Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 21, Issue 3, April 1975

The Postures of the Tunnel Web Spider Porrhothele Antipodiana: a Behavioural Study

page 108

The Postures of the Tunnel Web Spider Porrhothele Antipodiana: a Behavioural Study


Descriptions of the three main postures of the tunnel web spider P. antipodiana are given. A behavioural/ecological interpretation of each posture is attempted. Emphasis is placed on the importance of the strike posture; and an experiment suggests that the angle the spider assumes to strike at prey is related to the prey height.

Intra- and interspecific encounters are described and it is indicated that the behaviour of this spider may help protect it from attack by small animals such as mice.


The study of postures has been a recurring theme in the field of animal behaviour. Darwin wrote on the topic (1872), followed in the early years of this century by biologists like Heinroth and Craig. Most of the early studies were descriptive and many were concerned with the study of courtship in birds. Later writers began to use the study of postures as a basis for erecting models of behaviour, and the fixity of postures associated with instinctive behaviour was used in attempts to understand the functional organisation of the central nervous system in birds and fish. The so-called ‘hydraulic models’ of Lorenz and Tinbergen were the outcomes of these studies in the 1930′s and 1940's.

The study of postures is still popular with students of animal behaviour, but is more likely to be carried out as part of a wider study of animal communication than as a basis for interpreting the function of the nervous system. For example a recent paper by Wilson (1972) includes areas such as communication by chemical means and by sound as well as the role of postures in communication between animals.

In the field of spider behaviour, the most well known studies have probably been those of Crane (1949). These were concerned with the courtship behaviour of Salticid spiders, a group having good enough vision to communicate by posturing. Most spiders, however, have poor vision, so their postures are less elaborate and are most likely to be linked with the tactile rather than the visual senses. The present study is concerned mainly with an investigation of the rather spectacular aggressive or threat display of the large (body length often to 30 mm) New Zealand Mygalomorph spider, Porrhothele antipodiana, commonly called a tunnel web spider. This spider, like other Mygalomorphs, does not have a well developed visual sense; page 109 as a consequence its postures could be expected to be less elaborate than those of the Salticids studied by Crane.

Certain aspects of the natural history of P. antipodiana have been published already (Laing, 1973). This spider takes a wide range of prey types; it moves about the environment rather than staying in the one burrow as seems to be the case with the trapdoor spiders; it is strongly photo-negative; and it probably has some natural enemies — for example, the two Pompilid wasps Salius monachus and S. fugax. (For a wide-ranging discussion of the New Zealand Mygalomorphs, see Forster and Forster, 1973.)

The Main Postures of the Spider

page 110

1. The Resting or Withdrawal Posture (Fig. 1)

The legs are drawn up to cast the cephalothorax into shadow and the area occupied by the spider is at a minimum. This posture is seen when the tunnel in which the spider lives is cut open and the spider is exposed to the light. It is also seen when a spider that has been taken from its web is allowed to run away; in which case it will usually crouch in this posture in the first shaded area it comes to.

When in this posture, the spider may be stroked and prodded without eliciting much response. The same posture is shown by a wide range of spiders — the orb-web spider (Aranea), the large grass spider (Miturga), the Katipo (Latrodectus), and the trapdoor spider (Cantuaria), so it is probably a common posture dictated by spider morphology.

2. The Alert Posture (Fig. 2)

The legs are spread out, the spinnerets are lowered and the fangs may be drawn slightly out of their grooves. This is the posture seen when the spider is at the entrance of its tunnel at night, waiting for prey to cross its sheet web. In this posture, the spider is ready for rapid activity in any direction. When running toward the prey, the spider often overshoots the prey animal, coming to rest with its palps touching the prey. Because its body is raised slightly on the legs, the spider's ventral surface is some millimetres above the prey, and in this position the fangs need only be raised forward to be in position to strike. It is rarely necessary for the cephalothorax of the spider to be raised by more than 5-10 degrees from the horizontal for the spider to strike.

It seems to be a common assumption that because of the paraxial chelicerae (having fangs moving in a vertical path, not horizontally as in the true spiders) of the Mygalomorph spiders, raising of the body to a substantial angle from the horizontal is needed to bring the fangs into a striking position. The preceding description, which is based on several hundred observations — both day and night, and in the natural habitat of the spider — indicates that the posture of rearing up of the whole body is not used for routine prey capture by this spider at least. Table 1 shows a summary of the observations made on prey killing postures.
Fig. 1: The resting or withdrawl posture of P. antipodiana.

Fig. 1: The resting or withdrawl posture of P. antipodiana.

Fig. 2: The alert posture of P. antipodiana.

Fig. 2: The alert posture of P. antipodiana.

TABLE 1: Approximate angle to which body is raised for prey capture in P. antipodiana
AngleNo. of Observations
The anatomy of the fang may explain how P. antipodiana can strike at prey without rising up to a substantial angle: the length of the fang is about 5 mm in a spider of body length 25 mm and it can be swung through 130 (Fig. 3). This brings the fang into a position where its tip strikes vertically downwards in the initial stages of its movement. Prey such as slaters and the smaller beetles have a body height of 2-5 mm usually, and as P. antipodiana may be standing with the lower edge of the cephalothorax more than 5 mm above the web surface, an upward raising of the cephalothorax of 2-3 mm is all that is needed for the fang to clear the prey on the upward swing. In addition, each parturon is moved sideways from the midline of the body to an angle of 30 degrees (Fig. 4), and this further increases the probability that the fangs, by travelling outwards rather than directly over the prey, will clear the prey on their upward swing. page 111
Fig. 3: The fang movement of P. antipodiana.

Fig. 3: The fang movement of P. antipodiana.

Fig. 4: The chelicera of P. antipodiana: (left) closed or at rest; (right) parted for a strike.

Fig. 4: The chelicera of P. antipodiana: (left) closed or at rest; (right) parted for a strike.

The phylogenetic significance between the strike action of the Mygalomorphs and that of the true spiders has been pointed out to me by Dr. R. R. Forster (pers. comm.): the Mygalomorphs need to strike downwards, and so are somewhat restricted to a solid substrate where the pressure exerted by the fangs can be taken up. page 112 The true spiders, however, transfix their prey in front of them and so can take their prey while suspended.

3. The Aggressive or Threat Posture (Fig. 5)

The body is raised up on the second, third and fourth pairs of legs, with the first pair directed upwards and almost fully extended. The angle the body reaches relative to the ground is often 60 degrees or more; while the fangs are swung forward and drops of venom often appear at the tips. In this position the spider has exposed the central surface of its body — the vulnerable areas such as the leg to cephalothorax joints and the lung openings are very accessible to a predator when the spider assumes this posture.

This posture is often adopted by the spider when it is removed from its web and is prodded lightly with a piece of stick. It is also adopted when the spider is in contact with a small vertebrate such as a lizard or a mouse. As to how widespread the use of this posture is in the normal living activities of the spider, it is difficult to judge for there are possibly few occasions on which such a display would be provoked.

Petrunkevitch (1952) described a similar posture for the North American tarantula Crytopholis*; noting that ‘… touch excites no defensive response unless the approach is from above where the spider can see the motion, in which case it rises on its hind legs, lifts its front legs, opens its fangs and holds this threatening posture as long as the object continues to move’.

It is probable that the field of vision of the spider is curtailed by the upright posture, for P. antipodiana's ewes are placed quite high on the caput (raised area of cephalothorax where eyes are grouped), but the caput is not directed forward over the front margin of the cephalothorax. On morphological considerations it is probably difficult for this spider's visual field to extend downwards at a sharp angle. If the angle given in Fig. 6 is taken as a rough guide, then a spider of body length 20-25 mm, with the body raised at an angle of 50 degrees relative to the ground, would not be able to detect the movement of a small predator unless that animal was at least 20 cm distance from the spider. From this point of view, this threat posture could not be a very efficient form of defence as the spider cannot use its vision to help orient itself in relation to a predator such as a hunting wasp. Tactile stimuli do have a very marked effect on the spider while it is displaying — any touch to the legs results in the spider striking vigorously in the direction from which the touch came. The spider will rapidly re-orient itself in relation to the stimulus, swinging to the side or turning completely around to face in the appropriate direction. Once brought to the display posture, P. antipodiana will often remain in that position for more than five minutes, no extra stmulation being required to keep them this way. For comparison, Petrunkevitch's spiders resumed their prone posture once the external stimulation had ceased. page 113
Fig. 5: The threat display of P. antipodiana.

Fig. 5: The threat display of P. antipodiana.

What Determines Angle of Strike?

From the observations on strike behaviour — that is strikes resulting from prey killing or from when the spider is displaying — the hypothesis was formed that it might be the height of the object that touched it that influenced the posure assumed by P. antipodiana. The hypothesis was tested initially by presenting blocks of modelling clay of different heights to the spider. The blocks, mounted on thin wire, were lightly touched against the front legs of the spider to simulate an animal coming into contact with the spider (Plate 1). Three blocks were used: 5, 10 and 15 mm in height. Each was presented five times to each of the fifteen spiders used. The approximate heights to which each spider responded to each block were averaged out to the nearest ten degrees to be used as ranks in a Friedmann Two-way Analysis of Variance. The results (Table 2) showed the average response of the spiders to the three models to be significantly different at the 0.1% level. page 114
Fig. 6: A representation of the lower limit of P. antipodiana's field of vision when displaying. Line A is drawn through the body of a spider raised 50 degrees above the horizontal; line B is the probable lower limit of vision; the object in front of the spider represents a predator moving in the area not covered by the spider's visual field.

Fig. 6: A representation of the lower limit of P. antipodiana's field of vision when displaying. Line A is drawn through the body of a spider raised 50 degrees above the horizontal; line B is the probable lower limit of vision; the object in front of the spider represents a predator moving in the area not covered by the spider's visual field.

TABLE 2: Friedmann two-way analysis of variance of spider responses to blocks of three different heights
Spider No.5 mm blockRank10 mm blockRank15 mm blockRank
Rank totals:193349
FriedmannXr2— 51df — 2P — .001
page 115
Plate 1: A large female P. antipodiana strikes at a 15 mm high block. The angle of strike is about 40 degrees from the horizontal.

Plate 1: A large female P. antipodiana strikes at a 15 mm high block. The angle of strike is about 40 degrees from the horizontal.

The result of this testing seems a reasonable indication that raising of the body for striking is at least partly controlled by the height of the animal that contacts the spider. The spider must be ‘measuring up’ the animal or object that is contacting its front legs. It has two opportunities for doing this — first, when the initial contact is made with the prey and, second, when the spider has raised its legs for the sweep downwards to draw the prey in to the fangs. As the legs descend and contact the prey, the spider gains more information to assume a strike position appropriate to the size of the prey. The results of the experiment fit in reasonably well with the statements made earlier when describing the normal prey killing posture and they are also compatible with the comment on phylogenetic significance: that is, the spider rises to a height from where it can strike down at the prey; and in its web the spider is dealing with prey that are in the main 5 mm or less in height, so when it measures them up, the angle of strike that results will be quite low.

It is probable that this simple model based on a linear relationship between stimulus and response may not be detailed enough to explain why many spiders will jump up to extreme angles of 60 degrees or page 116 more, with legs outstretched to the full and body quivering. Sometimes even a slight touch may set off one of these extreme responses — rather than assuming a specific strike posture, the spider appears to be displaying itself as a large, formidable animal. This aspect has been given further consideration in the section on interspecific encounters.

The part played by vision in determining the strike posture was investigated briefly: five spiders were temporarily blinded by painting over their eyes with black dope. The three clay blocks were presented to them as in the first experiment. The results (Table 3) are consistent with those for the normal spiders, indicating that angle of strike is probably little affected by visual stimuli.
TABLE 3: Responses of five blinded spiders to blocks of three different heights (Figures are average of five trials with each spider to each of the blocks)
Angle to which body is raised5 mmHeight of Block 10 mm15 mm

Postures in Intraspecific Encounters

The possibility that the threat posture might be linked with territorial behaviour or with intraspecific interactions in general was investigated. Pairs of spiders were placed together in a container of size 30 cm x 30 cm wide and 5 cm high. The encounter took place in dim light — one 25-watt bulb at 4 metres distance being the sole illumination. Thirty spiders were paired in different encounters, some being used several times to give a total of 100 encounters. Male/male, male/female, and female/female pairings were tried, but as the activities of the spiders were no different from one group to another, the data were massed with the following results: in 70% of encounters, one spider quickly retreated; in 15% of encounters both spiders quickly retreated; and in the remaining 15% both spiders gave fight. In these same 100 encounters, the approximate angle to which the body was raised to deliver a strike was estimated. The results were: 0-10° raising of body 51%; 10-20° raising of body 42%; 20°+ raising of body 7%. At no stage in these encounters was an extreme threat posture noticed. When contact took place between two spiders, one or both usually threw out the front legs and delivered a strike in the direction where contact was made.

page 117

In further attempts to provoke threat displays between these spiders, two individuals were placed in glass containers and brought close together in daylight conditions; then one spider was prodded into displaying in front of the other. There was no observable response from the spider in the opposite container. Holding a mirror in front of this spider also failed to elicit a display, further indicating the small part played by visual stimuli in its behaviour.

Postures in Interspecific Encounters

One of P. antipodiana's common neighbours under rocks and logs is the large brown grass spider, Miturga. It is possible that encounters between the two do take place in their environment. To test their reactions to one another, individuals of both types were placed in the container used in the first encounter experiment. In 80% of the encounters Miturga quickly retreated when contact was made; in 15% of the encounters both spiders quickly retreated on making contact; and in 5% P. antipodiana quickly retreated on making contact. In these encounters, the angle of strike assumed by P. antipodiana varied as follows: 0-10° raising, 55%; 10-20° raising, 40%; 20° + raising, 5%. The angle of strike was consistent for both intra and interspecific encounters then, and P. antipodiana was not seen to display the high angle threat posture in front of Miturga.

It may be of interest to note that when Miturga and P. antipodiana are matched on a body length basis, and when grappling does occur, the grass spider is overcome in eight out of ten encounters.

In its normal habitat, P. antipodiana probably encounters small vertebrates such as skinks, mice, rats and possibly hedgehogs. Spider remains have been found in the stomachs of both rats (Daniel, 1973), and hedgehogs (Campbell, 1973) in New Zealand studies, so it is reasonable to assume that a large spider like P. antipodiana could be used as food by these animals. Experiments were conducted to see how readily mice could kill and eat this spider, and to find out how the spider responded to attack by a mouse. The mice used were a mixture of strains: some were the result of accidental matings between white laboratory mice and wild mice living around the building, while others were the white mice and black mice commonly used in breeding experiments. The half-wild mice proved to be excellent animals for the spider experiment — they were far more aggressive and speedy than the laboratory mice. When a spider was introduced into the cage of one of the half-wilds, the mouse first investigated the spider for several seconds, sniffing closely at it. The spiders usually responded quickly to this by rearing up to an angle of 70 degrees or more and displayed vigorously at the mouse. The mice always attacked shortly afterwards (Plate 2), using a series of quick, darting rushes, nipping at the outstretched legs of the spider. The spiders struck repeatedly at the mice, but the latter withdrew very page 118
Plate 2: A mouse attacks a large female P. antipodiana, which has thrown itself into a high angled threat posture.

Plate 2: A mouse attacks a large female P. antipodiana, which has thrown itself into a high angled threat posture.

quickly beyond the reach of the spider's legs and fangs. As the combats proceeded, the spiders usually tired and when pieces of their legs began to be bitten off, the resulting loss of body fluid spelled the end of the encounter. Finally, the mice would grasp the spider first in the front paws and proceed to eat most of the abdomen and the cephalothorax.
Several facts emerged from the mice/spider encounters: that mice will eat this spider; that the active defence put up by the spider does keep the mouse at bay for up to two minutes in some cases; that mice do have a behaviour pattern (probably innate) that enables them to attack this large spider in an efficient manner. The risk to the mice in these encounters is substantial, for when P. antipodiana does catch hold of a mouse and strike home for two to three seconds, then the effects of the bite prove fatal to the mouse. The immediate effects of the bite are that the mouse loses its orientation, runs around jerkily and shivers rapidly. Death follows in three to five hours. In some encounters, mice were bitten for a short time only — perhaps one second; their response to this was to break off the encounter at once and retreat to their nest. These mice were presented with the spiders again — sometimes up to one week following the encounter page 119 in which they received a non-fatal bite. All of the mice in this group did investigate the spider they were presented with, and all retreated rapidly as soon as the spider reared up in a threat posture. The results of ten mice/spider encounters are given in Table 4.
TABLE 4: The effect of non-fatal bites on the subsequent behaviour of mice
First EncounterSecond Encounter
Mouse No.BittenNot BittenAttacksRetreats

The data on these mice spider encounters are probably too few to draw any conclusions from, but it is tempting to view the display posture of P. antipodiana as a communicator of information and having a biological effect. Several other facts could be tied in with this view as well: Miturga, whose bite has little effect on mice, does not usually display to mice, but attempts to escape from the encounter as rapidly as possible. When mice do catch Miturga, they overcome it in a matter of seconds, showing little reluctance to come in close and grasp this spider in their paws.

Some Conclusions

(a)The first posture described seems to need little comment — it is primarily concerned with concealment of the spider.
(b)The second posture is of interest because of the common assumption mentioned in the text. It would seem that when observers have thought the spider must rear up to kill prey, the spider was either responding to a stimulus high above it, or was adopting its threat posture.
(c)The third posture is possibly the most interesting — and the question could still be asked: is it merely a response to a certain set of stimuli, representing a high angled strike position; or is it a threat display involving a state of great excitation in the spider? From the bias in the text it will be obvious what position this writer has taken on the question.page 120
(d)P. antipodiana is quite a successful spider — it has a far wider distribution than other New Zealand Mygalomorphs. Its range stretches from sand dunes to beech forest; it is at home in a variety of habitats: farm paddocks, stony hill outcrops, clay river banks, rotten forest logs, up trees, around dwellings, and so on. It certainly shares these habitats with a variety of other spiders, and possibly competes with some of them for web sites and food. One of its common co-habitants is the brown grass spider Miturga, a large and vigorous competitor. That P. antipodiana thrives may be due to a number of factors such as its ability to accept a wide variety of prey types., its ability to build its web in a wide variety of sites, its speed of running over its web in pursuit of prey, and the strength of its venom. Behaviour, too, must play a part in its success: its aggressiveness and fighting ability must help it in interspecific encounters, and in subduing large prey like ground beetles and bumblebees. Its readiness to defend itself against larger animals and the habit of displaying its potential for offence are probably important characteristics of this species, which would have to be taken into account to gain a fuller understanding of an animal like P. antipodiana.


My thanks are due to Dr. R. R. Forster. Director of the Otago Museum, for his helpful and encouraging comments on this work.


Bristowe, W. S., 1958: The spider's world. London, Collins.

Campbell, P. A., 1973: The feeding behaviour of the hedgehog (Erinaceus europaeus L.) in pasture land in New Zealand. Proc. N.Z. Ecol. Soc. 20: 35-40.

Craig, W., 1918: Appetites and aversions as constituents of instincts. Biol. Bull. 34: 91-107.

Crane, J., 1949: The comparative biology of Salticid spiders at Rancho Grande, Venezuela. Part IV: An analysis of display. Zoologica 34: 159-214.

Daniel, M. J., 1973: Seasonal diet of the ship rat (Rattus r. rattus) in lowland forest in New Zealand. Proc. N.Z. Ecol. Soc. 20: 21-30.

Darwin, Charles, 1872: The expression of the emotions in man and animals. London.

Forster, R. R., 1967: The spiders of New Zealand, Part 1. Otago Mus. Bull. 1.

—— and Forster, L. M., 1973: New Zealand spiders — an introduction. Auckland, Collins.

Laing, D. J., 1973: Prey and prey capture in the tunnel web spider Porrhothele antopodiana. Tuatara 20: 2, 57-64.

Petrunkevitch, A., 1952: The spider and the wasp. Sci. Am. 187: 2, 20-23.

Siegel, Sidney, 1956: Nonparametric statistics for the behavioral sciences. New York, McGraw-Hill.

Wilson, E. O., 1972: Animal communication. Sci. Am. 227: 3, 53-60.

* This spider is a large Mygalomorph common in the southern United States.

Bristowe (1958) gives several examples of spiders measuring up prey.