r Predation in Porcellio Spide BSTRAC A population of the theridiid spider Steatoda grossa in Pacific Grove, California was studied to observe its predatory behavior towards the terrestrial isopod Porcellio scaber and subsidiary prey items. The isopod and spider are both found to be nocturnally active. The carcasses of consumed prey are cut from the spider's web to collect in a debris pile which can be examined for an indication of long term dietary consumption. The isopod, though rejected as prey by most spiders because of repugnant tegumental glands, composes 84% of the diet of this spider as determined by web analysis. The spider from laboratory experiments was found to go about five days between meals but able to tolerate periods of three weeks or more without feeding. Spider Fredation in Porcell: INTRODUCTION Cloudsley-Thompson (1958), reported birds, reptiles amphibia, and many other insectivorous animals including spiders, harvestmen, mites and centipedes as predators of terrestrial isopods. For central Californian woodlice, Miller (1938) listed as vertebrate predators "at least two species of salamanders, several species of reptiles, birds, and insectivores; among the invertebrates, the black widow spider (Latrodectus mactans) and various species of centipedes". Gorvett (1956) has stressed the importance of the tegumental glands as defense mechanisms in isopods; these organs only occur in terrestrial species and the fact that their secretions are distasteful to many spiders suggests that spider predation has provided the strongest selective pressure for their evolution. Spiders thus appear to be at least potentially the major predators on woodlice, yet detailed studies of spider predation on isopod populations are almost non-existent. Porcellio scaber Latreille, 1804 is the terrestrial isopod which shows the greatest development of the tegumental glands (Gorvett, 1951). This species is a member of the cryptozoan community, a term coined by Dendry (1895) to describe the assemblage of small terrestrial animals found dwelling in darkness beneath stones, rotten logs, bark of trees and other similar situations. Though cosmopolitan in distribution. and one of the commonest isopods in the United States, little study has been devoted to its interspecific relationships, particularly those of predation. Several groups of arachnids were initially Spide edat in Torcellio observed in their contacts with Pscaber under laboratory conditions; these included solifugids, harvestmen, wolf spiders, orb weavers, pholcids and theridiids. The theridiid house spider Stoutoda grossa (C.L. Koch) was found to be the most active predator on P. scaber of all arachnids studied. It is the purpose of this paper to discuss the natural history of this spider, particularly its predatory behavior towards P. scaber - All studies were carried out at the Hopkins Marine Station and around private homes in Pacific Grove, California during the period April 27 to June 9, 1973. P. scaber populations were collected from both the grounds of the marine station and from a private estate near Asilomar Beach. The populations were maintained in the laboratory in 30 x 20x 10cm plastic tubs. A thin layer of dirt was spread over damp paper towels on the bottom; this was covered with a layer of grass and pine needles; and this, in turn, topped with small pieces of decaying wood and bark. Tubs were covered with aluminum foil and humidity was maintained at a high level. As long as the paper towels were kept damp an ideal environment was maintained. Most work on S. grossa was conducted in garages and other human habitats. The main study area consisted of one wall of a garage of a private home near Asilomar Beach (fig. 1), which supported 115 individual webs with 76 observed spiders. Other methods used in specific investigations are indicated in the appropriate sections which follow. spi rodation in orie IErIVTTY PTTEENS OE TONCTIOE Initially activity pattorns of the isopod were monitored in a natural onvironment on the grounds of the Hopkins Marine Station. Twenty-two traps consisting of plastic drinking cups were buried, with the lip of each cup at ground level, in dirt and sand surrounding a large patch of iceplant (Fosembryanthemum) bordering the east beach on Mussel Foint. These traps effectively captured and retained all isopods which wandered over the lip. In addition, four 0.25m2 arcas of nearby ranite rock surface were also monitored. Counts were made every hour for 24 hours on numbers of . scaber trapped in cups or detected on the rock surface. Since activity proved largely nocturnal. these populations were subsequently monitored on three successive nights, and data gathered on physical parameters. Temperature readings were taken with a thermocouple heat probe and relative humidity with a Honeywell portable relative humidity indicator from Scm above the ground surface. Temperature was also recorded from the iceplant habitat proper. The results of these studies (figs. 2, 3) show that activity in the P. scaber population increases at sunset and ceases with approaching sunrise. Relative humidity consistently shows a peak near 0300, with a corresponding temperature minimum. Activity is maximum earlier and it declining at this time. Activity from sunrise to sunset is negligible. P. scaber is sensitive to desiccation (Heeley, 1941); in this connection it maintains a nocturnal activity pattern, avoids dry places, and remains in a humid environment. Judging from the data in fig. 3, activity is most closely related to the absence of light ther than to particular conditions of temperature Spider Prodation in poi or humidity. The light regime provides the most constant parameter of the environment and one easiest for them to dotect. Numidity may also be of importance, however; on the second night, and especially on the third night of study (when bocause of heavy fog, the humidity remained close to 100%) nocturnal activity continued until a later hour at night. ATI TY PAT ETERN OT GTEATODA CROSCA Acr The activity pattorns of 76 3. grossa were monitored in the garage study area near Asilomar Beach (fig. 1). An atomiser which sprayed a very fine mist of water wha used to delingate the extent of each web. The dreplets adhered to the web without damaging it and without harm to the spider, and refracted the light brilliantly. Within 20 minutes evaporation restored the web to its original state. Each spider was observed every two hours over à 24 hour period. Spiders were scored as inactive if hiding in their daytime retreats, and as active if they were in a typical predatory stance in the web or were actively out and moving about. Temperature, relative humidity and light exposure were recorded every hour from one typical web. During the night spiders were observed with a flashlight covered with a red filter, a procedure which appeared not to disturb them. The results (fig. 4) showed there is much variance among the individual spiders with respect to their periods of activity. Some individuals were active in the predatory stance during the whole observational period; others were not active at all Spider Predation in Porcollio although they had been seen previously. However, on the whole the population showed much greater activity between sunset and sunrise than during the day (fig.5). Relative humidity and temperature showed a correlated maximum and minimum respectively, between 0600-0800. Thus it was found that S. grossa, like P. scaber, is principly nocturnal. Spiders are potentially diurnal having an integument containing a waxy epicuticle which acts to prevent desiccation. The more primitive groups of spiders are secondarily adapted to nocturnal habits, probably as a result of competition with more efficient species (Cloudsley-Thompson, 1958). The fact that S. grossa feeds upon P. scaber, a species rejected by more advanced spiders, may further indicate its inability to compete with these forms. In S. gross as in P. scaber, the period of activity corresponds best with the period of darkness rather than the temperature or humidity directly, emphasizing that behavior is cued mainly to light. Predatory behavior is dealt with in a separate section to follow, but a few observations of other aspects of S. grossa activity and behavior are included here. Several of the spiders maintained in the lab were observed to spin egg cocoons, one spinning a second two and a half weeks after the first (fig. 6). One typical egg sac contained 129 eggs. The eggs are enclosed within a hollow,silk sac with a cavity of larger volume than the eggs. When rotated the sac shows the eggs to sometimes be tightly packed together in a mass or else loose and able to roll around. Four weeks after formation of the cocoon, the eggs within had hatched and gone through at least one molt Spider Fredation in Porcellio to produce unpigmented pre-spiderlings with definite body form and capable of limited movement. Since S. grossa is sedentary, the question of where and how it obtains water is of interest. During experiments with the fine mist atomizer, the spiders were observed to drink water droplets from the threads of the web. The legs and pedipalps were also used to collect droplets from the web and bring them to the mouth. This suggests that collection of dew condensing on the webs provides an important source of water for the spider. PREDATORY B HAVIOR OF ST ATODA GROSSA Members of the family Theridiidae characteristically build irregular webs from the threads of which they suspend themselves in an inverted position while they await their prey (Levi, 1968). This is what I term the predatory stance in S. grossa (fig. 7). Both males and females were observed to spin webs, sex determined in males by the presence of pedipalps swollen at the distal end with the reproductive apparatus. All of the larger web spinners were females as evidenced by egg laying activity. The web of S. grossa is a more or less closely woven sheet extending in a single plane and consisting of threads running in all directions with no apparent regularity (fig.8). The sheet extends outwards from a hidden retreat in a crack or crevice, in which the spider is generally found when not in the predatory stance. These are the two basic states of activity. Occasionally a spider prowls its web laying new threads and actively repairing damaged portions of the web. Litter or dead carcasses are cut out of the web and dropped, where they form Spider Predation inT an accumulated garbage heap. As is typical of primitive web builders (Kullman. 1972), the web is used only as a means of getting information about the position of approaching prey, and does not use special viscid threads to snare its victims. Much silk is required and the webs are rather complicated and irregular. Vertical tangles of interconnecting scaffolding lines above, and drop lines below and to the sides anchor the main sheet of the web in place and provide additional means for detecting and entangling passing prey. Sometimes the webs are so compacted in a corner that they lose their characteristic shape. Once alerted to the presence of prey by vibrations of the web, the spider advances rapidly and touches the isopod or other prey with the pedipalps and first pair of walking legs. At this point the prey may be either accepted or rejected. If accepted, S. grossa proceeds to swath the ventral surface of P. scaber with viscid threads pulled from the spinnerets by the fourth pair of walking legs. The prey is not rotated as seen in species of Araneus, and not even held, but rather is crisscrossed ventrally with silk in the place of entanglement. The spider places occasional quy lines to the dorsal surface of the prey and attaches them to its web, acting to hitch the prey up into the web. It may pause occasionally and if the struggling is strong it continues to swath the prey in silk. A series of short bites with the fangs is then made through the ventral surface. One assumes venom is being injected but it is not always sufficient to kill P. scaber. The spider may then move a short distance away from the prey for a period of several minutes or less. If struggles continue it adds additional silk. Spider Predation in Porcellio It then sinks its chelicerne through the ventral surface of the proy and remains in this position for 1½ - 3 hours. Details of the feeding process were not followed; presumably enzymes are injected, and suction applied to ingest the organic soup created. The dead carcass is eventually cut from the main sheet of the web. If the packaged prey is small it may be carried from the web periphery to the center of the main sheet for consumption. While walking the spider carries the package with one of its third pair of walking legs, and it is held in a position between the spider and the web. If the prey is larger (and isopods up to five times the size of the spider were effectively preyed upon) then it is eaten in the place of capture. Additional details of predatory behavior were gained from a study of 25 specimens of S. grossa, 2 - 10mm in body length, which were kept in the laboratory for observation and experiment. I attempted to feed these nightly between the hours of 2200 - 0200. P. scaber of weight 20-35mg were introduced either by dropping them into the mainsheet of the web in the vicinity of the spider, or by placing them in peripheral parts, or wholly outside the web. Records of prey taken were kept for 19 S. grossa (fig. 9). Predation is relatively infrequent and the spiders averaged about five days between meals. Some spiders endured at least three weeks without feeding despite the presence of food and there is no reason to believe they could not starve for longer periods. The spiders were observed to have a poor sense of vision. Isopods walking even within 5cm of the spider never elicited a response unless the web was touched indicating that detection of prey relies Spider Predation in Porcellio primarily on vibratory stimulus. The isopods can alert the spider to their presence in several ways. They can fall directly into the web, trip over the scaffolding lines and fall into the web, or become entangled in drop lines or outlying guy lines. Subsequent struggles alert the spider to a stereotyped response. Isopods contacting the web are not necessarily trapped. It was observed that P. scaber can walk upside down either up or down a single thread of the web without much difficulty. The habit of S. grossa of discarding its old prey in a tangled heap below the mainsheet of the web made it easy to determine longterm diets for representative spiders. Accumulated carcasses from ten webs were taken from the garage study site and analyzed for contents. Similar collections were made from the webs of S. grossa in two other locations, an open carport and a closed woodbox from a home on Jack's Point in New Monterey. The results are shown in figs.10 and 11. The same animals were found fairly consistently in the webs, but by far the primary prey of S. grossa as determined from the webs is P. scaber with a frequency of 84%. Moths form the second most commonly taken prey. but only because of two very exceptional webs, occurring at heights of 100 and 170cm above the ground in the main garage study site. Both of these (and no others) were taken from the window ledges; the primary prey found here were moths and other flying insects, undoubtedly attracted to lights shining through the windows. Most of the animals classified under the category "Other" or "Others", figs.10 and 11, occurred in these two webs. They were mostly flying insects 10 Spider Predation in Porcellio such as horse flies, beach flies, dragon flies, etc. but there were also a few earwigs and other spiders. Aside from these two webs, the main prey of S. orossa clearly consists of crawling arthropods, which blunder into the web on foot. I expected to find dietary differences between spiders with webs close to the floor and those high on the walls, but no such differences were found. Webs were very abundant occupying almost every available corner or crevice in the study area, and everywhère except in the window webs, P. scaber is the major prey. Two points are of special interest in this regard. First, P. scaber is probably the main food here because it is most available rather than the most attractive prey. In feeding experiments in the laboratory I concluded that S. grossa does have a relative distaste for P. scaber. On several occasions when an isopod had been refused, a fly was introduced and elicited a complete predatory response resulting in consumption. A second point of interest is that P. scaber seems equally available to spiders on the floor and those near the ceiling. When P. scaber was released on the garage floor they were observed to have no difficulties climbing the garage walls, and on several occasions marked isopods released on the floor were captured as early as four hours Iater in webs halfway up the walls. Climbing is probably part of the normal nocturnal foraging activity of P. scaber. Bristowe (1941) found that specimens in captivity would eat spiders eggs. They also are observed to consume dead and decaying matter inclüding dead members of the species. The largest number found climbing on the walls were observed around 0400 when the humidity was reaching a 11 Spider Predation in Porcellio Maximum. It is reported that P. scaber prefers vertical surfaces where moisture collects but does not become excessive (Heeley, 1941). The spiders are reported to have a six-year lifespan (Kaston, 1953). The sizes of the middens of old prey below webs suggests that they remain in the same web as long as there is a good food supply and they are not disturbed. Should they be disturbed, they readily spin a new one in a suitable location as seen by those collected for the laboratory. The webs are discreet units separated from neighboring webs (fig. 1). Two spiders do not share the same web. When 10 hungry spiders were placed together they were found to be extremely cannibalistic, killing off each other until one remained. It then sat down to consume the dead losers. In view of the large number of annual offspring this activity would function to limit the population size to that suitable for the amount of available prey in the environment. . .* 12 Spider Predation i Porcellio ACKNOVLÉDGMEN I am deeply grateful to Dr. Donald F. Abbott and his wife Dr. Isabella A. Abbott for their guidance, inspiration, and use of Querida del Mar in my research as well as for sustenance during my field work. For their positive influences on my personal growth I will always be indebted. For help in my field studies, special thanks go to Dr. Eugene C. Haderlie for access to his home as a study site and to S. Randall Pratt for assistance with Porcellio scaber. .. Spider Predation in Porcellio TURE CITED LITERA Dristowe, W.S. 1941. The comity of spiders. 2 vols. Ray society, London. Cloudsley-Thompson, J.L. 1958. Spiders, scorpions, centipedes and mites: The ecology and natural history of woodlice, myriapods, and arachnids. Pergamon Press, New York, London, Paris, Los Angeles. 228 pp. Dendry, A. 1895. The cryptozoic fauna of Australasia. Rept. sixth meeting Aust. Assn. Adv. Sci. 6:99-119. Gorvett, H. 1951. The tegumental glands in the land Isopoda. B. The lobed glands: structure and distribution. Quart. J. micr. Sci. 92:275. 1956. Tegumental glands and terrestrial life in woodlice. Proc. Zool. Soc. Lond. 126(2):291-314 Heeley, W. 1941. Observations of life histories of terrestrial Isopods. Proc. Zool. Soc. Lond. 111(B):79-149. Kaston, B. 1953. The. spiders. Wm. C. Brown Co., Dubuque. 289 pp. Kullman, E.J. 1972. The convergent development of orb webs in cribellate and ecribellate spiders. A. Zool. 12(3):395-405 Levi, H.W. 1968. A guide to spiders and their kin. Golden Press, New York. 168 pp. Miller, M.A. 1938. Comparitive ecological studies on the terrestrial isopod crustaceans of the San Francisco Bay region. Univ. Cal. Pub. Zool. 43(7):113-142. 14 Spider Predation in Porcellio IGURE CAPTIONS Fig. 1. Blueprint diagram of the garage study area near Asilomar Beach, Pacific Grove, California, showing exposed studs, horizontal supports, and windows. Shaded black areas represent main sheets of webs; cross hatched areas represent the outlying drag lines; arrows point to locations of spiders in predatory stance. Graph showing number of P. scaber caught Fig. 2. in traps as an indication of relative activity level during a 24 hour period on April 26 to 27, 1973. Fig. 3. Graph showing number of P. scaber caught in traps as an indication of relative activity level on three successive nights, April 30 to May 3, 1973. Lines above show the relative humidity (RH), temperature 5cm above the ground (T), and temperature of the iceplant (Mesembryanthemum) habitat (T,). Graph showing periods during which individual Fig. 4. spiders in the garage wall study area (fig. 1) were active in the characteristic predatory stance over a 24 hour period, May 28-29, 1973. Blackened horizontal bars indicate activity. Fig. 5. Graph showing periods during which the population of spiders in the garage wall study area (fig. 1) were active in the characteristic predatory stance over a 24 hour period, May 28-29, 1973. Blackened vertical bars indicate the number of spiders active in an hourly period. Photograph of S. grossa in process of spinning Fig. 6. eg ocoon 15 redation in Porcellio Spider Steatoda grossa in its predatory Fig. 7. Photograph of stance, hanging inverted from its web of irregular threads. Fig. 8. A) Photograph of web of Steatoda grossa in garage wall study årea (fig. 1) showing ain sheet, scaffolding lines, drop lines, and daytime retreat. B) Photograph - closeup view of same web freshly misted. Fig. 9. Rates of predation in 19 specimens of Steatoda grossa maintained in the laboratory. Fig.10. Tabulated results of examination of discarded prey collected from ten webs in the garage wall study area (fig. 1) and four other webs from Jack's Point for comparison. Numbers indicate number of individuals of prey species. Fig.11. Graph showing percentage composition of prey according to numbers of prey individuals, in ten webs from the garage wall study area (fig. 1). 16 — — — — — — — — — — SEE d A A I 14 A A F S . 5 — 5 a L 0 30 20 10 O . . 00 ME Q aka . . O Fig. o600 0300 2400 2100 1800 O600 O300 2400 2100 1800 0600 20300 12400 22100 1800 RH% 00 5. GaddVUL ON .. RH 80 60 40 1 Z 24 18 12 60 50 40 30 20 10 X V 1 n-76 2 I — 1 TIME Fio 7 ge 5 5 J. . A . . — Fg . . .. Fig 8A ataaa- aakaaa- — g 8E — . O O 0 a O O Q O O O a! — 10 % o m 0) 0 0 1 2 Enclosed garage Open carport Wooabox — — o — 20 50 50 100 130 160 170 210 220 245 305 220 245 10 10 — 30 216 110 11 26 87 79 50 196 201 14 62 5. — 19 O 31 5 O F000 O — O O 4 O O O 23 O 2 — LE — L — — 3 s— - — 00 100 1