Ian C. Fergusson.
Natural History of the Spider, Hypochilus thorelli Marx (Hypochilidae).
Psyche 79:179-199, 1972.

Full text (searchable PDF, 4940K)
Durable link: http://psyche.entclub.org/79/79-179.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

NATURAL HISTORY OF THE SPIDER
HYPOCHJLUS THORELLI MARX (HYPOCHILIDAE)l BY IAN C. FERGUSSON
Department of Biology,
Western Carolina University, Cullowhee, N. C.' INTRODUCTION
Hypochilus is regarded as the most primitive of the three "hypo- chilomorph" spider families: the Hypochilidae, Austrochilidae (= Thaididae) and Hickmaniidae (Marples, 1968). These small, close- ly related families exhibit character states resembling both those of the primitive mygalomorph spiders (suborder Orthognatha) and the araneomorph spiders (suborder Labidognatha). They are consid- ered to be the most primitive araneomorphs (Gertsch, 1958; Mar- pies, 1968). The hypochilomorph families are geographic as well as phylogenetic relicts, being represented by limited populations in such widely-separated areas as the United States (Hypochilidae), China ( Hypochilidae) , Chile (Austrochilidae) , and Tasmania (Hick- maniidae).
Clearly, a better understanding of spider evolution may be gained from studies of hypochilomorph biology (Marples, 1968; Shear, 1970).
The four known species of Hypochilut occur only in the United States. All are allopatric and are found only in mountainous re- gions.
Most works on Hypochilus have used only morphological char- acters to determine the relationships of the Hypochilidae to other spider families.
Though mention has been made of web structure (Comstock, 1940; Gertsch, 1958; Hoffman, 1963), only two in- vestigators (Kraus, 1965, and Shear, 1970, with Hypochilus gert- schi) have published additional behavioral data. Life history data on Hypochilus are virtually non-existent. The present study was undertaken to obtain information on the natural history (ecology, life history, and behavior) of Hypochilus thorelli. It is hoped that these data will contribute to an under- standing of the evolutionary history of Hypochilus and the Hypo- 'From a thesis presented to the graduate faculty of Western Carolina University in partial fulfillment of the requirements for the degree of Master of Arts in Biology.
'Present address: 524 Benton Street, Missoula, Montana 59801 Manuscript received by the editor October 15,1972



================================================================================

I 80 Psyche [September
chilidae, and, in addition, help explain the success (abundance) of this relict species.
METHODS
Most of the results reported here are based on field observations made from July 1970 to June 197 I. For some behavioral observa- tions, spiders were kept in cardboard boxes in a humid basement room. The open sides of the boxes were covered with transparent plastic wrap. A headlamp covered with a red plastic "filter" was used for observing behavior at night.
I collected H. thorelZi in the following localities in North Caro- lina: Buncombe County- 3 mi S of Dillingham, Pisgah National Forest (females). Graham County - Joyce Kilmer Memorial Forest (egg cases). Haywood "County - Sherwood Wildlife Man- agement Area, 13 mi SE of Waynesville (immatures). Jackson County - Cullowhee Mountain, 8 mi S of Cullowhee (immatures, egg cases) ; Highway 107, 2 mi N of Glenville (immatures, egg cases) ; Little Mill Creek, 5 mi S of Tuckaseigee (males, females) ; Mull Creek, Caney Fork Game Area (males, females, irnmatures, egg cases) ; Tanassee Creek Reservoir, 12.5 mi SE of Cullowhee immatures) ; Wayehutta Creek, 2 mi NE of Cullowhee (male, egg cases) ; Whitewater Falls, 8 mi E of Highlands (immatures, egg cases) ; Wolf Creek, 5 mi S of Cullowhee (males, females, imma- tures, egg cases). Macon County - Dry Falls and Cullasaja River gorge, 3 mi NW of Highlands (males, females, immatures, egg cases). Swain County - Clingman's Dome, Great Smoky Moun- tains National Park (male) ; Kephart Prong Trail, GSMNP (egg cases) ; Nantahala Gorge, 12 mi SW of Bryson City (females, egg cases). Yancey County- "Lost Cove", 30 mi NE of Asheville (egg cases). H. thorelli was also observed but not collected in Cherokee and Henderson counties, North Carolina and in Blount and Sevier counties, Tennessee. J. Dan Pittillo collected a mature male at Yonah Mountain, White County, Georgia. Repeated observations were made in the following localities where large populations of 1H. thorelli were located: in Jackson County at Mull Creek, elev. 975 m (3200 ft), and Wolf Creek, elev. 730 m (2400 ft) ; and in Macon County at Dry Falls, elev. 1000 m (3300 ft)
NATURAL HISTORY
Geographic Distribution
Hypochilus thorelli occurs at elevations of about 300 m ( 1000 ft)



================================================================================

19721 Fergusson - Hypochilus thorelli I 8 I and above in the Appalachian Mountains of North Carolina, Ten- nessee, Georgia, and Alabama, and in the Cumberland Plateau of Tennessee and Kentucky. Kaston (1948) repo,rted that Petrunke- vitch introduced H. thorelli into Meshomasic State Forest in Con- necticut in 1932, and that they were "holding their own" in 1937. Petrunkevitch ( I 932) gave the Appalachian distribution as a triangle with apices at Blowing Rock, North Carolina, Tallulah Falls, Georgia, and Maqville, Tennessee. The line from Tallulah Falls northwest to Maryville has been extended farther into Georgia and Alabama. The Appalachian populations are separated from the Cumberland populations by the broad valley of the Tennessee River. While the distribution of H. thorelli in the Appalachians is appar- ently limited to the east, south, and west by low elevations, no such geographic barrier is apparent that would limit its northward dis- tribution, or the southward distribution of H. gertschi, known from West Virginia and northwestern Virginia. Yet Petrunkevitch ( 1932) and Hoffman ( 1963) searched in favorable habitats in the "isthmus" of western Virginia, and were unable to find any pop- ulations.
Habitat
The optimum habitat for H. thorelll in North Carolina is in stream gorges in humid deciduous forests at elevations of 600 to 1100 m (2000 to 3600 ft). Webs are built predon~inantly on sheltered rock ledges. The undersurfaces of stream bridges and logs overhanging or close to stream banks are also occasionally inhabited. The availability of web substrate is a factor which limits local distribution. Stable overhanging or vertical surfaces with a number of projections or nearby plants for guy line attachment are necessary. Humidity is apparently the most important factor limiting local distribution. <H. thorelli webs are more abundant close to a stream than farther away. Well-shaded ledges are more likely to be inhabited than poorly-shaded ones. Where the ledges are well-shaded, the webs are built at any place providing good attachment points. On poorly- shaded ledges webs are built close to the ground, usually behind low vegetation where increased shade and plant transpiration keep the humidity higher.
Marginal habitats located included an unshaded roadside ledge IOO m across cleared land from the nearest stream (Graham County) and a ledge in dense spruce-fir forest on the side of Clingrnan's Dome at an elevation of 1891 m (6200 ft).
Hypochilus was observed from 397 to 1891 m (1300 to 6200 ft), but is rare above 1370 m (4500 f t) .




================================================================================

182 Psyche [September
Figure 1.
Generalized web
structure of H. thorelli. S.S., substrate silk; L., lampshade; S.L., support line; F.L., frame line; T., tangle. Drawn from photos.
The densest population observed was at Mull Creek, Jackson County, where 59 individuals were counted one day in June on a rock ledge approximately I 5 m long and I to 3 m high. Petrunkevitch ( I 932) and Hoffman ( I 963) noted that 1H. thorelli occurs in noticeably drier situations in the Cumberlands than in the Appalachians. J. Beatty (pers. comm.) says that the Alabama popu- lations of H. thorelli occupy moist sites when available, and he does not consider them abundant in drier situations. Perhaps, as he sug- gests, the Cumberland populations have a wider range of moisture tolerance than do the Appalachian populations. Hoffman ( I 963) and Shear ( I 970) noted that H. g-ertschi prefers drier situations than does H. fhorelli, even though moist sites are available. Prey Capture
Web construction.
The web of Hypochilus is usually described ( Comstock, I 940 ; Gertsch, 1958 ; Shear, 1970) as a "lampshade- shaped" mesh. The narrower inner end of the lampshade is at-



================================================================================

19721 . Fergusson - Hypochilus thorelli 183 tached to a circular sheet of silk laid down on the substrate. The sides of the lampshade extend away from the substrate, flaring to a larger mouth. An extensive tangle of lines connects the lampshade to the surrounding substrate and serves to keep the lampshade taut. Sticky silk is found only in the lampshade and support lines (Shear, 1970)
The web of H. thorelli (Figures 1-3) is similar to that of H. gertschi, as described by Shear
( 1970) . This includes (following
Shear's terminology) ) the meshed lampshade, with 10- I 5 paired support lines attaching it to' a frame line, which is guyed to the sub- strate and to the extensive angle of threads below the lampshade. H. thorelli rests with its ventral surface appressed to the substrate, with all legs touching the lampshade at its base (Figure 4). This is somewhat different from the ~osition of H. gertschi, which holds on to the lampshade with its first two pairs of legs, and to the circular sheet with its last two pairs ( Shear, 1970). I measured lampshade base diameter, mouth diameter, and depth of 10 H. thorelli webs.
The range of values obtained was: base
diameter 2.2-7.4 cm, mouth diameter 3.3-14.2 cm, depth 1.5-9.5 cm. I also measured base diameter of 134 webs in an attempt to de- termine age classes (see Life History below). Cornstock (1940) observed web construction behavior in H. thorelli. He found that the disc of silk was first laid down on the substrate, then the foundations of the lampshade were constructed, and finally the lampshade itself was completed. The only construc- tion activity I
observed was web repair after prey-capture. The spider combs its cribellum silk with the calamistra on the last pair of legs, while applying silk to the damaged area. The edges of the torn area of the lampshade are pulled together with the other legs during this process. The spider often returns to the web base, pulls on the sides of the lampshade to test its tension, and then resumes work if the repair is not "satisfactory." Extensive web repair, and probably also web construction, takes place at dusk and night; it was not observed during daylight.
Predatory behavior.
H. thorelli does not respond to prey in the tangle below the lampshade unless the impact is violent, in which case it reacts as it does to a threat (see Defensive ~ehavior). A prey struggling in the sticky lampshade causes the spider, to test web tension by flexing its legs and tightening the web in order to locate the prey. (One very large and vigorous insect- an ichneumonid wasp - caused a defensive reaction when it was caught in the lamp- shade.)
The spider than orients so that it faces the prey. Usually



================================================================================

Psyche [Septexber
Figure 2, Web of H. Ihorr!11, lateral view. Sotc position of spider against substrate. Photo by L. Tucker.
Figure 3.
Web of H. thwelli seen from below. Note tangle and flaring of lampshade. The spider is at the center of the web base.



================================================================================

Figure +.
H. fhordi in web. Note cryptic coloration and the legs grasp- ing the lampshade. Photo by L. Tucker.




================================================================================

186 Psyche [September
it anchors its last pair of legs on the silk base and uses the other legs to grasp the silk near the prey and pull it toward its chelicerae. The struggling of the prey, combined with the pulling of the spider, often thoroughly entangles the prey. When it has pulled the prey close, the spider lunges forward and bites it repeatedly, frequently touching the prey with its pedipalps. Occasionally the spider may lunge at the prey without first pulling on it. It may also bite the prey and hold on for 2-3 minutes before releasing it. After the prey is dead (2-10 minutes), the spider usually cuts it out of the web with its chelicerae, carries it back to the web base, and assumes its usual resting position. It then feeds, holding the prey only with its chelicerae. After feeding, the spider simply drops the prey remains out of the web.
One prey, an immature Homoptera, was rejected after being killed. The spider cut it out of the web and allowed it to drop out. On no occasion did I observe H. thorel'li wrap prey or otherwise use silk to subdue prey, though I did observe one peculiar use of silk. A Hypochilus attacked and killed a small gnaphosid spider, then cut it out of the web. It placed the smaller spider against its spinnerets, where it was held by silk. The spider then climbed back to the web base and pressed the gnaphosid against the base sheet, where it remained attached. The spider then went back to the at- tack site and began repairing the web.
Prey. I collected 40 different prey remains from H. thorelli webs. Insect families represented were Tipulidae (8), Formicidae (4), Gryllacrididae (2), Ptilodact~lidae, Cerarnbycidae, Lampyridae, and Cicadellidae (one each). Nine specimens could only be identi- fied to order: Diptera (6), Coleoptera, Trichoptera, Lepidoptera (one each). Other arthropods included g opilionids and 4 spiders (one gnaphosid, one lycosid, one Hypochilus, and one Antrodiaetus.) The most common prey, opilionids (daddy-longlegs) and tipulids (crane flies), are extremely abundant in the situations in which H. thorelli constructs its webs. Ants (Formicidae) are also common on rock faces. Cave crickets (Gryllacrididae) are also abundant, but are apparently usually able to avoid PI. thorelli webs. At Wolf Creek, I observed a number of cave crickets leaving a deep crevice, at the mouth o,f which was a large HypochiZus web. The long an- tennae of the cave crickets enabled them to detect the web before they could become entangled.
The single case of cannibalism occurred with captive specimens. One female left her box compartment before building a web and wandered into the adjacent compartment co,ntaining a female that



================================================================================

19721 Fergusson - Hypochilus t horelli 187 had already built a web. The remains of the first female were di- rectly under the other's web, but there was no evidence of damage to the web, so I do not know if the first female became entangled or was attacked on the guy lines.
Web function. The circular sheet of silk laid down on the sub- strate serves as an attachment point for the lampshade, and provides an anchor point for the spider's last pair of legs during prey capture. The lampshade, containing sticky cribellate silk, is the actual prey trap. In addition, the lampshade serves as a protective retreat for the spider.
The support and frame lines in H. thorelli webs function chiefly in support of the lampshade. They are too far apart to form an efficient prey trap, but may help deflect flying insects into the lamp- shade.
These lines also serve to signal the approach of a potential predator (see Defensive Behavior below). The tangle of threads serves to support and anchor the lampshade. A secondary function of the tangle is to warn of potential predators. Although prey do become briefly caught in the tangle, H. thorelli does not attack them there.
The web of ~Hypochilus is not as efficient as a two-dimensional orb web (in terms of effective area covered per quantity of silk), but it has the advantage over such strictly aerial or other strictly ground webs of being able to capture both crawling and flying arthropods. The two most abundant prey (see above) were tipulids (flying arthropods) and opilionids (crawling ones). Crawling arthropods - ants, daddy-longlegs, cave crickets, and spiders - comprised 47.5 % of the 40 prey remains I collected. These ground arthropods prob- ably encounter the base of the lampshade while crawling on the rock surface.
The flaring of the lampshade enables it to capture these prey when they struggle and fall after contacting the lampshade base.
Thus the lampshade is able to intercept movements that are either perpendicular to the substrate or parallel to it. Since it builds an aerial web and does not use silk either to over- come or secure prey, Shear (1970) placed Hypochiius between steps 2 and 3 in the scheme of spider prey-capture evolution pro- posed by Eberhard ( I 967) :
I. No web is spun, prey subdued by biting. 2. Ground web spun, prey subdued by biting. 3. Aerial web spun, prey subdued by biting and wrapped to pre- vent loss during subsequent attacks.
4. Aerial web spun, prey subdued by biting and wrapping. 5. Aerial web spun, prey subdued by wrapping only.



================================================================================

188 Psyche [September
My own observations indicate that Hypochiluj may be placed somewhat closer to step 2, since its web, both in placement (on rock ledges) and in function (capturing ground and aerial arthropods in nearly equal numbers), is neither strictly a ground nor strictly an aerial web. Kaston ( 1964) hypothesized that early snares arose from a fringe of drag lines radiating from the mouth of a silk-lined retreat. Shear (1970) proposed that "the web of Hypochihs can easily be derived" from such a structure, and suggested that the lampshade may represent an extension of the original retreat, while the paired support lines may represent the capture (originally drag) lines in Kaston's scheme.
Defensive Behavior
The defensive behavior of H. thorelli is very similar to that de- scribed by Shear (1970) for H. gertschi. There are three major components to the defensive behavior : "vibrating," "running," and "death-feigning" ; and one minor component, "crouching." The vi- brating response is elicited by a mild disturbance of the frame and support lines. The spider reacts by oscillating its body rapidly in a plane perpendicular to the substrate surface. Disturbance of the mesh of the lampshade may trigger one or more of the other three responses. The "crouching" response is one in which the disturbed spider withdraws to one side of the web base and huddles there. Usually, if the disturbance is continued, this is followed by the "running" response, in which the spider cuts through the lampshade with its chelicerae and runs along the substrate. The spider may instead scramble over the mouth of the lampshade, and then run along the substrate. Rarely, the spider cannot be provoked into any action other than "crouching."
Touching the spider or suddenly disturbing the mesh often re- sults in the spider's leaping suddenly out of the web. Leaping (rather than simply dropping) enables spiders in webs on vertical rock faces to clear the edge of the lampshade.
Upon striking the ground, the
spider assumes the folded position shown in Figure 5. Spiders in
this position are very difficult to see in the ground cover beneath the web. They remain in the "death-feigning" position for j to I j minutes, then climb back up the substrate until they encounter either the tangled portion of the web or the web base. This procedure
usually presents no particular difficulty, as the spiders are directly under the web when they land. No dragline is spun when the spiders leave their webs by running or leaping. Probably gravitational cues are used in orientation, at least until silk is encountered.



================================================================================

Figure 5.
Mature female of H. tkordli ini "death-feigning" position. The legs of this specimen are less tightly flexed than usual, since the spider had to be moved in order to obtain the photo. Photo by L. Tucker.



================================================================================

190 Psyche [September
Mature males, when disturbed, either run or flatten themselves against the substrate, but none were observed to "feign death." The function of the vibrating response is unclear. It does not resemble the "testing" of web tension, which is accomplished in a different manner (see Predatory Behavior above). It may aid in camouflage by blurring the visual image received by a predator. Vibrating may function in dispersing a defensive chemical, as it does for some opilionids, or it may be mimicry of such behavior. Another important defensive mechanism is the cryptic coloration of Hypochilus (Figure 4), which allows it to blend in remarkably well with rock surfaces when in its usual resting position. H. thorelli does not display autotomy, a common defensive mech- anism among most other spiders.
Reproductive Biology
Morphological and behavioral changes. Males mature early in August. (I collected the first mature specimen on 5 August; they became abundant soon after then until October, and the last one was observed on 7 November.) The males undergo their final molt in a molting web. This web has a circular base of about the same size as the base of the usual web, but the sides of the web extend away from the substrate for about 12-15 cm, tapering to a mouth somewhat smaller than the base. The mouth is sealed off with a loose network of silk. Whether they alter their old webs or move to new locations to build the molting web is unknown. The molt- ing web probably offers greater protection from predators during the (presumably) more difficult final molt. W. A. Shear (pers. comm.) reports that the same type of web is constructed by males of H. gerischi.
Sexually mature males are strikingly different in appearance from immature specimens and females. In addition to other morphologi- cal changes, they develop relatively longer legs (Figure 6) in the final molt and the legs change in color from pale yellow and brown to dark reddish-brown. The greater length of the legs of the males is related to their use as tactile organs during courtship be- havior (see below), and may also improve locomotion during their search for females (see also below). The reddish-brown coloration is a result of heavier cuticular sclerotization, which lends greater strength to the elongate appendages. The color is not as cryptic as is the normal coloration.
Several changes in behavior occur with the final molt (see also Defensive Behavior above). The males cease feeding, let their webs



================================================================================

19721 F e r g u s ~ Hypochilus thorelli 191 Figure 6. Mature male and female of /I. #hot.dli. Note the greater leg length of the male (on left). Photo by L. Tucker. Figure 7. Mating position of H. ihareili. Male black, female in out- line. Drawn from field sketch.




================================================================================

192 Psyche [September
degenerate, and wander in search of females. Most of the wander- ing is nocturnal, and the males are usually inactive during the day- light, often resting within the circular base of an abandoned web. Females do not undergo a drastic physical metamorphosis during their final molt, and they continue to carry on their normal preda- tory and defensive behavior patterns.
Courtship and mating.
In all 19 observations of courtship be-
havior, wandering males were seen to use their first pair of legs as "feelers," often waving them about when advancing along a rock surface. When a male makes contact with silk, he stops immediately with one or both first legs touching the silk. (No males were seen to contact any silk but that of ~Hypochilus webs; what their reaction is to other webs is unknown.) After a brief pause, he strokes the web with one first leg. The stroke may be either a "tapping" of the silk, or a plucking motion wherein he extends his leg over the silk, then draws it across the silk back toward his body. If this precipitates an attack by the female, he backs rapidly away for 8 to 15 crn. If no rush occurs (usually the case when the male has con- tacted the guy lines of the web), the male proceeds a little farther