Non-response to stimuli from the sheet web during daylight may seem to be stating the obvious; however, when examined in detail, like any other biological variable it turns out to be complex. Mygalomorph spiders generally tend to be photo-negative and avoid exposure to light. They are reluctant to venture out of their tunnels in periods of bright light. There is a conflict of behaviour here, for it is one of the spider's most natural responses to react to web stimuli, for this is how they obtain their food. This conflict of drives can be demonstrated by dropping a slater into a sheet web and watching. Often the spider will appear at its tunnel entrance and then come no further. The author's interpretation of this behaviour is that at first the spider responded to the web stimuli, but on moving along its tunnel the light became progressively brighter until it inhibited the approach response to the prey in the web.

Not all individuals in a P. antipodiana population respond in the same way to web stimuli during the day; some appear readily at the tunnel entrance whereas others can never be enticed out of their tunnels. To illustrate this point, the response times of 42 spiders in one population are given in Figure 4. All of the spiders were second year or older and the test was performed with slaters; the slater was dropped on its back into the sheet web and left to entice the spider out. Times were recorded from when the slater was dropped in until the spider made its appearance at the edge of the sheet web. The P. antipodiana individuals which responded rapidly were described as ‘reactors’, while those individuals taking 60 seconds or longer to appear were described as ‘non-reactors’.

The reactors are the individuals which will be most at risk from the wasps, for they would be vulnerable, being out in the open and in strong light. These are most likely to be the spiders that Miller (1971) had in mind when he wrote of the wasp being naturally cautious and waiting until the spider had been enticed out of its tunnel before delivering the paralysing sting.

Evidence for the significance of this survival factor was obtained from the Johnsonville crib wall population. Here it was the page 14 page 15

Fig. 3: The six main survival tactics of the spider when the possibility of wasp attack occurs: a. non-response to web stimuli; b. dense silk cover over tunnel opening; c. aestivation; d. presence of a side tunnel; e. rapid evacuation from the tunnel; f. active defence in the tunnel.

Fig. 4: Histogram of reaction time for group of 40 mature spiders; test used being number of seconds for spider to react to a slater placed on the sheet web in daylight. This split into two groups for reaction rates is a common feature of P. antipodiana populations.

page 16

Fig. 5: An example of the relationship between spider reaction time and capture by wasps. The closed circles represent spiders from a crib-wall population, with the ‘rectors’ at the left-hand end. The arrowheads denote those spiders captured by wasps during November.

‘reactors’ which disappeared from their webs very early on in the wasp season. Fig. 5 shows the results from one summer investigation on reaction times and survival.

Results such as this confirm that a slow reaction time is likely to be important for the survival of P. antipodiana in the face of hunting wasp predation.

The work by Coville (1976) on the wasp Chalybion contains a similar inference; that the reactor spiders are those most likely to be captured by the wasp. This variable may well be of widespread importance in spider-wasp relationships.

Survival Factor (b)

During the summer months it was common to see the openings of P. antipodiana tunnels covered by layers of silk, the thickness of which varied from tunnel to tunnel. Why this was done is not clear, but it may have been a sign of temporary inactivity by the spider. Whatever the case, the cover certainly does act as a partial deterrent to hunting wasps intent on entering the tunnel. It did not represent a complete barrier, for a determined wasp could force its way through. Often, though, the wasps moved off to investigate other areas after some entanglement with the silken barrier. When the number of wasps seen entering covered tunnels was compared with those entering open tunnels, it was found that the silk provided a significant degree of protection (Chi-square = 8.8; P < .01).

Survival Factor (c)

On a number of occasions, wasps were seen to enter and then remain within a tunnel for ten seconds or longer before reappearing. Several of these tunnels were opened up to see if a spider was resident there. In some of these instances, the spider was found sheltering at one end of a divided tunnel. Side tunnels of this type (see Fig. 3, d) are not found in all P. antipodiana webs, but where conditions permit page 17 as in soft ground or under a long, the spider may construct more than one part to a tunnel. The effects of these split tunnels are even more difficult to assess than the other factors, for it cannot be known if the wasp did in fact find the spider but failed to carry out the capture for some reason. However, it is likely that split tunnels do lower the chances of the wasp making contact with the spider and so this factor could have some survival value.

Survival Factor (d)

One of the surprising features of P. antipodiana was the discovery that during the dry months of summer, that is any time from December onward, up to 20% of a population was likely to be aestivating. Aestivation over summer may be common in Mygalomorphs, for Forster and Forster (1975) have noted that the trap door spiders of the genus Cantuaria also aestivate over summer until autumn. Aestivation introduces a difficulty in assessing the numbers of spiders that have been captured by wasps, for during this phase the sheet web becomes broken and weathered, in the same fashion as when a spider is no longer resident. Counts made later in the summer or early autumn usually reveal which spiders have been aestivating and which have gone from their webs. The effects of aestivation, apart from conservation of body fluids and food reserves, are that wasps do not usually waste time investigating old webs; thus aestivation may protect a spider from wasp attack.

Survival Factor (e)

Experiments carried out by the author on spiders and wasps in captivity showed that the larger spiders could sustain several quick stings from a Salius wasp and yet continue to run, even if a little unsteadily. The effect of the first sting was usually to make the spider take evasive action. Observations under natural conditions have been similar; with spiders observed escaping from their webs and running fast enough to escape the wasp, which is usually left running excitedly in circles in search of the spider. Leaving the web at speed after the initial contact with the wasp must enable a number of spiders to survive.

Survival Factor (f)

It has already been pointed out that the mortality rate of the wasps that can be attributed to the bite of the spider is at least 2.6% and may be as high as 11%. Active defence on the part of the spider is certainly a survival factor. Often the combat does not proceed to the point where one party is overcome; the wasp has been observed leaving the tunnel entrance where it could be seen wrestling with the spider. These cases always involved large spiders, and it would seem that the wasp was often reluctant to proceed against spiders with a page 18 body length in excess of 25 mm. In the cases being discussed it was possible to confirm afterwards that the spiders were still active. Some cases were observed where the spider was left semi-paralysed just inside the tunnel entrance.

The reasons why wasps may not persist against large spiders are most likely to be or a combination of the following:

The second reason is certainly a practical one, for the diameter of the burrows dug by S. monachus ranged from 10-14 mm in the localities studied. A burrow of this size would not accommodate the very large spiders.