken to extend our present scanty knowledge of the nature and extent of the relationships between the parasite and host pop¬ ulations. Studies were carried out at Hopkins Marine Station of Stanford Univer¬ sity, Pacific Grove, California from April to June, 1972. All collections were made on the Monterey Boatworks beach located just east of Mussel Point, Pacific Grove. This beach was found by Bowers (1964) to contain a beach hopper pop¬ ulation consisting solely of Orchestoidea corniculata Stout 1913. All individ¬ collected ual amphipods were checked using keys in Bousfield (1959). Only 9. corniculata occurs on this beach at the present time. Mites on amphi pods The occurrence of mites on the host population Investigations were made to discover the extent to which the host population was infested by mites, and to see whether or not the mites show a preference for hosts of a particular size or sex. A total of 275 hosts was collected. Each individual amphipod was immediately placed in a separate vial of 75% ethyl alco¬ hol, thus any mites that became detached from their hosts could be counted and attributed to a single amphipod. In the laboratory, the contents of each vial were examined under a disect¬ ing microscope. Each amphipod was identified by species and sex, and the number of mites and their positions on the host were noted. Occasional mites were found to have come free of their hosts and were loose in the vials of alcohol. The sexes of the youngest hosts could not be determlned. In the case of brooding female hosts, the eggs or young were also checked for mites. The body length of each amphipod was then measured by straightening the body and recording the distance from the front of the rostrum to the tip of the telson. The O. corniculata col¬ lected ranged in length from 3mm, the size when the young first leave the mar¬ supium (McClurkin, 1953), to 24mm, the size of the largest males. Results of the survey (Fig. 1) show there is a positive correlation between host body size and the percentage of the host population infested. In fact, only amphipods one mite (a deutonymph) was found among the 63 of the smallest size class examined. The smallest amphipods have less exposed ventral surface area for mite attachment, and also show a pattern of behavior slightly different from that of the larger hosts. The youngest amphipods are more active than the larger ones during the days, and can sometimes be found hopping above the surface at that time (Bowers, 1964, and personal observations). Mites are very susceptible to desication (Cheng, 1964, p. 545) and conceivably, this host daytime behavior could subject the mites to drying conditions. Mites on amphipods The percentage of infested males of the largest size class was lower than that of the next smallest size class, while the females of the two largest sizes showed similar percentages of infestation. The decrease found in the largest males might be attributed to their tougher outer covering. Females of the largest size class were all close to the lower length limit of this range (20mm) and so were almost members of the next smaller class. The average size of the mite population on individual infested hosts increased slightly with host body size (Fig. 2), which is not surprising since larger hosts have more room for parasites, and have had a longer time in which to acquire them. One male in the largest size class carried a record-setting total of 33 mites. Among those 0. corniculata which were infested, small mite loads were more common than large loads (Fig. 3) Distribution and attachment of mites on the host body When live hosts were examined, occasionally unattached mites were observed crawling on the ventral or sometimes dorsal side of the host, or on its appendages. This behavior probably explains the presence of the small number of mites which detached from the host body on preservation. Attached mites, when observed on alcohol preserved hosts, were found exclus¬ ively on the ventral side of the amphipod body; here they ranged longitudinally from the area between the first gnathopods to a position directly above the first pleopods. Mites were found in every possible position on this flat, thin skinned surface, and also on sites on the five pairs of gills (Fig. 4). The ventral surface would be advantageous to a mite, for this area forms part of a protected and probably humid chamber when the amphipod curls its body and crosses its gnatho¬ pods and pereôpods across its ventral side during its daily resting period in its sand burrow. Mites on amphipods The number of mites on the ventral surface was greater anteriorly on the hosts while the number on the gills was higher posteriorly, where the gills are larger. In cases of heavy infestation, mites tended to cluster into dense packs in the center of the venter. The ventral surface of the host is relatively thinly armored and down the middle, passing close to the epidermis runs a large blood sinus (Cussans, 1904). The five pairs of gills also offer a rich blood supply. If the m tes are drawing fluids from the host, better places could not be found. Mites were never found attached to the host's tail area, probably because of the amphipods' mode of locomotion. When a beach hopper jumps, it curls its tail forward under the body, then straightens it suddenly in springing away. On brooding females, mites (8) were also found on the four pairs of oostegites or brood plates which retain the eggs or young. No mites were ever found on any of the eggs or young belonging to the brooding females examined. Mites were most often found attached to the amphipod by their mouth parts; their remaining appendages were held extended, but not attached. This position does suggest feeding. However, even the more heavily infested hosts showed no obvious signs of weakness, damage, or even inconvenience caused by the mites. Not all mites were attached at their mouths to the host. On the host's gills, mites often adhered by their posterior ends and had their appendages drawn close to their bodies. Possibly individuals are attached by a thread of sticky material ejected through their anus, as has been noted in some Uropodina (Hughes, 1959, p. 25). The behavior of mites during molting of the host was never observed, but it seems likely this presents few problems to the attached ectoparasites. Orches¬ toidea, after molting, remains dormant for a period, and then proceeds to eat its own shed exoskeleton (McClurkin. 1955). This host behavior would appear to give the mites plenty of opportunity to detach from the old skin and find a place on Mites on amphi pods the new, tender covering. The transfer of mites between hosts A small series of experiments was performed to determine if a shift in mites occurs from one living host to another, and from dead to living hosts. Tranfer between healthy hosts. One investigation was initiated to discover if the mites constantly remain with their hosts even in the presence of other heal thy amphipods. Live amphipods were first anesthetized with ethyl ether fumes and then were examined very carefully under a dissecting microscope. Both mites and amphipods quickly and completely recovered from the anesthetic. Amphipods were maintained in the laboratory using the methods of McClurkin (1953). Into two 2cmx 9cm glass vials were placed cotton pads moistened with seawater, and the tops of the vials were covered with cotton gauze to prevent amphipod escape. One infested and one uninfested 0. corniculata were placed in each of the two containers. After one to two days, each previously uninfested amphipod bore at least one mite. The mites are apparently relatively motile. They can shift from one living host to another, at least when the hosts are close together. This physical prox¬ imity occurs in the field especially during the night, when many closetly packed O. corniculata can be found feeding on the wrack. Mite reaction to host death. A similar experiment was set up to find out the reaction of mites to the death of their host. Two live 0. corniculata each infested with two mites were killed by stabbing. They were immediately placed in two McClurkin vials and left overnight. The dead hosts bore no mites the next morning, and even the cotton in the vials appeared to be free of mites. The same experiment was repeated in an aquarium tank filled with four inches of damp sand similar to that in which the amphipods burrow during the day. Five Mites on amphipods infested 0. corniculata were killed by stabbing and placed overnight in shallow holes in the sand to simulate the death of amphipods in their burrows. All the amphipods were found the next day to have no mites. Apparently the mites react to the decomposing host and leave. Einding a new host. Since mites leave after the death of their hosts, other experiments were performed to determine whether mites can locate new, healthy hosts. In one experiment, two dead infested amphipods were placed in separate Mcclurkin vials, and a live, uninfested 0. corniculata added to each vial. One dead amphipod bore one mite, and the other two mites, at the start of the ex¬ periments, but after 12 hours all the mites had transfered to the living amphi¬ pods. A similar experiment was carried out in the sand tank. Five infested O. corniculata with one mite each were killed and placed in holes in the sand, then three uninfested live amphipods were added. After nine hours, one living O corniculata individual bore all five mites and the remaining live and dead amphipods had no parasites. In these experiments, contact between the dead and living hosts was pos¬ sible. In the field, 0. corniculata have been observed eating dead individuals of their own species (pers. obs.), a contact which should provide a good oppor¬ tunity for mite transfer. Experiments were also performed to discover if mite transfer could take place without close contact between hosts. Under normal beach conditions this might involve mite movement across the beach. In order to test the mites' capacbilities of locomotion in sand, a round dish of beach sand was prepared. A mite was introduced at a marked spot on the sand, and allowed to move freely. The mite's progress was observed under the dissecting microscope, and it appeared to have no great difficulty traveling on the sand. Several mites, clocked for Mites on amphipods speed across the sand, averaged 3 inches per minute. When a living 0. cornicu¬ lata was placed at a point across the dish of sand from the mite, the mite was observed to crawl directly across the sand and climb on the amphipod. With these results in mind, a more complicated experiment was performed to test mite transfer from dead to living hosts over varying distances. A large aquarium tank was divided by metal sheets into separate compartments, measuring about 8cm x 42cm. Sand from the normal environment, treated with steaming hot water for fifteen minutes to kill any mites, was drained, dried to the dampness 0. corniculata prefers, then placed into each section of the aquarium, and the surface roughly leveled. Living, uninfested Orchestoidea were caged in cylinders of plastic screen of Imm square mesh, two animals to a cage. A cage of amphi¬ pods was placed at one end of each of the aquarium compartments. Three fresh 0. corniculata, each with three mites were killed and placed on sand in separate runways of the aquariùm. One dead infested amphipod was placed at 10 cm. away from the cage in runway A, one was placed 20 cm. away in runway B, and one at 30 cm. in runway C. Runway D contained live caged uninfested hosts and no dead host with parasites. After nine hours (overnight, in darkness), the cage in run¬ way A contained one amphipod with two mites and one with one mite (no loss of mites over 10 cm.).. In runway B, one caged amphipod bore three mites, the other none (no loss of mites over 20 cm.). In runway C, each caged amphipod bore one mite (loss of one mite over 30 cm.). No mites were found on the caged amphipods in control runway D. The results of these experiments suggest the mites may possess senses enabling them to locate hosts over considerable distances,a worthwhile subject for further study. Mites in other environments than the host. The foregoing observations and experiments suggest that some part of the mite population is existing apart from the 0. corniculata population at any one time. Alternative environments for these Mites on amphipods mites appear to be (1) other host species, (2) damp sand, and (3) damp wrack on the beach. As previously noted, mites also infest the Orchestoidea californiana population. Mites that had the appearance of those infesting 0. corniculata were found in older, partially buried wrack. Dr. Newell has identified them as the present species, G. brevisternalis, and noted (pers. comm. 1972) that he has found these mites in southern California exclusively in the wrack. Samples of the mites from Pacific Grove wrack, which appeared identical to the amphipod mites when compared under a compound microscope, were introduced into a vial with live uninfested amphipods. A day later they were found attached to the O corniculata. The mites can live for prolonged periods apart from the host. brevisternalis have been reported to live over a month on decomposing wrack placed in closed plastic bags (Helen Kompner, pers. comm., 1972). I have kept the mites stoppered in vials containing only tiny bits of beach wrack, and also in vials containing only a pad of sea water soaked cotton for at least three and one half weeks. In contrast, mites placed in an empty, corked glass vial over¬ night were all dead the next morning. Perhaps when the mites can not locate a new host quickly, they find in the wrack a humid home to await eventual arrival of amphipods which normally feed nightly on the wrack. The true relationship of the mites to the wrack and to their amphipod hosts still remains a mystery, but whatever the answers, these tiny animals seem to have worked out admirable means for survival. Summary The mesostigmatid mite, Gammaridacarus brevisternalis has been found and identified on both decomposing beach wrack and on the beach amphipods Orchestoidea corniculata and O. californiana. Mites on amphipods 2. The percentage of infestation of the host population increases with size of host and varied from 1.5% (hosts 3 - 7.9 mm) to 83.07% (hosts 16 - 19.9 mm). Mites showed no preference for male or female hosts. The number of mites per intested host increases slightly with amphipod size. 3. Mites were found attached, either by mouth parts or posterior end, exclu¬ sively on the ventral side. 4. The mites were found to leave dead hosts within 2 - 9 hours after death. 5. Nites without hosts can crawl over sand at an average rate of 3 inches per minute. Gammaridacarus brevisternalis were found, after leaving dead hosts, to attach to new, living hosts. When contact between dead and living hosts was pre¬ vented, the mites traveled up to at least 30 cm in finding new hosts. Ob¬ servations suggest that the mites may possess senses for locating healthy hosts at a distance. Acknowledgements I would like to thank Dr. Irwin E. Newell of the University of california. Riverside, for his helpful suggestions and identification of the mite. Sam Johnson, a graduate student at Hopkins Marine Station at the time of this study. was very patient with my many questions about amphipods. Most importantly.I would like to express my deep appreciation to Dr. Donald P. Abbott for his guidance and enthusiasm. Mites on amphipods Literature cited 1959. New records of beach hoppers (Crustacea: Amphipoda) Bousfield, E. from the coast of California. Nat. Mus. Canada Bull. 172: 1-12. Bowers D. E. 1964. Natural history of two beach hoppers of the genus Orchestoidea (Crustacea: Amphipoda) with reference to their complemental distribution. Ecology 45: 677-696. Canaris, A. G 1962. A new genus and species of mite (Laelaptidae) from Orchestoidea californiana (Gammaridea). J. Parasitol. 48 (3) 467-469. Cheng. Thomas C. 1964. The biology of parasites. W. B. Saunders, Phila. 727 p. Cussans, M. 1904. Liverpool marine biology committee memoirs. XII Gammarus. Williams and Norgate, London. 47 p. Hughes, T.E. 1959. Mites or the acari. Athlone Press. 209 p. Hull, H. V. 1912. Some marine and terrestrial acarina of Laguna Beach. First annual report of the Laguna Marine Laboratory. Pomona College Press. 218 p. McClurkin, J. 1., Jr. 1953. Studies on the genus Orchestoidea (Crustacea: Amphipoda) in California. Unpublished Ph. D. thesis, Stanford U, Publ. 5803 Univ. Microfilms, Ann Arbor, Mich. Mites on amphi pods Figure captions Percentage of the Orchestoidea corniculata population, by size class and Fig. I se e, found infested with mites. Sex of members of the smallest size class could not be determined.' Females of the largest size are very scarce. Fig. 2 Average number of mites per host on infested Orchestoidea corniculata, according to host size class and sex. Each column shows mean, range, and 95% confidence limits of the mean. Sex of members of the smallest size class could not be determined. Fig. 3 - Degree of infestation with mites of all 0. corniculata examined. Fig. 4 - Positions on host body at which mites were found attached in a survey of 275 amphipods. An 0. corniculata (male) is shown with appendages of the left side removed. Gill pairs are numbered 1 through 5 from front to rear. In addition, 8 mites were found on female oostegites. 5 38 5 2. 75 50 25 15 10 5 20 45 % of Q. corniculoto infested O 35 23 n=63 3-79 8-11.9 16-19.9 12-15.9 Size classes (body length, mm) Average no. mites — per infested host Q — 20 37 I 17 10 9 8-11.9 16-19.9 12-15.9 Size classes (body length, mm) 3-7.9 34 20-2? 9 N 9 20-23.9 i 2 40 30 20 10 ... I 44 + ..:1:: ..... ... .. N=21 :: ++ :: 18 32 34 8 10 12 14 16 No. of mites per hos + — — i ::: — .: i re 80 60 40 0 20 2 60 40 5 20- Z 2 A , A - + + — ...... . Ventral surface position 4 3 Gill pair . — 5 . L