Rolling by Armadillidium vulgare and Predation. Introduction Armadillidium vulgare is a terrestrial isopod with the capability of rolling into a ball. The purpose of this study was to determine what effect the rolling behavior has on predation by a scratching animal. Previous studies have linked the rolling behavior of Armadillidium vulgare to various stimuli. Saxena (1957) was able to induce the behavior by using a hard tactile stimulus. Salamanders have been observed eating the isopods by Miller(1938). He reported that large individ- uals were able to escape predation by rolling into a ball which the salamanders were unable to swallow. Waloff (1941) suggested that the behavior may be a re¬ sponse to suddenly changing microclimate conditions. Observations on Armadillidium vulgare A set of preliminary studies were conducted to pro¬ vide qualitative insights into possible factors rela- tingto the rolling behavior. Humidity To determine if low humidity conditions stimulate rolling, isopods were put in a Stender dish set on Dri-Rite desiccator in a 250 ml covered beaker. None of the isopods rolled in this 0% relative humidity condition, supporting the findings of Waloff (1941). edation Temperature To determine if constant, extreme temperatures stimulate rolling, isopods were placed in three covered finger bowls with moist paper towels on the bottom. One was kept at 0°C, one at about 25°C, and the last at 35°0. At the time of death of the isopods in the low and high temperatures, no rolling had been observed. No rolling was seen in the animals at room temperature. Touch Tactile stimuli were observed to immediately initiate the rolling behavior, supporting the find- ings of Saxena(1957). Touching them with any hard object, shaking them in a container or in the hand, or dropping them onto a hard surface all produced roll¬ ing. Optimal Environmental Conditions In collecting Armadillidium vulgare from under iceplant above Agassiz Beach at Hopkins, it was noticed that the isopods were usually rolled after the iceplant was lifted up. This observation led to a hypothesis that the rolled position is the normal rest- ing state of the isopod when environmental conditions are the most favorable for survival. Warburg (1964) and Waloff (1941) both found that the activity of Rolling by Armadillidium vulgare and Predation Armadillidium is at a minimum at 100% relative humidity. The habitat under the iceplant was very moist, probably reaching 100% relative humidity among the dead leaves where the isopods were found. The rolled position of the isopod is an inactive state. Therefore, the hypo¬ thesis that rolling is an inactive state in response to a very humid, favorable environment seemed reasonable. An experiment was set up to investigate this. 250 ml beakers were filled with loose, moist dirt, some deep enough to allow the isopods to bury, and some with only a thin layer of dirt. The hypothesis was that animals that were able to bury themselves in the moist dirt. where desiccation stress is absent, would be in optimal conditions and would roll up. An attempt was made to instantly kill the isopods while still in position in the dirt by quick-freezing the beakers in an alcohol- dry ice bath. This failed, but by carefully picking through the dirt with forceps it was possible to uncover many isopods without disturbing them. No formal data was taken due to uncertainty about the accuracy of the digging method. However, those isopods uncovered with¬ out disturbance were not rolled. With this suggestion that the rolled position is not the normal, resting position of Armadillidium vulgare, it was possible to concentrate on a second hypothisis, the one that became the working hypothesis for this study. Rolling by Armadillidi and Predation Hypothesis Since the isopods were probably not usually in the rolled position, the fact that they were observed rolled under the iceplant suggested that lifting the iceplant caused the behavior. In order to lift the iceplant, roots had to be pulled up also, which disturbed the soil and dead leaves the isopods live in. The hypothesis that led to this study was that a scratching predator, like a bird, could cause the same kind of disturbance of the soil and organic litter that uprooting the iceplant did. This dis- turbance could produce a tactile stimulus for roll¬ ing, or the change in microclimate conditions which has been suggested as a possible stimulus. Observations of Chickens, Scratching Predators, Feeding on Armadillidium vulgare Dr. Robin Burnett, in personal communication, reported observing chickens eating isopods, using a scratching behavior to find them in leaf litter. Detailed observations were subsequently made of these chickens feeding on isopods in an area covered about an inch deep with dry oak leaves. The scratching behavior of a chicken, in feeding on animals in litter, Rolling by Armadillidium vulgare and Predation consists of two scrapes with each leg, throwing debris diagonally back and away from the chicken. Then the chicken quickly jumps straight back scraping a large amount of debris under its feet, held close together under its body. The chicken then closely examines the uncovered patch of dirt and picks up food using quick stabs with its beak. The inspection of a cleared patch rarely lasts longer than ten seconds. Isopods were frequently uncovered and eaten. Armadillidium vulgare were present in the area in very small numbers, the dominant species being Porcellio scaber. Armadillidium were placed in the area so that observations could be made more easily on their response to the chickens feeding behavior. When Armadillidium were uncovered by a chicken's scratch, they often appeared on the bare ground in the rolled position. They also appeared in an unrolled state, crawling across the dirt, or lying on their dorsal surfaces with their legs moving rapidly. These active isopods were usually eaten immediately by the chickens. Rolled isopods were also eaten, but usually after the active ones, and less frequently. In one case, a chicken uncovered an isopod, which rolled up. After passing it over, the chicken moved forward and began to scratch in a new area. Just as the isopod Rolling by Armadillidium vulgare and Predation was unrolling it was hit by leaves the chicken had scratched back from the new area. The isopod immedi¬ ately rolled up again. It eventually unrolled and moved quickly under the layer of oak leaves. The Effect of Rolling on Predation: Materials and Methods Chickens were therefore used as the scratching predators for the experimental part of this study. They are docile and easy to observe and eat isopods readily. An experiment was conducted to determine quantitatively what effect the rolled position of Armadillidium vulgare had on predation by the chickens. A number of isopods were glued, while in the rolled position, with a drop of Eastman Kodak 911 glue placed where the head and uropods join. When the animals were prevented from opening until after the glue had set, they were effectively forced to remain rolled. Twenty of the rolled isopods, and twenty normal isopods were mixed together and scattered on a patch of bare dirt near a chicken. One hen was used for all the trials. After the chicken had eaten about half the isopods, the remaining ones were collected and the animals of each type not eaten were counted. Another test of the same kind, using unrolled animals, half with a spot of glue on their dorsal surfaces, had shown that the presence of glue on the isopods produced no significant effect on Rolling by Armadillidium vulgare and Predation the feeding. A similar experiment was performed using ten normal isopods and ten dead, unrolled animals per trial to detenmine theeffect that simple movement had on preda¬ tion by a chicken. Again, the feeding was stopped when about half the isopods had been eaten. Results The results of the first experiment appear in Table 1. The results of the second experiment are in Table 2. In both experiments the unrolled, active animals were eaten first. In many cases, a moving isopod was neatly plucked from among several rolled or dead ones that were never eaten. Discussion From these experiments it appears that a moving isopod draws the attention of a feeding chicken more than a motionless one. When given a choice, a chicken eats the moving isopods first. Saxena found the mean length of time of rolling after a tactile stimulus to be about ten seconds. If Armadillidium uncovered by a chicken remained rolled for this length of time, they would stand a good chance of not being eaten, as the chickens rarely examined an uncovered area longer than ten seconds. An isopod which does not roll when uncovered is very likely to be eaten, as Armadillidium in the unrolled state is almost always moving, either crawling, or if on it's dorsal surface. waving its legs. A rolled isopod could also be pro- Rolling by Armadillidium vulgare and Predation pelled away from a chicken when the predator scratches, due to the sphere shape which rolls well on flat surfaces. Exactly what causes the isopods to roll when un- covered by the scratching is unknown. Turning over a rock under which isopods are living can also cause rolling. The stimulus could be tactile, changing light intensity or a sudden change in microhabitat conditions such as humidity, or temperature. The same stimuli could elicit the response when a chicken scrapes away a layer of leaves looking for food. The extension of the results of this study to con- clusions about the adaptive significance of rolling to Armadillidium vulgare are not possible. However, the behavior does appear to be a successful defense against some types of predation. Evolutionarily, whether the behavior was a direct adaptive response to predation or a response to some other pressure, which had the effect of limiting some types of predation is a matter of conjecture. Low humidity, probably the most lethal pressure the isopods face, has been shown not to be a stimulus for the behavior. (Waloff) Rather, the isopods become more active in low humidities. (Waloff, Warburg 1964) The assumption by Heeley (1941) that rolled isopods were conserving moisture in the respiratory pleopods in con¬ ditions of low humidity is made even more tenuous by actual water loss data. Edney (1951) found that more Rolling by llidium vulgare and Predation Arm than 50% of the water lost by Armadillidium vulgare in low humidity is through the dorsal surface. This sug¬ gests that were the isopod to roll in low humidity, water conservation would be minimal, considering the extra dorsal surface that is exposed when the animal rolls. Paris (1963) reported that drowning is a major cause of winter deaths, and this environmental pressure could have been a factor in the evolution of the rolling be¬ havior. Summary 1. Scratching by chickens in feeding causes Armadillidium vulgare to roll up. 2. Chickens eat Armadillidium that are active more than rolled or dead ones. Acknowledgement I wish to thank Dr. Malvern Gilmartin for his help throughout this project and in preparing the manuscript. Special thanks go to Dr. Robin Burnett, gentleman farmer, for the use of his chickens, without which this project would not have been possible. Rollir llidium vulgare and Predation Literature Cited Edney, E.B. 1951. Evaporation of Water from Wood- lice and Glomeris. J. Exp. Biol. 28(1):91-115. Heeley, William. 1941. Observations on the Life- Histories of Some Terrestrial Isopods. Proc. Zool. Soc. Lond. Ser. B. 111(½):79-148. Miller, Milton A. 1938. Comparative Ecological Studies on the Terrestrial Isopod Crustacea of the San Francisco Bay Region. Univ. Calif. Pubs. Zool. 43(7):113-142. Paris, Oscar H. 1963. The Ecology of Armadillidium vulgare in California Grassland: Food, Enemies, and Weather. Ecol. Monogr. 33(1): 1-22. Saxena, S.C. 1957. An Experimental Study on Thanatosis in Armadillidium vulgare. J. Zool. Soc. India. 9(2): 192-199. Waloff, Nadejda. 1941. The Mechanisms of Humidity Reactions in Terrestrial Isopods. J. Exp. Biol. 18(2):115-135. Warburg, Michael R. 1964. The Response of Isopods Towards Temperature, Humidity and Light. Animal Behavior 12(1): 175-186. 10 Rolling by Armadillidium vulgare and Predation TABLE ROLLED IROLLET 1 16 18 14 16 5 11 N-20 .001 ABLE TRIA NEATEN IEATE N-10 01 11