Jumping Spiders Alternative Turns in the Abscence of Visual Cues

: The abilities of four species of diurnal jumping spiders (Helpis minitabunda, Portia fimbriata, Trite auricoma, and Trite planiceps) and one species of nocturnal clubionid spider (Clubiona cambridgei) to maintain approximately straight paths by alternating turns in the absence of visual cues was investigated. Under infra-red light (observed using infra-red video), individual spiders were run through a maze comprising a single forced turn and then a choice of turning in the same or opposite direction to the forced turn. At the second (free) turn, each species turned in the direction opposite to the forced turn (i.e., alternated turns) more frequently than it turned in the same direction. There was no evidence that species differed in tendency to alternate turns. In nature, jumping spiders may use this ability to navigate in the absence of visual cues when foraging or escaping predators in darkness. It is suggested that alternation of turns by jumping spiders depends on proprioceptive information gathered during previous turns.


INTRODUCTION
Turn alternation is the tendency of animals to turn in the direction opposite to the previous turn after a forced turn or at successive choice situations. This tendency has been described in a wide taxonomic range of animals (for review, see Hughes, 1989), and it has been suggested that turn alternation is used by animals in nature as a means of approximating straight paths in the absence of external directional cues (Dingle, 1965;Hughes, 1978). Maintaining straight paths may facilitate exploration, dispersal, foraging, and escape behaviours (Hughes, 1990;Richman et al., 1987).
Jumping spiders (Araneae, Salticidae) are unique amongst spiders, as they have complex eyes and acute vision (Williams & Mclntyre, and navigation abilities of jumping spiders have been studied (Hill, 1979;Tarsitano & Jackson, 1992;Jackson & Wilcox, 1993), other orientation mechanisms also underlie their behaviour. For example, Land (1971) found that, following target detection by lateral eyes, jumping spiders made turns to orient toward targets and that these turns did not depend on visual feedback during the execution of the turn.
That is, these turns are governed by an 'open-loop' system, whereby the turning angle is pre-set by the eyes, and the turn is mediated by internal information, rather than information received by the eyes during the turn.
Jumping spiders also retain a memory of their relative position in space, using combinations of route-referent, gravitational and visual cues.
They use this information while making detours to re-orient toward targets after moving to new locations (Hill, 1979). Although Hill (1979)  They were maintained using the standard procedures described by Jackson and Hallas (1986).

Apparatus
The testing arena was designed to present the subject with a single forced turn followed by a free turn on a flat perspex runway ( Fig. 1). All sections of the runway were 10 mm wide, and all turns were 90°. The 'starting platform', 'entry platform', and 'ramp' were all 50 mm long. The runway on each side following the choice point was 50 mm long.
The testing arena was placed in a tray containing water, and the water level was maintained so that a miniscus formed against the arena edge leaving only the upper surface of the arena out of the water.
This was usually adequate to keep the spiders on the arena because all of those tested tend to avoid water.
All experiments were conducted inside a wooden light-proof cabinet (800 mm high, 1200 mm long, 500 mm deep) and were observed using an infra-red (IR) video and light-source that was set up within the cabinet and connected to a monitor positioned outside.
There was a heavy black satin sleeve (500 mm long) attached to a 150-mm hole in the wall of the light-proof cabinet. By reaching through this sleeve, the experimenter could operate experiments without allowing light to enter the cabinet. The experimentor's arm remained within the sleeve while experiments were in progress, thereby eliminating the possibility that light could enter the cabinet during withdrawal of the arm. The inside of the light-proof cabinet was painted matt black and the door surrounds contained soft black rubber seals that overlapped the door, thereby excluding light. The spiders were maintained and tested under a laboratory photoperiod of 12L : 12D. All tests were carried out during the light phase, although no tests were conducted during the first and last 2 h of the light phase.
To begin a test, a spider was placed on the starting platform outside the light-proof cabinet. This was achieved by holding the platform in front of the spider and gently touching its hind-most legs with a camel hair brush so that it walked forward onto the platform.
The spider usually stopped at the far end of the platform long enough for the test to be started.
If the spider remained active, it was detained by repeatedly rotating the platform. Once the spider stopped moving, the starting platform was placed into the water (under IR light) by holding it with a hand that extended through the sleeve. The starting platform was then pushed toward the entry platform so that the two platforms joined to form a single 100-mm-long runway.
Spiders then walked off the starting platform and along the entrance platform to the 90°' forced turn'.
Half of the tested individuals of each species were forced right, and the other half were forced left. Following the single forced turn, the spider walked along the ramp until it encountered the 'choice point' where it was given the option of turning either in the same direction as the forced turn (repeating), or in the opposite direction to the forced turn (alternating).
A choice was defined as having reached the end of one of the runways following the choice point. Each subject was tested once. However, if a spider failed to complete a test by leaving the arena (e.g., jumping or walking onto the water surface) or did not complete the test within 10 min, it was retested at least 2 h later. Any spiders that failed to complete the test on the second attempt were discarded.

RESULTS
In the absence of visual cues for pre-setting a target or as feedback during locomotion, each species alternated turns more frequently than it repeated turns. The numbers of each species tested, the number and percent that alternated turns, and the result of onesample chi-square tests can be seen in Table I.
There was no evidence that the five species differed in frequency of turn alternation (test of independence, x^= 7.08, df=4, ;? > 0.1). The present study was designed primarily to investigate only whether jumping spiders alternate turns when visual cues are absent, rather than to identify the sensory mechanisms underlying navigation in these conditions. However, because tactile cues were minimised by conducting experiments on an arena that lacked walls, it seems unlikely that turn alternation was mediated solely by tactile-dependent responses such as 'centrifugal swing and thigmotaxis' (Schneirla, 1929; see Hughes, 1989). Instead, it appears likely that proprioceptive cues were involved.
Recent studies have indicated that proprioceptive feedback generated by 'bilaterally-asymmetrical leg movements' (BALM, Hughes, 1985) may be used by arthropods during turn alternation.
Jumping spiders appear to rely on proprioceptive cues gathered by stepping movements of legs to mediate orientation toward visual stimuli in light (Land, 1972).
That spiders also rely on proprioceptive cues to mediate turn alternation has been suggested previously by Hill (1979) and Schafer (1982).
Further studies investigating cues used by jumping spiders during orientation and navigation in darkness are clearly required.