Unformatted text preview: Behavioral Ecology Vol. 10 No. 5: 607–611 Prey attraction as a possible function of the
silk decoration of the uloborid spider
Department of Zoology, Faculty of Science, Kyoto University, Kyoto, 606-8502, Japan
Both laboratory experiments and ﬁeld observations were used to examine the prey-attraction hypothesis for the function of the
silk decoration on the orb web of Octonoba sybotides. The reﬂectance spectrum of the decorative silk showed that the decorations
reﬂect relatively more ultraviolet (UV) light. Choice experiments were conducted using Drosophila melanogaster, a common prey
species of the spider, to determine whether webs with silk decoration attract more ﬂies than undecorated webs. The choice
experiment showed that webs with silk decoration attract more ﬂies in light that includes UV rays. However, ﬂies choose their
ﬂight direction randomly in light without UV rays. This suggests that the silk decoration might attract prey insects that tend to
ﬂy toward UV-reﬂecting objects. Field observations comparing the prey capture rate between webs with and without a silk
decoration showed that more prey are caught in decorated webs. In this study, no difference between the two forms of silk
decoration, linear and spiral, was detected either in prey attraction in the choice experiment or in the prey capture rate in the
ﬁeld observations. Key words: choice experiment, Octonoba sybotides, prey attraction, prey capture rate, silk decoration, ultraviolet
light. [Behav Ecol 10:607–611 (1999)] W eb design is an essential part of the foraging strategy of
orb-web spiders. Although the orb web is made up of
three fundamental elements—radial threads, frame threads,
and catching spiral—some Araneid spiders are known to add
silky structures to the center of their orb webs (Eberhard,
1990; Nentwig and Heimer, 1987). The orb web represents a
behavioral and material investment in foraging by the spider
(Eberhard, 1986). The energetic return depends largely on
the prey-capture efﬁciency of the web, which is tightly related
to its design, including factors such as mesh size (width between spiral threads) and web size (area of the catching spiral) (Eberhard, 1986; Rypstra, 1982; Sandoval, 1994; Uetz et
al., 1978). Recent studies suggest that variation in orb web
structure is associated with an individual spider’s allocation of
energy to reproduction or prey capture (Higgins, 1990; Higgins and Buskirk, 1992; Sherman, 1994). However, the question of why spiders build web decorations, which must involve
energetic investment, has been debated for more than a century.
To date, various forms of the silk decoration, otherwise
known as stabilimenta, have been described, and several possible functions for these silky structures have been proposed.
These include adjusting web tension (Robinson and Robinson, 1970), strengthening the web (Robinson and Robinson,
1970, 1973), predator avoidance (Eberhard, 1973; Ewer, 1972;
Lubin, 1975), and advertising the presence of the web (Blackledge, 1998; Eisner and Nowicki, 1983; Horton, 1980). The
most recent and plausible hypothesis is the prey-attraction hypothesis (Craig and Bernard, 1990; Tso, 1996), which the authors examined in a study of orb-weaving spiders in the genus
Argiope. They also showed that the stabilimentum silk of some
Uloboridae species reﬂects more ultraviolet (UV) rays than
visible rays. They suggested that the reﬂected light might attract prey insects that tend to ﬂy toward UV-reﬂecting objects
Address correspondence to T. Watanabe. E-mail: [email protected]
Received 4 September 1998; revised 6 March 1999; accepted 12
1999 International Society for Behavioral Ecology and hence increase the prey-interception rate of the web. No
additional studies have tested the prey-attraction hypothesis as
an explanation of the function of the silk decoration on the
webs of uloborid spiders.
Octonoba sybotides is an uloborid spider that is widespread
in eastern Asia. This spider is known to add conspicuous,
white, linear or spiral silk decorations at the center of its orb
web (Watanabe, 1999; Yoshida, 1980). Although many O. sybotides webs have linear or spiral-shaped decorations at their
hubs, webs with a translucent disc sheet at the hub and webs
without silk decorations are also found in the ﬁeld (Watanabe,
unpublished data). Although the latter two types are rarer, all
four types can be observed in the same area. The function of
each type of silk decoration is not yet clear.
This study addressed two questions. (1) Do the stabilimenta
of O. sybotides attract prey? (2) Does the prey-attraction hypothesis account for the different forms of stabilimenta?
These questions were addressed with ﬁeld observations and
experiments. The prey-attraction hypothesis for the silk decoration on O. sybotides webs was examined by comparing the
prey-capture rate in the ﬁeld between webs with linear and
spiral silk decorations and those without silk decorations. A
choice experiment was also conducted to determine whether
Drosophila ﬂy to UV-reﬂecting silk decorations spun by O. sybotides.
Choice experiment in the laboratory
I collected mature female spiders from the Kyoto University
Botanical Garden in Kyoto, Japan. The spiders were kept individually in cylindrical cases (10 cm diam, 8 cm high) and
fed fruit ﬂies. Within a few days, these spiders spun horizontal
orb webs in their cases. I removed the silk decorations from
the webs with a pair of tweezers under a binocular microscope
and measured the reﬂectance spectra of the decorations alone
with a Shimadzu UV-visible recording spectrophotometer UV240, which measures wavelengths from 240 to 700 nm with an
0.3 nm. The light source for the spectropho- Behavioral Ecology Vol. 10 No. 5 608 periments. In the choice experiment, I used two types of light
to determine whether the UV-reﬂecting property attracts prey
insects; one beam contained short wavelength light (UV-plus;
without the UV ﬁlter), and the other was a control without
UV (UV-minus; eliminated UV component by the UV ﬁlter).
Field observations Figure 1
Apparatus used to test the preference of Drosophila. A ﬂy (D.
melanogaster) was put into the connecting tube from the entrance.
All the light entering the apparatus was shielded, except light from
the lighting tube connected to the upper side of each chamber,
into which rings with the central portion of webs were placed. tometer was dueterium, and barium sulfate (Merck) was used
for white standard.
I conducted a choice experiment similar to that of Craig
and Bernard (1990) to determine whether fruit ﬂies, a common natural prey insect of O. sybotides, are attracted to UVilluminated silk decorations. Two vinyl chloride cases, used for
keeping spiders, were connected with a T-shaped pipe (2.0
cm diam; Figure 1). The central portions of webs with the two
forms of silk decoration, or undecorated webs, were ﬁxed to
plastic rings 5 cm diam. The rings were placed in the cases at
the opposite ends of the T, at a 45 angle to the central connecting pipe. Light tubes (2.0 cm diameter) were connected
to the upper side of each case and focused on the centers of
the rings. I introduced a single fruit ﬂy (Drosophila melanogaster) through the central connecting pipe and then closed
the entrance. The species of Drosophila used for this experiment was a species commonly caught by the spider in the
ﬁeld. One ﬂy was introduced for each trial to avoid any effect
of interaction between ﬂies, and the same ﬂy was never used
again. The introduced ﬂy could see both webs from the junction of the pipes. In the next trial, the two rings with webs
were switched to eliminate site effects. After two trials, I replaced the two webs with two new webs. Each new web was
taken from a different spider. A total of 90 mature (cephalothorax length
1.5 mm) female spiders supplied the webs
(30 each linear, spiral, and no decoration) for the choice ex- To compare the prey capture rate between the different types
of webs, I conducted ﬁeld observations from late June until
late September 1998 in the Kyoto University Botanical Garden. O. sybotides constructs a horizontal orb web at the base
of trees, or between rocks, ﬂower pots, or piled logs in dimly
lit sites. The spiders tend to repair sections of their web, but
infrequently renew the whole web before dawn. Although the
spiders foraged both during the day and at night, they appear
to be diurnal foragers and to catch prey insects mainly during
the daytime (Watanabe, personal observation). Individual spiders build webs with linear or spiral decorations and sometimes build undecorated webs.
I observed the number of prey trapped on the webs or consumed by the spiders at 3- to 8-day intervals. Although I did
not identify individual spiders, no web was observed repeatedly because webs that had not been renewed since the previous observation could be distinguished by dust sticking to
the spiral thread of the webs. Because it could not be predicted when a spider would form one of the three types of
web, I randomly chose webs without silk decorations in the
study area. Then one web with a linear decoration and one
with a spiral decoration were surveyed within 1 m of each
undecorated web. In choosing webs, I excluded those with
repaired areas. Each trio of webs, an undecorated web and
webs with linear and spiral decorations, was deﬁned as an
observational unit. If there were several fresh webs with decorations within 1 m of the undecorated web, the nearest ones
were chosen. No web within a unit was nearer to webs in other
units than to the webs in the same unit. Spiders within the
same unit seemed to be in similar environments.
I identiﬁed 16–21 units every observation period and counted the number of prey captured on each web between 1430
and 1600 h. The prey interception rate (number of insects
per web per census) was calculated. By using observational
units, in which spiders seemed to construct webs in similar
environments, site effects never biased the observations for a
particular type of web. Because the daily capture rate varied
greatly, I log transformed the data on prey-capture rate to
normalize their distribution. I used ANOVA to compare the
rate of prey capture between the three types of webs. Webs
with a translucent disc sheet at the hub were not included in
the analysis because of a small sample size and the difﬁculty
in distinguishing a typical disc sheet, which varied and sometimes included a linear or spiral silk decoration. Undigested
prey were removed from the webs and identiﬁed later in the
In the study period, I also estimated the frequency of each
type of O. sybotides decoration in the study area by counting
all the spiders (total length 4.0 mm; including mature and
immature spiders) with identiﬁable web types in a monitoring
area (8 1 m) in the Botanical Garden.
Because the structure of a web can affect how many insects
it intercepts, I also estimated the mesh size and catching area
of each web with reference to Sherman (1994). I measured
the major and minor diameter of the orb web (from one outermost spiral to the opposite outermost spiral) to estimate the
total web area, and also measured two cross-diameters of the
hub and surrounding nonsticky spiral zone. I calculated the
catching area of each web (total web area less the hub and
nonsticky spiral zone areas). I counted the number of spirals Watanabe • Silk decoration as prey attractant 609 Table 2
Prey capture rate per census for three types of webs Web types Mean capture
SD n With linear form decoration
With spiral form decoration
Without decoration 0.40
0.09 Data from 20 days are used. Daily sampling sizes varied from 16 to
21. Total number of webs was 362. Decorated webs intercepted
more prey than undecorated webs (ANOVA; F2, 57
.001) Figure 2
Normalized reﬂectance spectra of stabilimentum silk spun by
Octonoba sybotides. Each point shows the mean of ﬁve
measurements. in each web’s two major axes and then estimated the mean
The silk decorations reﬂected relatively more UV light (Figure
2). The reﬂectance curve was nearly ﬂat between wavelengths
of 300 and 550 nm, with a slight peak around 300 nm. The
reﬂectance decreased at wavelengths longer than 650 nm or
shorter than 250 nm.
The preference for decorated over undecorated was significantly different between UV-plus and UV-minus in experiments 1 and 2 (Fisher’s Exact test, p
The fruitﬂies were more likely to choose a web with a silk
decoration than an undecorated web under UV-plus light (Table 1); there was no signiﬁcant difference in their preference
for linear or spiral silk decorations (Table 1). Under UV-minus light, there was no signiﬁcant difference in the ﬂies’
choice of webs.
The data from a total of 20 days were used for the ANOVA.
The daily sample size (number of webs per census) of each
type of web ranged from 16 to 21. Identiﬁable prey included
midges (39.2%), mosquitoes (6.8%), crane ﬂies (5.0%), other
small (total length
4 mm) and large (total length 4 mm)
ﬂies (42.7%), and other insects (6.3%) including Ephemerop- 143).
tera spp., Neuroptera spp., and Hymenoptera spp. (n
The mean prey-capture rate at the webs with decorations was
much higher than at undecorated webs (Table 2). There was
a signiﬁcant difference in the mean prey-capture rate between
the three web types (ANOVA; F2, 57
.001). Multiple comparison showed that (1) the prey-capture
rate of the webs with a silk decoration was signiﬁcantly greater
than that of the undecorated webs, and (2) there was no signiﬁcant difference in the capture rate between webs with linear and spiral silk decorations (post hoc Scheffe’s F test; linear
.001; spiral decoration
decoration versus no decoration, p
versus no decoration, p .001; linear decoration versus spiral
Although equal numbers of each type of web were observed
to compare the prey-capture rate, the daily frequency of decorations in the study area varied greatly. The total number of
webs in the monitoring area varied from 29 to 59 on 38 observation days. The mean frequency of undecorated webs was
much lower than that of decorated webs: linear decoration,
57.3 11.3% (mean SD); spiral decoration, 32.9 13.7%;
undecorated and others, 9.9
The mean mesh sizes and catching areas of each type of
web did not differ statistically in the 20 observations (all F
.15, for tests of between-day variation for
all web types). Therefore, data from the 20 days were combined to compare the features of the three web types (Figure
3). ANOVA and multiple comparison tests (Scheffe’s F)
showed that the mesh size of webs with spiral decorations was
signiﬁcantly smaller than that of the other two types of web
.0001). There was no statistical differ(F2, 1083
ence between the mesh size of the webs with a linear decoration and the undecorated webs. On the other hand, the
catching areas of the webs with the spiral decoration and the
undecorated webs were signiﬁcantly larger than that of the
webs with the linear decoration (F2, 1083
The results of the laboratory experiments and ﬁeld observations support the prey-attraction hypothesis for the function Table 1
Results of the choice experiments
No decoration vs. linear No decoration vs. spiral Linear vs. spiral Light type n Linear p n Spiral p Linear Spiral p UV
ns Under UV-plus light, Drosophila were attracted to the chamber in which the web with silk decoration
was placed more often than to the chamber with an undecorated web. Under UV-minus light, they
chose the chamber randomly. P values calculated from binomial distribution. 610 Figure 3
(A) Mean ( SD) mesh size (mm) and (B) catching area (m2) for
three types of O. sybotides webs. The results of the multiple
comparison test (Scheffe’s F) are shown in the graph: **p
.0001. of the silk decoration on the web of O. sybotides. Webs with a
silk decoration intercepted prey insects more frequently than
undecorated webs, although there was no difference in the
rate of prey intercepted by webs with different types of silk
In the ﬁeld, the prey species are mainly ﬂying dipteran insects. Some dipteran insects (mosquitoes, fruitﬂies, and drone
ﬂies) are known to have photoreceptors that are very sensitive
in the ultraviolet range (Bishop, 1974; Hu and Stark, 1977;
Muir et al., 1992). There are several explanations for this sensitivity. UV-reﬂecting objects may attract insects searching for
food resources, mates, oviposition sites, or escape routes. Insects might respond to UV light as a cue that indicates these
objects, although it is uncertain if a single mechanism attracts
prey to the web.
Craig and Bernard (1990) showed that the catching silks of
uloborid spiders also have a high UV reﬂectance (see also
Craig et al., 1994). This poses a problem for the prey-attracting hypothesis, because a UV-reﬂecting catching silk might
enable ﬂying insects to avoid the web more easily. However,
O. sybotides is found in the dim forest understory. Craig et al.
(1994) showed that uloborid spiders generally forage in a diurnal forest or nocturnal light environment. Such light conditions reduce the visibility of the catching silks, although in
dim light the effectiveness of prey attraction is also likely to Behavioral Ecology Vol. 10 No. 5 decrease. However, more light must be reﬂected from the silk
decoration than from the catching strings because the surface
area of the stabilimentum threads themselves, which are
densely arrayed, is much greater than the surface area of the
much thinner catching strings, even though the catching web
covers a larger area. Therefore, the reﬂectance of the silk decoration that attracts prey insects might decrease less than the
reﬂectance of the catching silk.
One alternative explanation for the differences in the preyinterception rate between webs is the effect of web structure.
Many studies suggest that mesh size, catching area, and the
visibility of the web affect the interception of insects by webs
(Craig, 1986; Rypstra, 1982; Sandoval, 1994; Uetz et al., 1978).
However, this hypothesis might be rejected in O. sybotides because the prey-capture rate between webs with linear and with
no decorations differed. Both had similar mesh sizes, but the
catching area of undecorated webs was signiﬁcantly larger
than that of webs with linear decorations. In spite of their
smaller size, the webs with linear decorations captured more
prey insects than the undecorated webs. In addition, webs
with a smaller mesh must be more visible than those with a
larger mesh, so that a ﬂying insect would be expected to avoid
a web with a dense mesh more easily. Therefore, the low preycapture rate of an undecorated web is unlikely to result from
a difference in the visibility of the web to prey insects. Tso
(1996) also argues this point.
If prey are attracted to decorated webs, why don’t all spiders
decorate their webs? Several researchers argue that energetic
constraints may limit decorating behavior (Eberhard, 1990;
Hauber, 1998; Herberstein et al., 1997). In the ﬁeld, however,
O. sybotides spun undecorated webs less frequently than those
with decorations, and in the laboratory O. sybotides that initially spun undecorated webs subsequently added spiral or linear decorations if they were not fed (Watanabe, personal observation). Therefore, undecorated webs might be part of the
construction process, and there may be an interlude before
the decoration is added. I cannot make any presumptions
about the energetic state of the spiders on the undecorated
In this study, the prey capture rate did not differ signiﬁcantly between webs with linear and spiral silk decorations.
However, the mesh of webs with linear silk decorations was
signiﬁcantly larger than that of webs with spiral silk decorations, whereas the catching area of webs with linear decorations was signiﬁcantly smaller than that of webs with spiral
decorations. These structural differences should affect the
quantity (number) and quality (size) of insects intercepted by
the web. The energy gain of a web combines the quantity and
quality of prey. The results of this study showed that there was
no numerical difference in prey interception. To further examine the effect of structural differences on prey interception, it is necessary to take prey size into account and measure
the qualitative difference in the intercepted prey. Further
studies are needed to clarify the relationship between the
form of silk decoration and the structural characteristics of
webs in order to elucidate the functional signiﬁcance of the
decoration. I am extremely grateful to M. Yoshida, T. Miyashita, T. Masumoto, and
the members of the Kansai Spider Study Group for encouraging me
during the early stages of my work. M. Imafuku kindly allowed me to
use the spectrophotometer. I am also grateful to K. Nakata, M. Urabe,
E. Honda, and S. Ishida for comments and English corrections in
drafts of this paper. Thanks are also extended to members of the
Laboratory of Animal Ecology, Kyoto University, for their assistance
and advice. Watanabe • Silk decoration as prey attractant REFERENCES
Bishop LG, 1974. An ultraviolet photoreceptor in a dipteran compound eye. J Comp Physiol 91:267–275.
Blackledge TA, 1998. Stabilimentum variation and foraging success in
Argiope aurantia and Argiope trifasciata (Araneae: Araneidae). J
Craig CL, 1986. Orb-web visibility: the inﬂuence of insect ﬂight behaviour and visual physiology on the evolution of web designs within the Araneoidea. Anim Behav 34:54–68.
Craig CL, Bernard GD, 1990. Insect attraction to ultraviolet-reﬂecting
spider webs and web decorations. Ecology 71:616–623.
Craig CL, Bernard GD, Coddington JA, 1994. Evolutionary shifts in
the spectral properties of spider silks. Evolution 48:287–296.
Eberhard WG, 1973. Stabilimenta on the webs of Uloborus diversus
(Araneae: Uloboridae) and other spiders. J Zool 171:367–384.
Eberhard WG, 1986. Effect of orb-web geometry on prey interception
and retention. In: Spiders: webs, behaviour, and evolution (Shear
WA, ed). Stanford, California: Stanford University Press; 70–100.
Eberhard WG, 1990. Function and phylogeny of spider webs. Annu
Rev Ecol Syst 21:341–372.
Eisner T, Nowicki S, 1983. Spider-web protection through visual advertisement: role of the stabilimentum. Science 219:185–187.
Ewer RF, 1972. The devices in the web of the West African spider
Argiope ﬂavipalpus. J Nat Hist 6:159–167.
Hauber ME, 1998. Web decoration and alternative foraging tactics of
the spider Argiope appensa. Ethol Ecol Evol 10:47–54.
Herberstein ME, Craig CL, Elgar MA, 1997. Optimal foraging behaviour: web investment and web decorations in Argiope keyserlingii
[abstract]. Adv Ethol 32:185.
Higgins LE, 1990. Variation in foraging investment during the intermolt interval and before egg-laying in the spider Nephila clavipes
(Araneae: Araneidae). J Insect Behav 3:773–783.
Higgins LE, Buskirk RE, 1992. A trap-building predator exhibiting
different tactics for different tactics for different aspects of foraging
behaviour. Anim Behav 44:485–499. 611 Horton CC, 1980. A defensive function for the stabilimenta of two
orb-weaving spiders (Araneae, Araneidae). Psyche 87:13–20.
Hu GG, Stark WS, 1977. Speciﬁc receptor input into spectral preference in Drosophila. J Comp Physiol 121:241–252.
Lubin YD, 1975. Stabilimenta and barrier webs in the orb webs of
Argiope argentata (Araneae, Araneidae) on Daphne and Santa Cruz
Islands, Galapagos. J Arachnol 2:119–126.
Muir LE, Thorne MJ, Kay BH, 1992. Aedes aegypti (Diptera: Culicidae)
vision: spectral sensitivity and other perceptual parameters of the
female eye. J Med Entomol 29:278–281.
Nentwig W, Heimer S, 1987. Ecological aspects of spider webs. In:
Ecophysiology of spiders (Nentwig W, ed). Berlin: Springer; 211–
Robinson MH, Robinson B, 1970. The stabilimentum of the orb web
spider, Argiope argentata: an improbable defence against predators.
Can Entomol 102:641–655.
Robinson MH, Robinson B, 1973. The stabilimenta of Nephila clavipes
and the origins of stabilimentum building in araneids. Psyche 80:
Rypstra AL, 1982. Building a better insect trap; an experimental investigation of prey capture in a variety of spider webs. Oecologia
Sandoval CP, 1994. Plasticity in web design in the spider Parawixia
bistriata: a response to variable prey type. Funct Ecol 8:701–707.
Sherman PM, 1994. The orb-web: an energetic and behavioral estimator of a spider’s dynamic foraging and reproductive strategies.
Anim Behav 48:19–34.
Tso IM, 1996. Stabilimentum of the garden spider Argiope trifasciata:
a possible prey attractant. Anim Behav 52:183–191.
Uetz GW, Johnson AD, Schemske DW, 1978. Web placement, web
structure, and prey capture in orb-weaving spiders. Bull Br Arachnol Soc 4:141–148.
Watanabe T, 1999. The inﬂuence of energetic state on the form of
stabilimentum built by Octonoba sybotides (Araneae: Uloboridae).
Yoshida H, 1980. Six Japanese species of the genera Octonoba and
Philoponella (Araneae: Uloboridae). Acta Arachnol 29:57–64. ...
View Full Document
- Spring '11
- Ecology, Spider, Spider web, webs, Argiope, Takeshi Watanabe, silk decoration