POST-IMMOBILIZATION WRAPPING OF PREY BY LYCOSID SPIDERS OF THE HERBACEOUS STRATUM BY

The use of silk or wrapping prey is generally associated with those spiders that construct trapping webs; nevertheless, there are indications in the literature that wandering spiders sometimes use silk for this purpose. In some of the latter cases the silk is applied after the prey has been subdued by biting, a use corresponding to that of the diguetid and linyphiid web-weavers studied by Eberhard (1967). Post-immobilization wrapping in the latter two families was suggested to prevent the prey from falling out of the web during subsequent attacks (ibid.), and in araneid spiders to have the same and additional roles, depending on species and prey size (Robinson et al., I969). The function of this behavior in the non-web-weaving ctenids (Melchers, I963), theraphosids (Eberhard, I967), and lycosids (Rovner, I97I) remained unknown and was the subject of the present study. Our indings suggested that post-immobilization wrapping by wandering spiders serves the same general unction that it probably does in web-weavers--to prevent prey from dropping rom the spider’s elevated location down to the ground whenever it is released rom the chelicerae during (ceding, grooming, or subsequent capture attempts.

rom the spider's elevated location down to the ground whenever it is released rom the chelicerae during (ceding, grooming, or subse- quent capture attempts.

METHODS
We observed individuals of Lycosa rabida Walckenaer (females I2-I9 mm), Lycosa punctulata Hentz (emales I3-I5 rain), and Schizocosa crassipes (Walckenaer) (females--8-IO ram) for the presence of and nature of prey wrapping.XVhereas S. crassipes was collected during spring on forest leaf litter, both species of Lycosa were found in grassy /elds, L. rabida during the summer and L. punctulata during early fall.The spiders were collected in Athens Co., Ohio, USA.
Spiders were housed individually in plastic cages (70 X 125 X 70 mm high) with water available ad lib., and given mealworms (larvae of Tenebrio molitor) for maintenance feedings.Laboratory 1This study was supported in part by National Science Foundation Grant GB 35369 to J. S. Rovner.Manuscript received by the editor September 20, 1974.To determine the functions of the various spinnerets in wrapping, we sealed pairs of spinnerets with paraffin while emale L. rabida were under CO2 anesthesia.Two females underwent sealing of the anterior spinnerets; two had the anterior and median spinnerets sealed; and two, the median and posterior spinnerets.These spiders were subsequently observed and filmed (Bolex Macrozoom Super 8 mm camera, Model 16o) during prey-wrapping.We also exam- ined the silk on the prey and substratum, after chasing away the female from her just-wrapped prey.
To study the preference of L. rabida (and, to some extent, L. #unctulata) for the ground vs. the herbaceous stratum, we con- structed artificial "field habitats" in three terraria (o.2 X o.4 >( o.2 m high).Cardboard "foliage" was fixed in a plaster base which provided the "ground surface" on of the terrarium bottom.The remainder of the glass bottom was left uncovered to provide an alternative ground surface in case the piaster had a repellent effect on the spiders (which turned out not to be the case).Vertical and sloping surfaces projecting into space were provided by the artificial foliage (Fig. I).Spiders could climb to a height (limited by a glass lid) of o. 19 m.
Three female L. rabida (tagged with non-toxic enamel) were housed in each terrarium.(Three individuals were cannibalized in the course of this experiment and were replaced with equivalently tagged substitutes.)To prevent possible conditioning effects, we fed the spiders on the ground as often as on the foliage.(Feeding was done at times other than the daily observation periods.)We observed the nine spiders for 2-hr blocks of time each day for o days during the afternoon (on most days, I3OO-15oo), and recorded the duration of time each spider spent on the ground vs. the foliage.At -hr intervals we also measured the height of those spiders that were on the foliage.After completing this phase of our study, we observed and filmed prey capture and wrapping by spiders on the artificial foliage.Analysis of these and other films was aided by use of a Bell & Howell Super 8 mm Multi-motion projector.
Field observations were made on L. rabida during daylight hours (-usually 3oo-I 5oo), and during both day I3OO-I 5oo) and night (2o3o-22oo) conditions on L. I)unctulata.The spiders were detected at a distance and their location within the habitat noted.By ap- proaching carefully, we were able to film undisturbed L. rabida in the field.

PREY-WRAPPING BEHAVIOR
Whereas individuals of S. crassipes did not use silk after prey capture, those of L. rabida and, to a lesser extent, L. punctulata did so.In the latter two species, silk was not used to.immobilize the prey, since wrapping began after the prey had ceased most of its struggling, usually several min after capture (Table ).

1974]
Rovner q Knost--H/'ra#loing of Prey 4Ol Table I.Bouts of prey-wrapping by female Lycosa rabida (*) and Lycosa punctulata (f) when offered multiple prey during a 30-min period.A total of 12 spiders was tested on each type of prey.(Sample When beginning a bout of wrapping, the spider spread and wiggled its spinnerets, flexed its abdomen ventrad, and then pressed its anterior spinnerets to the substratum.(Sometimes additional attachment disks were placed on the substratum in this manner within an arc .of60 or less.)Then the spider began to pivot above the prey.
In about 0% of the cases the spider pivoted revolution (or sometimes revolution), fixed an attachment disk to the substratum, and then began to turn in the opposite direction.However, in most cases the spider pivoted in only one direction, while remaining cen- tered above the prey.During the first revolution, the silk was attached to the substratum at 2-5 points, the number perhaps de- pending on the prey's bulk.Fewer attachment disks were placed during the next turn or two, and usually only one or none during later revolutions.
During the first two or three turns, the spider usually held the prey in its chelicerae (Fig. 2).Consequently, as the spider revolved, the prey animals beneath it pivoted around with the spider.(Very large animals, while not held in the chelicerae during wrapping, and therefore not having their weight supported by the spider, were contacted by the palps, chelicerae, and sternum of the spider pivoting above.)If, as was the case with most prey, the spider held the prey in its chelicerae during early revolutions, it then released the prey during a subsequent turn and continued to pivot for one or more revolutions (Fig. 3).Throughout the process, the palps were used to contact or manipulate the prey.In L. rabida 8o% of the wrapping bouts involved four or fewer revolutions (mean--3.3+/-2.tSD; range -4; n 06).Some views of one such bout are sketched in Fig. 4.
All three pairs of spinnerets were used during prey-wrapping.
Data obtained from experimental animals indicated the same roles Psyche [September-December Fig. 2. Female Lycosa rabida placing an attachment disk on the sub- stratum during the early phasz of prey-wrapping.The multiple prey are held in her ehelieerae and palps.Fig. 3. Female Lycosa rabida during the later phase of prey-wrapping.The prey have just been released from her cheliceral grasp.(Her palps continue to contact the prey.)Swathing silk issues from her posterior spinnerets.

1974]
Rovner & Knost [/[/'rapping of Prey 403 for the spinnerets in lycosids as in other spiders that wrap prey.When the anterior spinnerets were sealed, no attachment disks were produced.The prey was wrapped but not fixed to the substratum.The same was true when the anterior and median spinnerets were sealed.When the median and posterior spinnerets were sealed, attachment disks were fixed to the substratum and dragline laid around the prey, but no swathing silk.In this case wrapping was not effective.Multiple prey were not tied together; indeed, the prey items were often pushed apart by the legs of the pivoting spider.
As the result of normal wrapping behavior, silk was placed around the bodies of the prey animals and attached to. the substratum at intervals.Although the pivoting spider had pulled the threads taut, the silk was not dense enough to produce a very tightly wrapped package, as is produced by orb-weavers, for example.Indeed the wrapping was so sparse that it would be unnoticed by the casual observer.Nevertheless, the prey group did form a more compact mass than it had prior to wrapping.When the spider resumed feed- ing, it lifted the prey away from the substratum a. short distance (rather than lean down to feed) however, the lines running to the substratum generally remained intact due to their elasticity.
After prey-wrapping, the spider sometimes groomed its chelicerae and palps, while remaining directly above the prey, and then resumed feeding.Additional prey that approached within reach were often captured; and one or more bouts of wrapping followed.However, regardless of whether further captures were made after the first bout of wrapping, subsequent bouts on the previously wrapped prey occurred later in the observation period in many cases.
Table II.Number of Lycosa rabida making single vs. multiple captures when offered multiple prey during a 30-min period (second figure) and the number o.f those captures in which wrapping occurred (first figure).Twelve spiders o.f each sex were tested on each type of prey.Rovner Knost--fralling of Prey 405 L. rabida wrapped various types of prey, especially in situations in which multiple prey were captured (Table II).While only I9% of the males captured multiple prey, 68% of the females did so.

Males
Considering both sexes together, wrapping occurred in 2o% of the cases in which only single captures were made, and in 95% of the cases in which two or more prey were captured.Such correlations meant that females wrapped prey much more often than males, performing at least one bout of wrapping in 78% of the observation periods in which they captured prey (single or multiple), while males did so only 23% of the time.None of our spider species wrapped fruit flies.
As shown in Table I, individuals of L. I)unctulata wrapped prey much less often than L. rabida and did so less intensively (fewer turns and shorter bout duration).Furthermore, it was only four of the twelve females that did so; none of the twelve males wrapped prey.

STRATUM SELECTION
Field observations.
We watched penultimate and adult L. rabida on sunny, calm days (early July) in a grassy field near the edge of a woods.Undisturbed individuals rested motionless on the leaves and stems of grasses and other vegetation.The spiders typically were on horizontal or sloping surfaces (rather than vertical ones).Those individuals we happened to observe were on the upper leaves of relatively low vegetation or part-way up the stems and leaves of taller vegetation.
In response to mild disturbance, the spiders walked or climbed a short distance and then resumed their motionless stance.Strong disturbance (as that which resulted from our walking near them during our collecting efforts) caused the spiders to run and climb .Sketches, based on a Super 8 mm film, showing dorsal views of prey-wrapping by a female Lycosa rabida.The several items of the multiple prey-capture are drawn as an homogenous mass (stippled) for convenience, and should not be regarded as a wrapped bundle.Lighting conditions permitted seeing the silk in views B, E, and G. Times were rounded to the nearest see. A. Placing 1st attachment disk (Time--'0).B. One-quarter into the 1st revolution (+ 4 see).C. Near end of 1st revolution (+ 7 see).D. One-third into 2nd revolution (+ 9 see).E. Near end of 2nd revolution (+ 12 see).Prey were released from the ehelieerae at the beginning of the next revolution.F. One-third into 3rd revolution (+ 14 see).G. Beginning of 4th revolution (+ 18 see).

Psyche
[September-December swiftly through the foliage, sometimes breaking into a series of leaps and dashes that carried them long distances along the top of lowgrowing vegetation in a few seconds.Thus, these animals usually remained in the herbaceous stratum; only occasionally did an in- dividual plunge downward and hide near the ground.
Penultimate L. lunctulata were seen in the fall in the same fields, and typically traveled through the herbaceous stratum when dis- turbed.At night these spiders were seen resting on or moving slowly upon the vegetation or, much less frequently, upon the ground.
Laboratory st.udies.The results of our study of diurnal stratum selection by undisturbed female L. rabida in an artificial habitat are summarized in Fig. 5. (Temperature was 25-26 and relative hu- midity > 6o% throughout the o days.)During the 2o-hr survey, the spiders were on the cardboard foliage 73% of the time.Most of that time the animals rested motionless on one of the "leaves" at an average height of o. m, i.e., more than half-way up the "plant." The long axis of their bodies was usually in a plane ranging from nearly horizontal to a slope of about 6o from the horizontal.In the latter situation the spider faced up-or downward.In those cases in which the spider rested vertically on a "stem" the spider Rovner & Knost--Wrapping of Prey 4o7 typically aced downward.Individuals rested more oten in an upright position on the upper surface o the lea (illumination- approx.4oo lux) than they did in an inverted position beneath the lea (illumination approx.I3o lux).
We considered each spider's behavior within the 2-hr periods o observation and noted that each remained within the "herbaceous stratum" during the entire 2 hr in 5.3 (range---4-7) o the ten periods.Each spider remained on the ground throughout the 2-hr period in only .6 (range--=-0-3) of the ten periods.The spider moved rom the ground to, the oliage or vice versa in the remaining 3.I (range ---2-4) o the ten periods.
Some observations made on several emale L. punctulata in the artificial habitat indicated that these spiders also tended to spend their time resting on the oliage.However, we did not quantiy this part of our study.
Ater completing the o-day survey o stratum selection by L. rabida, we observed and filmed prey-wrapping by these spiders on the artificial oliage.The behavior resembled that performed on the floors o( the housing containers, as we described above.Wrapping Fig. 6.Female Lycosa rabida wrapping prey on artificial foliage.This leaf sloped at an angle of about 60 to the horizontal.Psyche September-Decembe usually was performed at the capture site and therefore was carried out in all planes, including the vertical.The spiders behaved on sloping leaves and vertical stems in a manner similar to that seen on the horizontal cage floors; however, pivoting was slower on the narrow surfaces, apparently due to the spider's having to seek foot- holds as it revolved "above" the prey (Fig. 6).
In some cases, we intentionally disturbed the spider before or after prey-wrapping.In the former situation, the prey sometimes was released from the chelicerae and dropped to a lower leaf or to the ground.On the other hand, if wrapping already had occurred, they prey did not fall from the capture site when released from the chelicerae, even on a vertical surface.When multiple prey were wrapped on a sloping surface, they: () remained together rather than roll apart and (2) hung at the wrapping site rather than drop to the ground, when the spider was chased away.
A further bit of evidence for the a.daptedness of L. rabida to the herbaceous stratum was provided later by the construction of a "shelter web" (about 50 mm) and, within it, an egg sac (infertile) by a female on a nearly horizontal leaf o.3 m above the ground in the artificial habitat.

DISCUSSION
Preference for herbaceous stratum.--8,crassiDes is well-known as an inhabitant of woodland leaf-litter (Kaston, I948; Fitch, I963)   indeed, lycosids are generally regarded as ground-dwellers (Lowrie,   968).However, L. rabida and L. punctulata can be collected by sweeping grass and shrubs (Kaston, I948).Kuenzler (958) also found L. rabida on the lower trunk and branches of trees in open woodland.Whitcomb et al. (I963) describe finding L. rabida in cotton fields on the ground during the day, but at night "... half- way up the cotton stalks feeding on bollworm or cabbage looper moths ."as well as on noctuid moths.At night, Eason and Whitcomb (I965) collected 80% of their specimens of L. rabida 2/3 rn or more above the ground in tall grass and bushes.
According to Eason and Whitcomb (965), "'L.punctulata is mostly captured on or near the ground, since it has less tendency to climb into bushes than does L. rabida.'"While finding both species in grassland, Fitch (963) reported that L. punctulata preferred "... relatively open or barren situations as compared to L. rabida, which prefers a tall-grass habitat."Kaston (948) stated simply that both species have the same habits.Our field and laboratory observations are in agreement with the general idea expressed above uL.rabida and L. punctulata are adapted for spending a significant part of their life in the herbaceous stratum, rather than being, like most lycosids, ground-dwellers.(Obviously, stratum selection may vary during the life of an in- dividual under different conditions, both external (temperature, humidity, wind, etc.) and internal (hunger, sexual state, etc.)).A difference between the species in regard to habitat or stratum selec- tion, suggested by some authors, was not examined by us, but might be revealed by quantitative field and laboratory studies.The rela- tively longer legs of L. rabida suggests a greater adaptation to the herbaceous stratum than that of L. punctulata, as does the stronger tendency of L. rabida to wrap prey.
While these lycosids are often seen running about, perhaps mi- grating from one habitat to another in response to micro-climatic conditions, or perhaps merely reacting to the collector's movements, most of their time is probably spent at rest on the foliage.This seemed to be the case in the field, and was certainly the situation in the artificial habitat.As has been noted elsewhere (Edgar,t969), lycosid spiders generally remain motionless and wait for their prey to come within reach.Like the web-weaving spiders, the wandering spiders of the herbaceous stratum are resting on a medium that pro- vides effective transmission of vibrational cues from the prey.Function of prey-wrapping.--Wehypothesize that post-immo- bilization prey-wrapping in L. rabida and L. punctulata is a be- havioral adaptation for life in the herbaceous stratum.It reduces the possibility of losing the prey items when the spider releases them from its cheliceral grasp between bouts of feeding.The spider may release its grasp to make an additional capture, to groom, to drink, or to locate another point on the prey's body for further feeding.Also, when startled by another animal or by shaking of the sub- stratum due to wind, the spider's momentary release of the prey would often result in its loss.No such problem exists for ground- dwelling forms.
Our hypothesis is supported by the absence of wrapping after the capture of small prey, which are easily held in the chelicerae and ingested in a relatively short time.It appears, then, that post-immo- bilization wrapping serves the same primary function in wandering spiders as it does in web-weaversto prevent the prey from dropping to the ground from the spider's elevated location.Also, it is im- portant when multiple prey are captured by a wandering spider, Psyche September-December since the number o prey that the spider can hold in its chelicerae during eeding is limited.
As to the unctions o the components o] prey-wrapping, we sug- gest the ollowing.The swathing silk serves to enclose the prey, a role that is not only important or holding multiple prey together, but which is also o value in holding together the pieces that oten result rom the process o ingesting a single prey item.This latter unction may be one reason or the occurrence o additional bouts o wrapping during the course o eeding.The dragline may be entangled in the swathing band and serve to attach the prey package to the substratum via the attachment disks, although it seemed that the swathing band itsel was connected to the attachment disks.
Since attachment to the substratum seems to be o prime value in preventing loss o acquired ood material, the re-anchoring that occurs in additional wrapping bouts during the course o eeding probably makes up or the breakage o lines that accompanies the physical and chemical processing o the prey by the eeding spider, and thus may be another reason or extra bouts o wrapping.
The temporal pattern is one o providing multiple attachment points, along with swathing, during the early phase o each bout (while the prey is held in the chelicerae) and then devoting most o the remaining time to swathing (ater the prey has been released ]rom the chelicerae).The early phase (attachment) may be im- portant or effective wrapping, since it enables the spider to release the prey and pivot around it, pulling the svathing lines taut.
Role of aciniform and other silk glands.Comsto.ck 92)and others have described the aciniform glands o web-weaving spiders as the source o the swathing silk used or prey-wrapping, as well as the fine fibers o egg sacs.These glands are present in lycosids, but their function had remained unknown (Richter and Van der Kraan, 97o).On the basis o histological data, Richter (97oa) suggested roles for the acinifo.rmglands in lycosid spiders in producing silk involved in molting, egg sac construction, and sperm web construction.
Our data indicate that the aciniorm glands also have one o the same unctions in lycosids as they do in the web-weavers produc- tion o swathing silk or wrapping prey.When the spigots or these glands were closed experimentally (by sealing the median and posterior spinnerets), wrapping behavior was ineffective.Such spiders pivoted, fixed attachment disks, and laid a dragline; however, no broad bands o silk were placed on the prey.Indeed, the several items o multiple prey were oten pushed apart by the legs o the pivoting spider.Our sealing of the anterior spinnerets (site ot: spigots ior the piriform glands) prevented attachment disk production, as was well- known in the literature.Such spiders could wrap the prey but not attach it to the substratum.This was also the case with spiders that had their anterior and median spinnerets sealed, with the further limitation of being incapable of producing draglines.In summary, for complete wrapping of the prey, these lycosids are using all three pairs o spinnerets and employing silk from three types of silk glands (the third being the ampullaceal glands or production of drag- lines---whose spigots are on the anterior and median spinnerets).
This agrees with observations made by Melchers (96) on prey- wrapping by normal individuals of Cupiennius salei, a large ctenid spider.
Since the wrapping of prey by lycosids (and at least some other wandering spiders) probably correlates with a preference for the herbaceous (or higher) stratum, we hypothesize that the size of the aciniform glands, source of swathing silk, would also.This expecta- tion is based on Richter's (97ob) demonstration of such a corre- lation between ampullaceal gland size (dragline silk) and preferred vegetation structure in Pardosa spp.Indeed, we should expect lyco- sids with larger ampullaceal glands to have larger aciniform glands, based on our hypothesis and on Richter's findings.
Stimuli releasing prey-wrapping.The correlation between the frequencies of prey-wrapping and multiple prey-capture suggests that the latter event provides a stimulus situation to release prey-wrap- ping.The value of wrapping for retaining multiple prey has been discussed above.However, wrapping certainly occurs after single prey are captured, if these prey are about as large or larger than the spider's body size.It seems, then, that the stimulus for post-immo- bilization wrapping is that of a relatively large volume or mass of prey material beneath the spider, not the absolute number of prey captured.We suspect volume rather than weight to be the stimulus, since very large prey are wrapped without being held in the cheli- cerae following a capture in which the bite was maintained without the spider supporting the weight of the prey resting on the sub- stratum.
Prey-wrapping was initiated after the prey had ceased most or all of its struggling.Groups of small crickets or grasshoppers, each member of which succumbed quickly to the spider's bite, were wrapped within a minute or two after capture.Mealworms, which performed active twisting of the free end of their body for a long time, were not wrapped until about o min after capture.Very Psyche [September-December large, single prey were wrapped only after a lengthy period, hr or more, had elapsed.It appears that the absence oi: vigorous movements by the prey is a necessary part of the stimulus situation for the onset of wrapping.
Post-immobilization wrapping in other wandering spiders.Our two species of Lycosa performed post-immobilization wrapping, whereas Schizocosa crassipes did not.We modified our original speculation that this behavior might be characteristic of the Lycosidae to the notion that it might be at least genus-specific.Even this is no longer tenable, in light oi: our preliminary observations on other species as well as data obtained by R. J. McKay (personal commu-  nication).We now regard post-immobilization prey-wrapping as a behavioral adaptation that may be included in the repertoire of various unrelated wandering spiders that have specialized for spend- ing part or most of their time some distance above the ground on vegetation.
Post-immobilization wrapping occurs after the prey has stopped struggling, occupies a. relatively brief amount o time during the lengthy feeding process, and is used for prey only when the total volume exceeds some threshold value.These are the probable rea- sons why this behavior has not yet been reported in many species in which it is likely to occur.Authors specifically studying feeding behavior have seen post-immobilization wrapping in two other fam- ilies of wandering spiders--ctenids (Melchers, 96, 963) and theraphosids (Eberhard,967), although neither of these authors determined the function.
A personal communication from E.-A.Seyfarth, that wrapping occurred in Cupiennius salei, called our attention to the work of Melchers 96i, 963) on this ctenid species.(Comparison with our species was facilitated by our being able to view Melchers' (96) film of this behavior.)Post-immobilization wrapping by C. salei is very similar to what occurs in our Lycosa spp. as to the sequence of events and movements, stimuli releasing the behavior, etc.However, two differences are obvious: () C. galei waits for prey on a vertical surface.If it tumbles to the ground during prey capture, it carries the prey back up to an elevated point on the vertical surface the glass wall o the terrarium in Melchers' study.Individuals o,f our species never did the latter, but readily stayed on the ground to feed if the capture ended there.(2) The density and breadth of the swathing bands are much greater in C. salei than in our Lycosa spp.
Points and 2 are probably interrelated and suggest that this ctenid Rovner & Knost--trapping of Prey 4r3 spider is even more specialized for living on vegetation than are our Lycosa spp.
Scopula hairs, which enable spiders to climb smooth surfaces, have been studied in lycosids, theraphosids, and other wandering spiders 974).Since these structures are an im- portant adaptation (or many of the spiders which spend their lie on vegetation, they may serve as a morphological indicator of behavioral correlates o this li(e style, including post-immobilization prey-wrap- ping.Indeed, a scopula is present in those species in which this behavior has been seen.On the other hand, since scopula hairs on the tarsus and metatarsus are probably important for the capture and manipulation of prey (Rovner and Knost, in prepa,ra.tion),their presence may not always be indicative of habitat preference and related behaviors.
Experimentally modified male L. rabida that are unable to achieve palpal insertions "tie-down" emales during attempted copulation (Rovner, 97), a behavior resembling post-immobilization wrapping of large prey.Male thomisid spiders perform a similar behavior as part of their normal pre-copulatory activity.(In neither case does the wrapping have any restraining effect on the emale, who easily tears the threads and departs when the copulation is completed.)Perhaps the pre-copulatory wrapping in thomisids is a ritualized behavior derived from a present or past use of silk for wrapping prey in this family.Eberhard (967) associated post-immobilization wrapping with certain primitive types of aerial web-building, similar to that ex- hibited by the present-day diguetid spiders.The behavior shown by our Lycosa spp. of the herbaceous stratum, and by other wander- ing spiders, suggests that the evolution of post-immobilization wrap- ping may not be as closely associated with the building o aerial webs as Eberhard had hypothesized.

SUMMARY
Post-immobilization wrapping o large, single prey and groups of smaller prey occurs in Lycosa rabida and, to a lesser extent, Lycosa punctulata.Observation of these spiders in both the natural habitat and an artificial habitat indicated their preference for the herbaceous stratum.Post-immobilization wrapping is probably an adaptation for life in this stratum, since it reduces the possibility of losing prey when it is released from the cheliceral grasp.The swathing bands, at- tachment disks, and draglines involved in wrapping are produced Fig. 1.Female Lycosa rabida (with prey) on artificial foliage.

Fig. 4 .
Fig. 4. [Opposite page].Sketches, based on a Super 8 mm film, showing dorsal views of prey-wrapping by a female Lycosa rabida.The Stratum selection by female Lycosa rabida in an artificial habitat.During 20 hr, spread over 10 days, we recorded the proportion o time spent on cardboard oliage (solid bar) vs. on the ground.At halt- hr intervals within observation periods, we also measured the height o those spiders that were on the oliage(hatched bar).1974]