Table Of Content1999. The Journal of Arachnology 27:663-671
ANTERIOR MEDIAN EYES OF LYCOSA TARENTULA
(ARANEAE, LYCOSIDAE) DETECT POLARIZED LIGHT:
BEHAVIORAL EXPERIMENTS AND
ELECTRORETINOGRAPHIC ANALYSIS
J. Ortega-Escobar: Faculty of Psychology, Univ. Autonoma, 28049 -Madrid, Spain
A. Munoz-Cuevas: Laboratoire de Zoologie (Arthropodes), M.N.H.N.-C.N.R.S., 61
—
Rue Buffon, 75231 Paris Cedex 05, France
ABSTRACT. We studied solar light cues that may be used by Lycosa tarentula (Linnaeus 1758) (Ar-
aneae, Lycosidae) for homing. Experiments performed under clear skies, under overcast skies, and under
clearskies as seenthroughaplastic sheet(whichchangedthepolarizationfromlineartoelliptical),allowed
us to discover which attributes of daylight were used by the spiders during orientation and homing. We
found that patterns of linearly polarized light in the natural sky were sufficient to allow accurate homing
by the spiders. The homing behavior of individuals having the anterior median eyes (AME) or all other
AME
eyes blinded allowed us to determine that were responsible for the reception of polarized light.
Electroretinography of all eyes confirmed that only the anterior median eyes were differentially sensitive
to the orientation ofpolarization in linearly polarized light.
Arthropods are known to detect linearly po- Linearly-polarized light is among the visual
larized light by means of structural speciali- cues used by some spiders. In this group, an
zations of their eyes, whether compound or orthogonal disposition of rhabdoms in the
single ocellar. An orthogonal arrangement of ventral part of the retina of the anterior me-
rhabdoms in certain parts of the retina is the dian eyes (AME) has been found in the age-
anatomical basis of polarized light perception lenid Agelena gracilens C.L. Koch 1841 (see
(for review see Wehner 1989). This has been Schroer 1976) and the lycosid Lycosa taren-
demonstrated in the compound eyes of some tula (Linnaeus 1758) (see Kovoor et al. 1993),
insects. A retinal specialization, called “POL and in the peripheral part of the retina of L.
area” (Wehner & Strasser 1985), is located in erythrognatha Lucas 1836 and L. thorelli
the marginal dorsal part ofthe compound eyes (Keyserling 1876) (see Melamed & Trujillo-
and oriented towards the zenith when the an- Cenoz 1966), while Baccetti & Bedini (1964)
imal is walking. In other insects, for example could not find a similar arrangement in the
in the water bug Notonecta glauca, the POL lycosid Arctosa variana C.L. Koch 1847. In
area occurs in the ventral rather than dorsal A. variana, Magni et al. (1965) analyzed the
part of the eye; and this insect uses polarized electroretinographic responses ofthe different
light reflected from the water surface to detect eyes to polarized light. They found that both
the ponds into which it dives (Schwind 1984). AME and PME (posterior median eyes) are
It has been experimentally shown that the po- capable of analyzing the plane of polarized
POL
larization-sensitive area plays a role in light.
celestial navigation in Cataglyphis spp. (Weh- Behavioral studies on orientation by polar-
ner 1982) andApis mellifera (Wehner & Ros- ized light in spiders were initiated by Papi
sel 1985; Wehner & Strasser 1985). These Hy- (1955a, b) in lycosids and by Corner (1958)
menoptera were unable to orient themselves in agelenids. Both investigators found that the
correctly when the POL area was masked. only eyes involved in polarized-lightdetection
For spiders, Wehner (1992) reviews the were the AME (Comer & Claas 1985; Magni
processes and mechanisms by which they can et al. 1964). Papi & Syrjamaki (1963) studied
return home: they can use idiothetic, tacto- the orientation of an Arctic population ofLy-
chemical or visual information for homing. cosa fluviatilis{= Pardosa agricola (Thorell
663
—
664 THE JOURNAL OF ARACHNOLOGY
1856)), which exhibited a correct solar ori- parent open glass container and transferred to
entation throughout the day. the center of an open field 90 cm in diameter,
In the present study, the ability of Lycosa and left in the center of it, and in a different
tarentula to use celestial cues, mainly the po- direction from that of the burrow. The walls
larized light pattern of the sky, for homing of the open field were 60 cm high and com-
was examined experimentally. The role ofthe pletely white. The periphery was divided into
anterior median eyes in the expression of this 10° sectors to identify the direction followed
behavior was determined through eye-painting by the spider. The position of the spider was
experiments, and polarization sensitivity was recorded when it was at 40 cm from the cen-
studied by electroretinography of all the eyes. ter.
METHODS Experiments were carried out under five
— conditions, in all cases with the sun obscured
Subjects. Lycosa tarentula is a ground- by an opaque screen (Table 1). Eyes were
living lycosid which constructs a burrow near- made non-functional by covering them with
ly 15 cm in depth and 3 cm in diameter. As three coats of black paint (Pelikan Hobby
do the other members of the family, it has Tempera #11).
eight eyes arranged in three rows. The front It was further ensured that the acrylic plas-
row is composed of the anterior lateral eyes tic (a Plexiglas® sheet with a polyethylene
(ALE) and anterior median eyes (AME). The film) changed linearly-polarized light to ellip-
middle row comprises the posterior median tically-polarized light with maximum efficien-
eyes (PME); and the posterior one, the pos- cy at a specific angle between the electric field
terior lateral eyes (PLE). This spider is active of the light and the orientation of the sheet.
during the day and at night. This arrangement was accomplished by using
Experiments were performed on adult fe- a He/Ne laser (X = 632.8 nm) and introducing
males ofL. tarentula collected at Canto Blan- the Plexiglas sheet or the polyethylene film
co, 25 km from Madrid. They were captured between crossed polarizers, together with a
as immature individuals and maintained in the Soleil-Babinet compensator. We were able to
laboratory under an artificial light/dark cycle compensate for the phase change introduced
of 12:12 h (light on at 0800 h local time) and by the sheet and produce zero light at the de-
at 25 ±2 °C. They were fed with mealflies tector. The transmission characteristics ofthis
(Calliphora vomitoria) and crickets (Acheta sheet, measured with a spectrophotometer
domestica). Spiders were used for experi- (Hitachi U-2000), are shown in Fig. 2.
ments at least 5 days after their last molt. Statisticalanalysis: The directions followed
Voucher specimens of the adults have been by the animals are shown as circular distri-
deposited in the Museum National d’Histoire butions, which were analyzed using circular
Naturelle (Paris, France). statistics (Batschelet 1981), calculating the
Behavioral experiments. Adult females mean resultant vector for every distribution.
were transferred from theirindividual contain- Appendix I shows how we calculated the
ers to a terrarium measuring 60 X 30 X 35 mean angle of the sample and the angular de-
cm placed on the roofofthe Faculty building. viation and it describes the Rayleigh and Mar-
This terrarium had a 15 cm deep substratum dia-Watson-Wheeler tests. The statistical evi-
of soil; in the middle of one long side of the dence of directness was tested following the
terrarium, an artificial burrow was built, sim- Rayleigh test. If directness was evident, the
ilar to that which the spider digs in the field. confidence interval for the mean angle was
Experimental procedure: After 5 days of calculated to test whether the mean direction
habituation to the terrarium, experiments be- of the sample deviated significantly from the
gan. Spiders were gently pushed forward in direction of the burrow. For each condition
one of two paths running, right or left from (e.g., overcast sky, blue sky, etc.) the Mardia-
the burrow, along half-the-length and the full Watson-Wheeler test was used to test whether
width ofthe terrarium (Fig. 1). The orientation the two samples (animals that should orient
of the shortest return path was 30° NE for one towards 30° NE versus animals that should
direction and 300° NW for the other and a orient towards 300° NW) differed significantly
distance of 35 cm. When the spider arrived at from each other. —
the end ofthe path, it was placed into a trans- Electroretinographic analysis. Lycosa
ORTEGA-ESCOBAR & MUNOZ CUEVAS—POLARIZED LIGHT DETECTION IN LYCOSA. 665
—
Figure 1. Apparatus used to study homing in L. tarentula. Right, top view of terrarium in which the
animal lived during the study; arrows indicate the possible outward paths. Left, dorsal view of the open
field in which the animal was left after being taken from one of the comers opposite to the burrow. 0°
was always oriented towards the north. The big arrow indicates the translation ofthe animal to the center
of the open field (shown at half of its actual size in relation to the terrarium).
tarentula females from Canto Blanco (Ma- trode was athin wire, positionedonthe cornea
drid, Spain) were maintained in the laborato- by means of a Prior micromanipulator. The
ry, in Paris, at 20 °C and under natural light- electrodes were connected to a high input im-
dark cycles (LD: 10/14). Animal weight was pedance solid state amplifier. The animal, the
about 2 g. Diurnal electroretinograms were micromanipulator and the amplifier were en-
obtained as described in Carricaburu et al. closed in a Faraday cage. The output of the
(1990). For the recording of ERGs, the ani- amplifier was connected to a cathode ray os-
mals were placed on a metallic plate used as cillator (CRO), the ERGs were displayed on
the indifferent electrode. The different elec- the screen and photographed. The light stimuli
—
Table 1. Experimental conditions that were used in this study and cues available to the spiders in each
one.
Experimental conditions Cues available to the spider
Releases under a clear sky Linearly polarized light pattern
All eyes functional Light intensity gradient
Unintended landmarks
Releases under an overcast sky Unintended landmarks
All eyes functional
Releases under a clear sky filtered through plastic Elliptically polarized light pattern
All eyes functional Light intensity gradient
Unintended landmarks
Releases under a clear sky Linearly polarized light pattern
Only anterior median eyes not functional Light intensity gradient
Unintended landmarks
Releases under a clear sky Linearly polarized light pattern
Only anterior median eyes functional Light intensity gradient
Unintended landmarks
A
666 THE JOURNAL OF ARACHNOLOGY
Wavelength(nm)
—
Figure 2. Transmission (%) characteristics of
the components ofthe plastic sheet. —
Figure 3. Electroretinogramofan arachnid.Ab-
breviations: ERG = electroretinogram; CPE = pho-
were given by an electronic flash and were toelectric cell response; A = Amplitude; L = La-
conveyed to the eyes by an optic fiber. A small tency.
device was placed between the tip ofthe fiber
and the eyes making it possible to insert one
of two polarizing sheets (Polaroid) and con^ each spider remained motionless for several
sequently to stimulate the eye by linearly po- minutes and then began to walk following a
larized light, the plane of polarization being linear path, sometimes after a turn to re-orient
either vertical or horizontal. A light flash was itself. When each spider had run 40 cm or
delivered to adapt the eye to light and after a more, it began to make nest-searching move-
variable lapse oftime named duration ofdark ments: legs I and II flexed towards the spider
adaptation, a second flash elicited therecorded body. These movements have also been ob-
ERG. The durations of dark adaptation were served in the terrarium when the spider was
1 s, 2 s, 5 s, 10 s, 20 s, 60 s, and 300 s. In returning to the burrow (the entrance ofwhich
arachnids, the full ERG is composed of two had been blocked by soil) and do not occur in
negative waves, (3 and 7, and a positive wave, other situations. These movements, when seen
8 (Fig. 3), in contrast to insects, in whichthere in the open field, indicated that the spider was
is a first positive wave, a, prior to p, 7, and in search of its burrow entrance. Figure 4
8 (Fouchard & Carricaburu 1972). For each shows the distribution ofmovement directions
ERG, the amplitude (between the isoelectric by spiders after open-field releases under a
line and the top ofthe p wave) and the latency clear sky. The spiders show a correct orien-
(between the stimulus and the top of the p tation towards the burrow direction (meanval-
wave) were measured: both were found to be ue for 30° NE burrow direction sample: 34° ±
highly dependent on dark-adaptation and the 5°, r = 0.98, n = 5, Rayleigh test: P < 0.001;
NW
hour of recording. mean value for the 300° burrow direction
Experiments were carried out on three an- sample: 290° ± 18°, r = 0.83, n = 21, Ray-
imals for 24 hours. Only one spider provided leigh test: P < 0.001), although abimodal dis-
a complete electroretinographic record series, tribution was observed, with some spiders
without any trouble. Results shown in Fig. 6 searching for the nest 180° away from it.
relate to this animal. TNhEergerowuapsaandsitghniefNicWantgdriofufper(eMnacredbieat-wWeaetnsotnh-e
RESULTS— Wheeler test, P < 0.01).
Behavioral experiments. Releases under Releases under an overcast sky: Some spi-
a clear sky: As a control, prior to the release ders began by making a systematic search in
of spiders in the open field, the homing be- the open field. This behavior consisted of cir-
havior of each individual was observed in the cular movements by the animals beginning at
terrarium in order to see if it could return the release point and increasing in radius with
home and was well-adapted to the burrow. time. From time to time the spider returned to
Only those spiders well-adapted to the burrow the release point. Such spiders were not in-
(i.e., those immediately entering it upon con- cluded in the analysis. Figure 4B shows that
tact with the first pair of legs) were used in the orientation ofthe rest ofthe spiders under
the open-field experiments. Upon release. an overcast sky was random (mean value for
ORTEGA-ESCOBAR & MUNOZ-CUEVAS—POLARIZED LIGHT DETECTION IN LYCOSA. 667
/ Releases underaplastic sheet. As indicated
in the Methods section, when sun light passes
through the plastic sheet, its polarization
changes from linear, which is its predominant
characteristic (Waterman 1981), to elliptical,
with a small modification ofintensity. Spiders
released in the open field in these conditions
either exhibited the behavior of systematic
search or headed in a random direction, as
they did under an overcast sky. Only the latter
individuals were considered in the analysis.
a=256° Fig. 4C shows that (mean value for 30° NE
rn==05.42 burrow direction sample: 270°, r 0.35, n =
NS 9, Rayleigh test: P — 0.342; mean value for
300° NW burrow direction sample: 332°, r =
0.16, w = 11, Rayleigh test: P = 0.710). There
was no significant difference between the NE
NW
group and the group (Mardia-Watson-
Wheeler test, P > 0.368).
Releases with AMEs blinded: When the
a=270° AMEs were blinded by black opaque paint, a
r=0.35 non-directed distribution was observed (Fig.
n=9 5A). In this case, spiders did not show the
NS
systematic search behavior described in Ex-
periments B and C. The behavior of these an-
imals was completely normal except that their
orientation was not to the burrow (mean value
for 30° NE burrow direction sample: 45°, r =
Figure4.—Directions followedbyindividual spi- 0.04, n = 6, RaNylWeigh test: P > 0.900; mean
d=erRselineatsheesoupnednerfiaeldcluenardesrkyd;ifBfer=entRecloenadsietisonusn.deAr v1a44l°u,e rfor=300.02°2, n =bu1r0r,owRaydilreeicgthiontessta:mPple=:
ashneeotv.erAcrarsotwsskyo;utCsid=e oRfeltehaesecsirculnedseproitnhte tpolawsatridc 0b.e5t7w4e).enThtheereNEwasgrnooupsiagnndifitchaentNdWiffegrreonucpe
the two possible directions forreturning to the nest, samples (Mardia-Watson-Wheeler test, NS).
according to the outbound paths: 30° (thick arrow) Releases with PMEs, PLEs, and ALEs
and 300° (thin arrow). Filled circles correspond to blinded: When all eyes, except the AMEs,
home directions of animals that should orient to were blinded (Fig. 5B), a clear orientation of
300°, while open circles correspond to home direc- the spiders to the burrow direction was ob-
tions ofanimals that should orientto 30°. The thick served (mean value for 30° NE burrow direc-
tfaohrrerottwhheiinsnsaaimrdpreloewthoeifncsiairndcielmetahilsestcthiheractmleeshaiosnutlvhdeecrtmeoteruardnnirtveoecct3ti0oo°;rn tNlieoiWnghsatmesptl:e:P3<1° 0±.050°1,;rm=ea0n.97v,alnue=fo5r, R3a0y0-°
direction for the sample of animals that should re- burrow direction sample: 298° ± 20°, r
turn to 300°. Abbreviations: a — mean vector di- = 0.90, n = Rayleigh test: P < 0.001).
rection; r = length ofthe mean vector; n = sample There was a significant difference between the
size. NE group andthe NW group (Mardia-Watson-
Wheeler test, P < 0.01). —
Electroretinographic analysis. The elec-
30° NE burrow direction sample: 256°, r == troretinographic responses of the AMEs were
0.42, n = 5, Rayleigh test: P = 0.440; mean very different, depending on the plane oflight
NW
value for 300° burrow direction sample: polarization. During the photophase, vertical
45°, r — 0.14, n = 12, Rayleigh test: P = polarization resulted in a much higher ERG
0.797). There was no significant difference amplitude than did a horizontal one (Fig. 6
between the samples (Mardia-Watson-Wheel- shows the change of amplitude with dark ad-
er test, NS). aptation). A maximum was reached at 20 s of
668 THE JOURNAL OF ARACHNOLOGY
a=45°
r=0.04
n=6
NS
—
Figure 6. Electroretinograms (ERG) under po-
a=31° larized light. Diurnal amplitudes (mV) of AME
r=0.97 ERGs, recorded for different times of dark adapta-
Pn=<5 tion. The 0 = the horizontal plane of polarization;
0.001 the • = the vertical plane ofpolarization.
tional criterion for dead reckoning [path
integration]”(p. 56). Our results show that L.
— tarentula is capable of homing by means of a
Figure 5. Directions followedbyindividual spi- mechanism which does not use familiar ref-
dAer=s= iRneltehaeseospeonf afnieilmdalwsitwhitdhifAfeMreEnst ebyleisndebdl.inBded=. erences and it is based on visual information
Releases of animals with PMEs, PLEs and ALEs as external reference to calculate home direc-
blinded. Symbols as in Fig. 4. tion. In experiments with Arctosa (Magni et
al. 1964; Papi 1955a, b), the animals weretyp-
ically placed in the center of a cylinder and
dark adaptation, and the amplitude did not their escape directions were registered. They
change for adaptation up to 5 minutes. Laten- display so-called zonal orientation that gen-
cies of ERG responses obtained for vertically erally is not considered as homing (Papi
(5 mn in daytime) or horizontally polarized 1992). In contrast, we have used a paradigm
light were similar, about 30 ms. ERGs ofother that has been used to testhoming throughpath
eyes (ALE, PME and PEE) were not signifi^ integration in arthropods and mammals
cantly different, whether light was polarized (Etienne et al. 1998). In this paradigm, the
or not. subjects followed an L-shapeddetourandthey
returned directly to the point of departure. In
DISCUSSION
our study, L. tarentula also shows the ability
Path integration is a route-based homing to shortcut the outward path when returning.
which allows the animals a straight return af- Generally, our spiders did not retrace the out-
ter a more or less winding outward trip (Papi ward path. As with other spiders (Gomer &
1992). It is a process that allows an animal to Claas 1985; Seyfarth et al. 1982), L. tarentula
deduce its position, in relation to a point of could use tacto-chemical information, visual,
departure, from its own movement. To achieve or idiothetic information to return home. Tac-
this the animal has to measure two compo- to-chemical information is used by lycosid
&
nents ofits outwardjourney: the direction and males to find females (reviewed by Tietjen
the distance. The first component can be mea- Rovner 1982), and virgin females of L. tar-
sured using external references to calculate entula leave silk threads placed several mil-
the directions followed or using internal ref- limeters over the substratum (unpubl. data)
erences such as centrally stored recordings of that can be used by males to recognize fe-
their own movements. Following Etienne et males’ sexual status and to find them. How-
al. (1998), “The ability to “home” irrespec- ever, in our study, this information was not
tive of familiar references from the environ- available to the animal because the spider was
ment remains the hallmark and safest opera- placed in an open field that did not contain
ORTEGA-ESCOBAR & MUNOZ-CUEVAS—POLARIZED LIGHT DETECTION IN LYCOSA. 669
this source of information. Idiothetic infor^ PMEs, PLEs, and ALEs blinded. Although in
mation was not used by L. tarentula in hom^ both samples we are at the low limit ofsample
ing because the spiders did not follow any size, we think that the results are not invali-
particular direction under an overcast sky. If dated because we have an equal sample size
L. tarentula used idiothetic information, we for the animals that should orient towards 30°
should have observed successful homing be- NE under an overcast sky or with the PMEs,
havior under an overcast sky, as in the idioth- PLEs, and ALEs blinded, and in the lattercase
etic orientation of blind Cupiennius salei the distribution is at random. This bimodal
(Keyserling 1877) (see Seyfarth et al. 1982) distribution has also been observed in the ce-
towards a prey from which ithas been chased. lestial orientation of sandhoppers (Ugolini et
In this latter ctenid spider it has been argued al. 1993) in the morning and at sunset.
that homing orientation could be based on The skylight polarization pattern is detected
non-visual cues, given its nocturnal habits. On by the anterior median eyes of L. tarentula.
the other hand in L. tarentula, a diurnal as InArctosa variana, a wandering ripicolous ly-
well as nocturnal spider, visual and non-visual cosid species, Magni et al. (1964) have shown
cues can be used for orientation. that polarized light detection is carried out by
Which celestial cues are used by L. taren- the AMEs and PMEs. The indirect eyes
tula to return home? The sun’s position is ex- (PMEs, PLEs and ALEs) have no role in lin-
cluded by the experimental conditions of our early-polarized light detection by L. tarentula,
study. Other cues could be the skylight pat- since a completely random distribution is ob-
terns of linearly“polarized light and the inten- served when the AMEs are blinded. This is in
sity gradient. L. tarentula may use these cues accordance with a morphological analysis of
since it has a good orientation towards the the AME retina (Kovoor et al. 1993) which
burrow under a blue sky, while it becomes has shown that, in ventral photoreceptors, suc-
disoriented under an overcast sky. Under this cessive lines ofrhabdoms are oriented orthog-
condition, the spider lacks information about onally to each other; such an arrangement was
the sun’s position and the skylightpolarization not observed in any other eye type of L. tar-
pattern. We obtained similar results when we
entula.
used the plastic sheet, which did not signifi- From the analysis of the ERGs, it can be
cantly modify the intensity gradient but did concluded that the anterior median eyes ofL.
cancel out the skylight pattern of linearly po- tarentula perceive polarized light well, which
larized light. So, we thinkthattherelevantcue is unlikely for the other eyes. This correlates
forhoming is the skylightpolarizationpattern. perfectly with the results ofbehavioral exper-
In Arctosa perita (Latreille 1799) Papi
iments. The change in the direction ofthe po-
(1955b) also found that this cue was more im-
larization plane produces a large change in the
portant for orientation than the light intensity ERG
response (amplitude) of the of the
gradient.
How could one explain the presence of a AMEs. During the photophase, Lycosa tar-
entula being inside the burrow in an almost
bimodal distribution under ablue sky and with
the PMEs, PLEs, and ALEs blinded? We sug- vertical position or outside the burrow with its
body axis parallel to the ground, both vertical
gestthat the time ofday at whichobservations
and horizontal planes ofpolarization will play
resulting in orientation opposite to the burrow
direction were made is the determining factor. a role in homing. Ongoing experiments will
These observations were carried out at the determine if a change in the direction of the
first hours of daylight (near sunrise) or when polarization plane induces a change in the
the sun was at its noon location. Under either homing direction taken by the spiders as it has
ofthese conditions, a symmetrical distribution been shown for other spiders like Agelena la-
of the e-vector patterns occurs around the so- byrinthica (Gomer & Claas 1985).
lar-antisolar meridian. So, the spiders tested at APPENDIX
these hours could be confused by the sym- I
metry ofthe e-vector pattern. This would also Suppose that ({>1, (|)2,. . .cj)„ are the directions
be the reason for the absence of a bimodal taken by n animals to return home, then to
distribution for the animals that should go to- calculate the mean angle and the angular de-
wards 30° NE under a blue sky and with the viation of the sample we proceed as follows:
670 THE JOURNAL OF ARACHNOLOGY
ACKNOWLEDGMENTS
X = -(cos + cos ^>2 + • ' • + cos #n) We wish to thank P. Carricaburu for kindly
obtaining the electroretinograms. We also
y = -n(sin <Fi + sin $2 + • • • + sin wish to thank J.-M. Cabrera (Fac. Sciences,
U.A.M.) for determining the polarization
and the mean angle of the sample will be characteristics ofthe plastic sheet and F. Jaque
(Fac. Sciences, U.A.M.) for determj.nieg the
== arctan^ if X > 0 terdagnesmJi.sS.sioRnovcnhaerractaenrdistaincsaonfoint.yWmoeusackrenfoewrle-e
forreviewing the original manuscript and sug-
# = 180° + arctanr if x < 0. gesting various changes.
X
LITERATURE CITED
To calculate the angular deviation we use Baccetti, B. & C. Bedini. 1964. Research on the
structure and physiology ofthe eyes ofa lycosid
180^
s — V2(l - r). spider. I.-Microscopic and ultramicroscopic
IT structure. Arch. Italiennes BioL, 102:97-122.
Batschelet, E. 1981. Circular Statistic in Biology.
The length of the mean vector, r, is a mea- Academic Press, London. 371 pp.
sure of the dispersion of the data and it is Carricabura, R, A. Munoz-Cuevas & J. Ortega-Es-
calculated with the following formula cobar. 1990. Electroretinography and circadian
rhythm in Lycosa tarentula (Araneae, Lycosi-
r ^ ^ Etideanen)e.,AAc.tSa.,ZoJo.l.BeFrleinen,icJa.,G1e9o0r:g6a3k-o6p7o.ulos & R.
Maurer. 1998. Role of dead reckoning in navi-
With F2, sample size, and r, length of mean gation, Pp. 54-68, In Spatial Representation in
vector, the Rayleigh test gives us the critical Animals. (S. Healy, ed.). Oxford Univ. Press,
level, P, that the parent population is random- Oxford. &
ly distributed. Fouchard, R. P. Carricabura. 1972. Analyse de
Felectroretinogramme de ITnsecte. Vision Re-
Mardia-Watson-Wheeler test: The purpose
search, 12:1-15.
of this test is to discover whether two inde- Corner, R 1958. Die optische und kinasthetische
pendent samples of«i and W2 observations dif- Orientierung derTrichterspinneAgelena labyrin-
fer significantly from each other. This test is thica, Z. Vergl. Physiol., 51:111-153.
also known as a uniform-score test because it Gomer, P. & B. Claas. 1985. Homingbehaviorand
uses only the order in which the observations orientation in the funnel-web spider, Agelena la-
ofboth samples are arranged. We pooled both byrinthica Clerck. Pp. 275-297, InNeurobiology
samples and we ranked the elements of one of Arachnids. (EG. Barth, ed.). Springer-Verlag,
casuralemapttlheee,8sfi=ozrese[xo36fa0mb7po(lt«ehits+haemp«slm2ea)]sllweahsntedroweneeW.jmWualentdicpalHlyj- KovB1oe9or9lr3i,.n.J,M,iAc.roMaunnaotzo-mCyueofvatshe&anJt.erOirotregmae-dEisacno'beayre.
and its possible relation to polarized light recep-
each rank by 8, transforming each initial value tion in Lycosa tarentula (Araneae, Lycosidae).
to pj. The resultant vector of the first sample Boll. ZooL, 60:367-375.
has the following components Magni, E, F. Papi, H.E. Savely & R Tongiorgi.
C, = X cos Pi, Si = 2 sin Pi o1f96t4h.e Reeyseesarocfhaolnyctohseidstrsupcitduerr.e aIIn.dTphheysrioolleogoyf
aVnCd^th+e lengatnhdoafstahetersetsuslttatainsttivcecwteorusiseRB, == Magtdniiofinf,.erAeErn,cthF.p,aiIPrtasapliio,efnnHee.ysEe.sBioiSnLa,vaes1lt0yr2o:&n1o2m3Ri-c1aT3lo6n.ogriioerngtia.-
R?. 1965. Research on the stracture and physiology
If n > 17, we use the quantity of the eyes of a lycosid spider. III. Electroreti-
nographic responses to polarized light. Arch. It-
R2
= 2(n - 1)—^ aliennes Biol., 103:146-158.
nin2 Melamed, J. & O. Trajillo-Cenoz. 1966. The fine
structure of the visual system of Lycosa (Ara-
which is approximately distributed as cM- neae, Lycosidae). Zeit. Zellf., 74:12-31.
squared with two degrees of freedom. Papi, E 1955a. Astronomische Orienterang bei der
ORTEGA-ESCOBAR & MUNOZ-CUEVAS—POLARIZED LIGHT DETECTION IN LYCOSA. 671
Wolfspinne Arctosa perita. Z, Vergl. Physiol., nisms and Ecological Significance. (P.N. Witt &
37:230-233. J.S. Rovner, eds.). Princeton Univ. Press, Prince-
Papi, F. 1955b. Ricerche suU’orientamento diArc- ton, New Jersey.
&
tosa perita (Latr.) (Araneae, Lycosidae). Pubbl. Ugolini, A., B. Laffort, C. Castellini G. Beugnon.
Staz. Zool. Napoli, 27:76-103. 1993. Celestial orientation and ultraviolet per-
Papi, F. 1992. General aspects. Pp. 1-18, In Ani- ception in Talitrus saltator. Ethol. Ecol. EvoL,
mal Homing. (F. Papi, ed.). Chapman & Hall, 5:489-499.
London, New York. Waterman, T. 1981. Polarization sensitivity. Pp.
Papi, F. & J. Syrjamaki. 1963. The sun-orientation 281-471, In Handbook of Sensory Physiology,
rhythm of wolf spiders at different latitudes. vol. VII/6B. (H,-J. Autrum, ed.). Springer-Ver-
Arch. Italiennes Biol., 101:59-77. lag, Berlin.
Schroer, W.-D. 1976. Polarisation sensitivity of Wehner, R. 1982. Himmelsnavigationbei Insekten.
rhabdomeric systems in the principal eyes ofthe Neurophysiologie und Verhalten. Neujahrsbl.
funnel spiderAgelena gracilens (Arachnida: Ar- Naturf. Ges. Zurich, 184:1-132.
aneae: Agelenidae). Entomol. Germanica, 3:88- Wehner, R. 1989. Thehymenopteran skylightcom-
92. pass: Matched filtering and parallel coding. J.
Schwind, R. 1984. Evidence for true polarization Exp. BioL, 146:63-85.
vision based on a two-channel analyser system Wehner, R. 1992. Arthropods. Pp. 45-144, In An-
in the eye ofthe waterbug, Notonecta glauca. J. imal Homing. (F. Papi, ed.). Champan & Hall,
SeyCfaormtph., EP.h-yAs.i,olR..A,Her1g5e4n:r5o3d-e5r7,. H. Ebbes & EG. WehLcnooemnrpd,aosnRs..:&AS.caRsoesssetlu.dy19in85.behTahveiobuerea’lsnceeulersotbiia-l
Barth. 1982. Idiothetic orientation of a wander-
ology. Fort. Zook, 31:11-53.
ing spider: compensation of detours and esti- Wehner, R. & S. Strasser. 1985. The POL area of
mates of goal distance. Behav. Ecol. SociobioL,
the honeybee’s eye: behavioural evidence. Phy-
11:139-148.
Tietjen, W.J. & J.S. Rovner. 1982. Chemical com- siol. Entomol., 10:337-349.
munication in lycosids and other spiders. Pp. Manuscript reveived 17 August 1998, revised 28
249-280, In Spider Communication: Mecha- May 1999.
