DISPERSAL BY UMMIDIA SPIDERLINGS (ARANEAE, CTENIZIDAE): ANCIENT ROOTS OF AERIAL WEBS AND ORIENTATION?

2006·5.3 MB·English

by  William G Eberhard

2005. The Journal of Arachnology 34:254-257
SHORT COMMUNICATION
DISPERSAL BY UMMIDIA SPIDERLINGS
(ARANEAE, CTENIZIDAE): ANCIENT ROOTS OF
AERIAL WEBS AND ORIENTATION?
William G. Eberhard: Smithsonian Tropical Research Institute, and Escuela de
Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica
Keywords; Mygalomorph ballooning behavior, orientation
It is well known that many araneomorph spiders headband magnifier. Some individual silk lines
disperse by ballooning (e.g. Decae 1987; Suter were extremely fine and difficult to see; checks for
1999), but similar dispersal abilities of mygalo- unseen lines were made by moving an objectwhere
morph spiders are much less well established. Pre- a line might have been, and noting whether this
vious publications are easily summarized. Themost movement produced tugs on the spider or nearby
complete observations are those of Coyle (1983, silk lines. A voucher(maturefemale) specimenwill
1985) of spiderlings of Sphodros sp. (prob. S. at- be deposited in the Museum of Comparative Zo-
lanticus Gertsch & Platnick 1980) (Atypidae), and ology, Cambridge, MA 02138, USA.
Ummidia sp. (Ctenizidae). Spiderlings ofboth spe- At about 09:00 on the cool rainy season morning
cies moved along bands of silk lines, and launched of 17 Oct. 2002 with only intermittent sunshineand
themselves into the air after dangling at the ends of weak, erratic wind (a very fine, weak, intermittent
draglines. Other descriptions of mygalomorph bal- drizzle fell briefly at 12:30, but it did not rain until
looning did not provide details on how spiders took after 15:00), I noted a small ball of spiderlings at
to the air. Baerg (1928) carefully observed move- the tip of a long thin leaf growing at the edge of
ments of Ummidia carabivora (Atkinson 1886) the deck of my house (Figs. 1, 2). A file of spider-
(originally described in the genus Pachylomerus) lings toiled slowly up the edge of the leaf toward
spiderlings from their mother’s burrow along wide the ball. Below the ball of spiders, I traced a more
silk trails to elevated sites, but did not witness the or less straight trail of shiny silk downward along
spiders taking off. Enock (1885) saw that Atypus the edge ofthe leaf, across an open space ofabout
piceus (Sulzer 1776) spiderlings followed ascend- 10 cm to the rock wall supporting the deck, and
ing silken cords to upwardly projecting objects, then across this irregular surface (Fig. 3) for about
from which they were “blown off into midair . . . 70 cm to its end, the edge of the extremely well-
until they became attached to other sticks” (p. 394). camouflagedctenizidburrow lidcoveredwithgreen
Muma & Muma (1945) also observed silk bands moss (Fig. 5). In places the trail ran along the sur-
produced by the spiderlings of Sphodros rufipes face ofthe wall as a wide band, about the width of
(Latrielle 1829) (= Atypus bicolor Lucas); they a spiderling’s body. In others, where it bridged
stated that the spiderlings dispersed by ballooning, cracks and spaces of up to 7 cm between rocks, it
but gave no details. Cutler & Guarisco (1995) ob- narrowed to a single thick, aerial line (Fig. 3). Spi-
served a group of spiderlings ofS.fitchi Gertsch & derlings were scattered along this trail. The burrow
Platnick 1980 and apparent ballooning attempts at entrance was about 10 cm above the ground, in de-
the top of a small tree. Main (1957) suggested that tritus in a shallow indentation in the wall. Spider-
Conothelemalayana (Doleschall 1859) (Ctenizidae) lings were emerging one by one from underthe lid,
spiderlings balloon, but only on the basis of ob- which was hinged at the top and slightly open, and
serving large numbers of fine threads of silk pro- climbing up along the trail. Judging by their rates
duced by spiderlings held in collecting tubes. This ofmovement, they must have been emerging for at
note reports an observation of dispersal and bal- least 30 min previously. My tentative attemptto lift
looning by spiderlings of the another ctenizid, an the lid was answered by a sharp tug that closed and
undetermined species of Ummidia. held it tight, showing that the female resident was
Observations were made in San Antonio de Es- evidently holding theinnerside ofthe lid (seeBond
cazu, San Jose Province, Costa Rica (el. 1325 m; & Coyle 1995).
approximately 9° 51' N, 84° 10' W). Observations The silk trail alsocontinued in theotherdirection
with the naked eye were complemented using a 2x beyond the ball, A few flimsy lines radiated from
254
EBERHARD—BALLOONING IN UMMIDIA 255
deriings in the bail dispersed in the space of 10--15
min around 10:45. No obvious change in wind
strength or intensity was noted at this time, but the
wind was so light that a subtle change could have
gone unnoticed. The last spiderling was seen at
about 11:30, In each of five cases in which I ob-
served take-off behavior from the beginning, the
spiderling first descended on a dragline that was
attached to a horizontal aerial line, and then glided
smoothly upward and laterally, moving in the same
direction as the wind. The longest glide I was able
to follow took the spiderjust over 5-10 m until the
line it was on became entangled in a bush. Another
glide, in which I lost sight of the spider when it
m
v/as about 5 above the ground, probably tookthe
spider at least 10 m, judging by the direction in
which it was moving and the closest objects in that
direction.
Much of the spiderlings’ behavior near the ball
was apparently tentative. Spiders moved back and
forth on the more or less horizontal lines, and as-
cended and descended vertical draglines. Most de-
scents were followed by ascents of the same line
rather than by glides. The draglines were too thin
to be easily observed directly, and only occasion-
—
Figures 1--5. Dispersal by Ummidia sp. spider- ally, when lighting andbackgroundconditions were
lings. 1. aerial lines to post ofrailing; 2. spiderling appropriate, did I succeed in seeing the silk as the
(right) leaves a ball ofspiderlings to walk along an spider descended. Nevertheless, the spider’s move-
aerial portion of the trail; 3. relatively straight silk ments (slow descent straight downward with the
trail followed by spiderlings across uneven rock spider facing downward, with its legs more or less
face; v/here the trail was aerial, the band narrowed spread and moving little ifat all; legs nevermoving
to a single, thick thread (arrow); 4. horizontal por- as if walking along a line), left no doubt that they
tion of the trail of silk along the top of the railing were descending at the tips oflines they were pro-
(arrow indicates fork where some spiders went to ducing. The spinnerets were spread, atleastinsome
the right, others to the left); 5. the trail (single ver- cases, as the spider descended.
tical line, arrow) ends on an adult female’s burrow Spiders produceddrag lines as they walkedalong
lid, which is camouflaged with green moss. horizontal lines. Some spiders kept their spinnerets
spread as they walked, and in these cases it was
clearthatthe spiderproduced atleasttwolines, and
the tip of the leaf, forming aerial bridges to other probably more. When the spinnerets were directed
leaves 10-20 cm away, and a longer and stronger, rearward, these lines apparently merged into a sin-
approximately horizontal line about 30 cm long gle thread. Coyle (1985) reported that each spider-
connected the leaf tip to the post of the railing of ling of Ummidia sp. produced a band of numerous
the deck (Figs. 1,2). From here it ran horizontally fibers as it moved. In no case did I see a spider
along the railing for about 6 m (Fig. 4) until it went break and reel up a line as it walked along it (as is
straight up the corner of the house for about 4 m typical of araneoid spiders—Eberhard 1982, 1990;
totheunderside oftheroof, where severalthinlines Griswold et al. 1998). Nor did a spiderling ever
bridged to the underside ofthe eaves. Several lines slide tarsus IV along the dragline as it emerged, or
that streamed downwind from the eaves waved break and reel up aerial lines as it moved along
about in the light breeze as if their tips (which I them, other traits that are common in araneomorph
was not sure I could see) were free; the nearest spiders as they ascend draglines and produce span-
object in that direction was some 10 m away. It is ning lines (Eberhard 1986, 1987).
possible that these were lines which had broken I was not able to decipher with certainty how
when the spider was several m from this takeoff spiders initiated the horizontal aerial lines along
site. An estimated 50-100 spiderlings were seen at which they walked, or the lines with which they
different points along the trail during the next two ballooned. As in Sphodros and Ummidia (Coyle
hours, nearly all moving away from the burrow. 1983, 1985), ballooning was preceded by descent
The spiderlings dispersedaerially,bothfromnear on a dragline. But in no case did the spider give
the ball and from under the eaves. Most ofthe spi- any sign that this dragline broke as it glided up and
256 THE JOURNAL OF ARACHNOLOGY
away, as described in these other species (Coyle eses, as well as the dragline breaking hypothesis (F.
1983, 1985). The spider’s upward and lateral glid- Coyle pers. comm.). Resolution of this uncertainty
ing movement was observed carefully: it was very will unfortunately have to await further lucky oc-
smooth, and was not interrupted by any perceptible casions when ballooningbehaviorby mygalomorph
jerk that wouldbe produced when alinebroke. Nor spiders can be observed again. Perhaps the most
did the spider move its legs as if reeling in lines, useful technique to employ in such a situation
as occurs in some araneomorphs (Eberhard 1987). would be to lightly dust the lines with cornstarch
These details suggest that no lines v/ere broken. It or talcum powder, to make additional fine lines vis-
was as if the spider smoothly lengthened its drag- ible.
line while being pulled by another airborne length Several details of pre-ballooning dispersal by
of silk. mygalomorphs merit comment. In both genera that
One long (>1 m) horizontal line was established have been observed, the spiderlings migrate as a
by a ballooning spider and then used by several group from their mother’s burrow to the ballooning
other spiders, supporting the idea that at leastthese site, forming astrongbandofsilk(Baerg 1928, Fig.
early stages of flight during ballooning did not in- 3). Spiderlings ofthe theraphosid Brachypeima va-
volve breaking the drag line. The new horizontal gans (Ausserer 1875) also migrate in single file on
lineran injustthe direction in whichthe firstspider the ground, perhaps also following a band of silk
glided away several minutes previously. This line (Reichling 2000). Such mass movement, and the
was not present before this spider glided away, be- resultingformationofcompactaerialsilkhighways,
cause I had walked past this spot several minutes is very unusual in araneomorph spiders. I know of
earlier and would have broken any lines there. only one other case; the highways produced when
In a second case, a line was apparently formed colonies of the social theridiid Achaearanea wau
by one spider during the time it hung more or less Levi 1982 migrate (Lubin & Robinson 1982). The
motionless at the tip of a vertical line. The spider general pattern for dispersing araneomorphs seems
was first observed dangling atthe end ofadragline. to be for each spiderling to strike out on its own.
I passed my hands through the air at its sides and Spiderlings may benefit from moving as groups;
between the spider and my ov/n body without hav- following lines established by nest mates may fa-
ing any effect on its position, thus confirming that cilitate rapid movement to ballooning sites.
there were no unseen lines running from the spider The ability (and readiness) of Ummidia sp. spi-
or its dragline in these directions. Nevertheless, in derlings to walk upside down along aerial cables
the following minute, during which the spider re- (Figs. 1, 2) was surprising. Such dexterity in walk-
mained at the tip ofits dragline, a line was formed ing under aerial lines may thus be a very ancient
that connected the spider or the dragline near it to trait, and itcouldhavebeenimportantinfacilitating
mybody (perhaps 30 cmaway): eachtimeImoved, the evolution of aerial webs in other groups.
the spider was displaced. A minute or so later, this How did these mygalomorph spiderlings orient?
spider then glided smoothly away out of sight in a Perhaps a partial answer is related to a further re-
slightly different direction. markable detail ofthe highways: the trails are quite
These observations of Ummidia sp. ballooning extraordinarily straight (Figs. 3, 4), Baerg (1928),
are compatible with two different hypotheses re- who observed about 30 different trails of U. cara-
garding the initiation of ballooning lines: the spin- bivorus ranging from 10-68 feet long, also noted
neret spreading idea ofBlackwell (fig. 1C in Eber- that trails were “a straight line to the nearest tree
hard 1987); and the “second line” method (fig. 2 of considerable size. A tree less than 6 inches in
in Eberhard 1987). It is not clear whether my in- diameter is usually ignored, even ifit is much near-
ability to confirm the third “dragline breaking” er than some larger tree.” Coyle (pers. comm.) has
technique for initiating ballooning lines, which was also seen a straight 4 m trail of Ummidia across a
proposed by Coyle (1983, 1985) for other myga- grassy lawn to the base of a small holly tree. The
lomorphs, was due to limitations in the resolution trail of the theraphosid B. vagans may also be rel-
ofmy observations imposedbymygeneralinability atively straight; the text description says it “snaked
to see the lines the spiderlings were producing, to its way” along aroad, but the accompanyingphoto
my inability to follow spiders for longer distances shows a straight line of spiderlings (Reichling
(perhaps they break their draglines after having 2000). If one makes the apparently reasonable as-
moved several meters through the air), to differenc- sumption that the path of a trail reflects the path
es between species in the process ofballooning, or followed by the first spiderlings to emerge from the
to imprecisions in previous descriptions. If, as in maternal burrow, it seems likely that these animals
the observations reported here, the extraballooning must have used some sort of landmark orientation
lines (in addition to the dragline) were difficult to or a sun compass to maintain such straight trajec-
see in the Sphodros and Ummidia species observed tories when crossing irregular terrain, an ability
by Coyle, his observations are consistent with both documented in some araneomorphs (Corner 1973,
the spinneret-spreading and the second line hypoth- 1986; Tongiorgi 1959). Such an ability is surprising
EBERHARD—BALLOONING IN UMMIDIA 257
in spiders that probably seldom venture from their Eberhard, W.G. 1986. Trail line manipulation as a
burrows once they are established, andthat are gen- character for higher level spider taxonomy. Pp,
erally thought to depend largely on substrate vibra- 49-51. In Proceeding of the Ninth International
tions rather than visual stimuli to orient in other Congress ofArachnology, Panama. (W. G. Eber-
contexts (Coyle 1986). One possibility is that this hard, Y. D. Lubin & B. Robinson, eds.) Wash-
possibly ancient orientation ability may have ington. DC, Smithsonian Institution Press.
evolved to enable males to search more effectively Eberhard, W.G. 1987. How spiders initiate airborne
for females, instead of simply wandering randomly lines. Journal of Arachnology 15:1-9.
(Bell 1991). Male S. abboti Walckenaer search for Eberhard, W.G. 1990. Early stages oforb construc-
females during the day, and may orient visually to- tion by Philoponella, Leucauge, andNephila spi-
wardtree trunks (Coyle & Shear 1981). The search- ders (Araneae: Uloboridae and Araneidae). Jour-
ing behavior of mature male mygalomorphs (and nal of Arachnology 18:205-234.
for that matter, of mature male spiders in general) Enock, E 1885. The life-history ofAtypus piceus,
seems to be little known, and would repay further Sulz. Transactions of the Entomological Society
study. of London 1885:389-420.
I thankJason Bond foridentifyingthe spider, and Gorner, P. 1973. Beispiele einer Orientierung ohne
Fred Coyle and an anonymous reviewer for many richtende Aussenreize. Fortschritte der Zoologie
helpful comments on the ms. and references. 21:20-45.
LITERATURE CITED Gorner, P. 1986. Adjustment ofthe opticalreference
direction in the optical orientation ofthe funnel-
Baerg, W.J. 1928. Some studies ofatrapdoorspider web spider Agelena labyrinthica Clerck. Pp.
(Araneae: Aviculariidae). Entomological News 109-112. In Proceedings of the Ninth Interna-
39:1-4. tional Congress of Arachnology, Panama 1983
Bell, W. 1991. Searching Behaviour. New York, (W. G. Eberhard, Y. D. Lubin & B. Robinson,
Chapman and Hall, eds.) Washington, DC, Smithsonian Institution
Bond, J. & EA. Coyle. 1995. Observations on the Press.
natural history of an Ummidia trapdoor spider Griswold, C.E., J.A. Coddington, G. Hormiga & N.
from Costa Rica (Areneae, Ctenizidae). Journal Scharf. 1998. Phylogeny ofthe orb-web building
of Arachnology 23:157-164. spiders (Araneae, Orbiculariae: Deinopoidea, Ar-
Coyle, EA. 1983. Aerial dispersal by mygalomorph aneoidea). Zoological Journal ofthe Linnean So-
spiderlings (Araneae, Mygalomorphae). Journal ciety 123:1-99.
of Arachnology 11:283-295. Lubin, YD. & M.H. Robinson. 1982. Dispersal by
Coyle, EA. 1985. Ballooning behavior of Ummidia swarming in a social spider. Science 216:319-
spiderlings (Araneae, Ctenizidae). Journal ofAr- 321.
achnology 13:137-138. Main, B. 1957. Occurrence of the trap-door spider
Coyle, EA. 1986. The role of silk in prey capture Conothele malayana (Doleschall) in Australia
by nonaraneaomorph spiders. Pp. 269-305. In (Mygalomorphae: Ctenizidae). Western Austra-
Spiders, Webs, Behavior, and Evolution. (W. A. lian Naturalist 5:209-216.
Shear, ed.) Stanford, CA, Stanford University Muma, M.H. & K.E. Muma. 1945. Biological notes
Press. on Atypus bicolor Lucas. Entomological News
Coyle, EA. & W.A. Shear. 1981. Observations on 56:122-126.
the natural history of Sphodros abboti and Reichling, S.B. 2000. Group dispersal in juvenile
Sphodros rufipes (Araneae, Atypidae) with evi- Brachypelma vagans (Araneae, Theraphosidae).
dence for a contact sex pheromone. Journal of Journal of Arachnology 28:248-250.
Arachnology 9:317-326. Suter, R.B. 1999. An aerial lottery: the physics of
&
Cutler, B. H. Guarisco. 1995. Dispersal aggre- ballooning in a chaotic atmosphere. Journal of
gation of Sphodros fitchi (Araneae, Atypidae). Arachnology 27:281-293.
Journal of Arachnology 23:205-206. Tongiori, P. 1959. Effects of the reversal of the
Decae, A.E. 1987. Dispersal: ballooning and other rhythmofnocturnal illumination on astronomical
mechanisms. Pp. 348-356 In Ecophysiology of orientation and diurnal ac—tivity in Arctosa vari-
Spiders (W. Nentwig, ed.) Berlin, Springer-Ver- ana C. L. Koch (Araneae Lycosidae). Archivo
lag. Italiano de Biologia 97:251-265.
Eberhard, W.G. 1982. Behavioral characters for the
higher classification of orb-weaving spiders. Manuscript received 9 September 2003, revised 15
Evolution 36:1067-1095. March 2004.