Changes in web-monitoring forces associated with web reduction in ...

1028

Changes in web-monitoring forces associatedwith web reduction in the spider family Uloboridae BneNr D. Opell Departmentof Biologv, Virginia Polvtec'hnicInstitute ond State University,Blacksburg, VA, U.S.A. 24061 R e c e i v e dJ u l y 2 3 , 1 9 8 6 OnELL, B. D. 1987. Changesin web-monitoringforces associatedwith web reductionin the spider family Uloboridae. C a n .J . Z o o l . 6 5 : 1 0 2 8 - 1 0 3 4 . Membersof the genera Uloborus, H1'ptiotes,and Miagrammopeshave similar web-monitoringpostures,but very different websand tacticsfor monitoring them. Orb weaversof the genus Uloborusconstructhorizontalwebsand hang from their hubs, whereasreduced-webuloborids constructvertical webs and monitor them from a single thread. To determineif changesin spiderstrengthaccompaniedweb reduction, resting and maximum force measurementswere taken of a developmentalseries of Hy'ptiotes cavatus, Uloborus glomosus, Miagrammope.sanimotus, Miagrammopes pinopus, and an undescribed Costa Rican Miagrammopesusing a glass needle strain gauge. Both carapacelength and spider weight were used as indexesof spider size. Regressionanalysesof forces show that H. cavatusexertsthe greatestrelative force and Miagrammapesspecies the least. This is consistent with requirementsfor the operation of each web type: Hyptiotes cavatus tenses its entire triangularweb and suddenlyreleasesthis tensionwhen a prey strikesits web, whereasa Miagrammopesspeciesjerks a single threadthat hascaptureda prey. Within the genusMiagrammopes,the specieswith the most highly modifiedcarapaceexpressed the greatestresting force. OPELL,B. D. 1987. Changes in web-monitoring forces associatedwith web reduction in the spider family Uloboridae. C a n .J . Z o o l . 6 5 : 1 0 2 8 - 1 0 3 4 . Les espdcesdes genresUloborus, Ht'ptioteset Miagrammopesont desposturessemblableslors de l'inspectionde leur toile, mais leurs toiles et les tactiquesqu'elles emploient pour en faire I'inspection sont trds diffdrentes.Les tisseusesde toiles circulairesdu genre Uloborus construisentdes toiles horizontaleset restentsuspendues au milieu; en revanche,les Uloboridae qui construisentdes toiles plus petitesles fabriquente la verticaleet les surveillentd partir d'un fil unique. L'utilisation d'un indicateur d'effort d aiguille de verre a permis de mesurer la tension au repos et la force maximale chez des s6ries en ddveloppement d'H)'ptiotes c'avatus, Uloborus glomosus, Miagrammopes animotus, Miagrammopespinopus et une espdce encore in6dite de Miagrammopesde Costa Rica, mesuresrelev6esdans le but de d6terminersi la r6duction de la toile est reli6e d un changementde force. La longueur de la carapaceet la massede I'araigndeservaientd'indices de la taille de I'araign6e.Les analysesde rdgressiondes forces indiquentque c'est H. cavatusqui exercela plus grandeforce relative et les espdcesde Miagrammopes,lesforcesles moins grandes.Cesdonndessontcompatiblesavecles forcesrequisespour I'opdration de chaquetype de toile : Hyptiotes cavatusexerce une tension sur la totalit6 de sa toile triangulaireet relAchebrusquement cettetensionlorsqu'uneproie frappe la toile, alors que les espdcesde Miagrammopesne secouentque le fil qui a capturdune proie. Chez les espdcesdu genre Miagrammopes, celle qui possddela carapacela plus modifi6eest celle qui conservela plus grandetensionde repos. [Traduit par la revue]

Introduction In the spider family Uloboridae, web form correlateswith carapaceshape and the position, size, and number of eyes (Opell 1984a).This associationis explainedby changesin webmonitoring tactics that accompanyreduction of the family's primitive orb web (Eberhard 1912; Muma and Gertsch 1964; Opell 1979)to one of severaltypes of simpler webs. Unlike orb-weavinguloboridsthat typically hang from the hubsof their webswhile waiting for prey to contacttheir snares(Opell and Eberhard 1983), reduced-webuloborids monitor their webs nearan attachmentpoint of one of the web's taut anchorthreads (Fig. l). Here they extend their first legs directly forward and more actively monitor and manipulatetheir webs, and, in at least one case, exert greater force than orb weavers (Opell 1985).To facilitatetheseposturaland force changes,prosomal muscles,particularlythe extrinsicmusclesof the first legs,and the endosterniteplate to which they attachhavebeenreoriented (Opell 1984a). These appear to be primary changes that facilitateboth the effective paraxial movementof the legs and the exertionof force along the plane of the web's monitoring thread.Otherchangesin leg articlelengthsand leg musclesmay optimize either leg strengthor the total amount of leg movement.Externally,this prosomalreorganizationis expressedin modifiedcarapacefeaturesof the reduced-webgeneraHyptiotes and Miagrammopes. Hvptiotes speciesspin triangle webs (Opell 1982a; Wilder

1875;Peters1938)which consistof a monitoringthreadfrom which four taut radii diverge. The outer two radii attach to surroundingvegetationand are connectedby a frame threadthat completesthe triangle.The inner two radii attachdistallyto this frame thread. Evenly spaced, sticky (cribellar) prey-capture threads extend between the four radii. The carapace of Hyptiotesis broadand its four anterioreyesare small and offset from the anteriormargin of the carapace(Opell 1979, 1984a). The four posterior eyes are large and the posterior lateralsare borneon prominenttubercles. Members of the genus Miagrammopes construct an even more reducedweb which shows a much less stereotypicform (Akerman 1932: Lubin et al. 1918: Lubin 1986). This web usually consistsof a roughly horizontal, nonsticky resting thread from which one or several vertical or diagonal preycapturethreadsextend(Fi gs. 1,2).In somecase s,capt ur esilk may be depositedalong the restingline. Even within a single species,the number, orientation,and lengthsof the capture threadsarequitevariable(Lubin 1986;Lubin et al. 1978;B. D. Opell, unpublishedobservations).Miagrammopeslacks anterior eyes and, like Hyptiotes, has large posterior eyes and prominentposterior lateral eye tubercles(Opell 1984b;Opell andC ushi ng1986). The relativeleg lengthsof reduced-webuloboridsalsodiffer from thoseof mostorb-weavingmembersof the family: the legs of Hyptiotes are proportionately shorter and the legs of

r029

OPELL

Frcs. I atd2. Miagrammopes a imotus,Fig.I . Femalemonitoringa capturewebin a typicalcrypticposture,Scalebar: I cm.Fig.2. Thesame femaleresponding to a preyon the vefticalcaptureline. Scalebar : 0.5 cm. Miagrammopes,proportionately longer. These differencesare consistentwith requirementsfor the use of eachtype of reduced web. After constructinga triangle web, a Hyptiotes positions itself nearthe attachmentpoint of the monitoring thread, where it pulls in and holds slack threadequal in length to three to four body lengths(Opell 1982a).The spider periodically flexesits first legs as it monitors its web. When a prey strikes the web, Hyptiotes instantaneouslyreleasesits grip on the apex line, causingthe web to jerk as the slack thread previously held betweenthe first and fourth legs is taken up. By contrast,thecapturelinesof a Miagrammopesweb extend at less regular and more oblique angles from the typically horizontalmonitoring line, a situation that limits the spider's ability to greatlyalterthetensionsof thesecapturelines. When a Miagrammopes detects web vibration, it moves from the monitoringsiteonto its web, beginstugging on the captureline from which thesevibrationsemanate(Fig.2), andthenviolently jerks the web, sendinga wavelike loop down it (Lubin et al. 1978).This prey-capturebehavior appearsto be enhancedby the longer legs of these spiders, which permit greater distal displacementwhen they are flexed. Because the legs of Miagrammopesdo not have distinctly greaterrelative diameters, this increaseddisplacementis probably gained at the expenseof mechanicaladvantage(strength). An earlierstudy(Opell 1985)supportedthesepredictionsby showing that when it hangs from a single, horizontal resting thread,the triangle-webspiderHyptiotes cavatus(Hentz) exerts greater force throughout development than does (Jloborus glomosus(Walckenaer),an orb weaver with a similar webmonitoringposture(Opell and Eberhard1983).The purposeof this study is to test further the hypothesizedfunctional links betweenweb form, carapacefeatures,andleg lengths.This was done by measuringthe forces exerted by U. glomosus and H. cavatusand comparingthem with values of three Miagrammopesspecies,representingthe two subgroupsof the genus (Opell 1984b). Becausethese Miagrammopes subgroupsare characterizedby differencesin the degreeof carapacemodification, this study also determinesif subsequentprosomal modificationwithin this genusis associatedwith an increasein expressed force. Additionally, it comparesthe maximum force

expressed by each species and, unlike my earlier study (Opell 1985) using laboratory-reared H. cavatus, emplc.ys exclusively measurements taken from natural populations. Methods and materials All field studieswere conductedduring 1985. I collected Uloborus glomosuson the Virginia PolytechnicInstitute campusduring May and June and Hyptiotes cavatus from Giles and Montgomery counties, Virginia, from May through August. I studied Miagramrnopes animotus Chickering at the El Verde field station of the Center for Energy and Environment Researchin the Luquillo National Forest of PuertoRico during Januaryand February and Miagrammopespinopus Chickeringat the Virgin IslandsEcologicalResearchStation,St. John, U.S. Virgin lslands,during February.During late Juneand early July, I studied an undescrrbedMiagrammopes speciesat the La Selva field stationof the Organization for Tropical Studies, near Puerto Viejo de Sarapiqui,Heredia Province, Costa Rica. This latter Miagrammopes speciesbelongsto a different subgroupand hasa more highly modified, shortercarapacewith more prominent posterior lateral eye tubercles, and more widely spacedand posteriorly displacedposteriormedian eyes (Opell 1984b). Adult male (sixth instar) uloborids and spiderlingsnewly emerged from egg sacs(secondinstars)either do not construct capture webs or constructatypical capture webs (Eberhard 1976; Lubin et al. 1978; Opell 1982a, 1982b). For this reason, I used only third- to fifth-instar juveniles and mature (sixth-instar)femalesin this study. Only one setof resting and maximum force measurementswas taken from each Miagrammopes individual. Uloborus glomosus were collectedfrom a populationof approximately200 individualsliving on campusshrubbery and H. cavatus were collected from three hemlock groveseach with a populationof about 80 individuals. Samplingand releaseof theselattertwo specieswas donesystematicallyto insurethat an individual was not measuredmore than once in a given stadium, althoughthis methoddid not precludean individual's being usedmore than once during the courseof this study. After collecting specimens,I measuredtheir carapaceand first femur lengthsusinga dissectingmicroscopeequippedwith a micrometerand weighed them on a Cahn millibalance. Specimenswere kept individually in small, cotton-stopperedglassvials for I to 3 days before their forceswere measured.To preventdehydration,thesevials were placedin a plasticbag containinga pieceof moist cotton.Using a small brush, I induced each specimen to spin a resting thread between the fixed applicatorstick and the moveableglassneedleof a straingauge

c A N . J . Z O O L .V O L . 6 5 , 1 9 8 7

1030

o z o

a z -8.5

F

B" u Z -a ^

B -9.0

z U o (r

o r o

- 9.5 -10.0

L

z

> l

- d^ .- 5 -9.0

-10.5

F @ U c

o

o

-11.5

) )

-12.O

tr

-12-5

o

".;/

5 - ro.o

-11.0

I

.-" -,t . ---

o - 11.0

Hyptiotes cavatus M F = 4 . 7 O 1C a 6 2 0 x 1 0 - 1 5 r, 2 = 0 . 9 0

J

l F

5 -r1s

a -

ts -13.0 6.05

e

Uloborus glomosus M F = 2 . 1 4 4 C 2 . 8 6 0x 1 O - r 3 r, 2 . O . B O

z - 12.O

z

6.15

6.25

6.35

6.45

6.55

NATURAL LoG

6.65

6.75

6.85

6.95

7.O5 7.15

7.25

6.05

6.15

6.25

oF CARAPAoE LENGTH (pm)

6.35

6.45

6.55

NATURAL LoG

6.65

6.75

6.85

oF cARApAcE

6.95

retcrH

7.O5 7.15

7.25

7.35

(pm)

@

6 z - A

O

z o

q

]""

3 -n.o 3 U o {!

-9.5

u

-10.0

U Z

-e^

U (J

^-

o

o z - 10.5 ; tt U cc -i1.0 u o _ __ 1 1 . 5 o t

-12.o

I l

-*-.:..2

5 - ro.o

...,3: '

I

o . ---

E - 1 a -t ' " I F

a -

-

-1.25

-0.50

A.25

1.OO

NATURAL LOG OF wEIGHT

-11.o

J

Uloborus glomosus RF=2,314 Wo6e3 x 1O-s,r2.O.74

z -13.0

5

Hypttotes cavatus

R F = 3 . 0 7 1W o 6 6 6 x 1 O - 5 , r 2 . O . 8 2

J

1.75

2.50

z -12.O

-2.OO

9.25

(ms)

O

- 1. 2 5

-0.50

1.75 0.25 1.00 N A T U R A LL o G o F w E r c H r ( m s )

2.50

3.25

FIGS. 3-6. Comparison of the rcsting (Figs. 3 and 5) and maximum (Figs. 4 and 6) forces exerted by the orb we aver Uloborus glomosus and the triangleweaverl/)priotes caeatus^s afunction of carapacelength (Figs. 3 and 4) and weight (Figs. 5 and 6). RF, resting force; C, carapacelength; MF, maximum force. Tent-n |. Mean temperaturevalues ("C) at which force measurementswere taken, results of KolmogorovSmirnov D-testsof normality of temperaturevalues,and samplesizes Hyptiotes cavatus Mean temperature(SD) Probability of norrnal distribution Stadium sample size and mean temperature Third Fourth Fifth Sixth Total sample size

Uloborus glomosus

2 s . 2 ( 0 . 8 ) 24.4(0.7)

Miagrammopes sp.

Miagrammopes pinopus

Miagrammopes animotus

2 s . 2( t . s )

2 s . 3( 1 . 0 )

2 r . 8( 0 . 4 )

0.92

0.98

0.89

0.99

0.92

31,24.4 2 4 ,2 5 .5 53,25.6 4 2 ,2 5 . 2

2l,24.1 2l,24.2 13,24.0 4 5 ,2 5 . 0

68,24.8 4 9 ,2 5 . 6 26,25.5

48,25.4 3 3 ,2 5 . l 14,24.8 2 1 ,2 5 . 3

156

113

t70

123

20,21.7 2 7, 2 1 . 8 1 4 ,2 l . 7 3 7, 2 2 . 0 tt4

(Opell 1985).After the spiderhad assumeda typical resting posture, tuggedon the threadto evaluateand adjust its tension, and produceda stabledeflectionof the glassneedle,I recordedthe needle'sposition on the scale. This value was subsequentlyconverted to its milligram equivalent,multiplied by the acceleratingforce of gravity, and termed "restingforce." Next, I proddedthe spiderwith a small brush, causing it to increase the tension it exerted on the thread. I recorded the maximum deflection I could cause the spider to achieve and, after conversion,termed this "maximum force." Immediately after measuring thesevalues for a spider, I recorded the ambient temperature. To obtain the summary force values reported in Table 2, I assigned an instar value to each specimen. The carapaceor first femur lengths usedfor this purposewere establishedby noting discontinuitiesin the plotted size data and are reported in Table 2. This approach was

)1?5?

satisfactoryfor all speciesexcept U loborusglomosus. For this species, well-definedbreaks were not apparent and a cluster analysis (unweighted, pair group method using arithmetic averagesand squared distances)basedon abdomen,carapace,and first femur lengthswas usedto divide individualsinto groupsto which instarvalueswerethen assigned.

Results at For all speciesexceptM . animotus,the meantemperatures which forceswere measureddiffered by only 0.9"C (Table 1). The mean temperaturefor M. animotrrswas 2.6'C lower than that of the specieswith the next higher value. KolmogorovSmirnov D-tests showed that the temperaturesfor eachspecies

1031 -8.O

o z 3 U z U (r

-8.5

a z

- 9.0

F

B U z U (r o

- 9.5 -10.0

L

u

-10.5

o

-11.O

J

- t1.5

I I

J

J J

E l

z

- 12.O

l F

z

- 12.5 6

U. glomosus Resting Force

- 13.0 6.O5 6.15

6.25

6.35

6.45

NATURAL

6.55

LOG OF

6.65

6.75

cARAPAcE

6.85 leHcrx

6.95

7.O5 7.15

7.25

-o.50

7.35

0.25

1.00

N A T U R A LL o c o F w E r c H T ( m g )

(pm)

M. animotus

-8.O,

o z o

U) z I

B U z U (r

B

I

o o

3

(r f

-10.O,

-

-lo.5,

O o

) 1

a-

Max imum Force

MF.3.877 C2'sx o---

-n.r'

U o

I

asB

t

E z

- e.o,

Resting

1O-t2

Force

-11.0, - 1 1 . 5 .

(I

P -r2.o. z

RF=1.427 C2t2' x 1O-tr

-12.5. -13.0. -2.00

10

-0.50

- 1. 2 5

0.25

1.00

1.75

N A T U R A LL O G O F W E I G H T( m g )

lv!.pinopus

o z F

3

U z q

(r ts

9 J f

? z

1| 1|

-2.O0

6.05 6.15 6.25 6.35 6.45 6.55 6.65 6.75 6.85 6.95 7.O57.15 7.25 7.35

(pm) NATURAL LOGOF CARAPACE LENGTH Miagrammopes sp.

-1.25

-0.50

O.25 1.75 1.OO NATuRAL Loc oF wErGHr(ms)

12 Miagrammopes

2.5O

3.25

sp.

FIGS. 7- 12. Comparison of the resting and maximum forces exerted by Miagrammopes animotu.r (Figs. 7 and 8) , Miagrarrtnopes pirropus (Figs. 9 and lO), ard Miagrummopassp. (Figs. I I and 12) as a function of carapacelength (Figs. 7, 9, I l) and weight (Figs. 8, 10, l2). A regressionline oJ the restingforce of Uloborus glomosus(dotted line) has been addedto thesefigures so that they may be related to Figs. 3-6. RF, resting force; C. carapacelengthl MF. maximum forcc.

were reasonably normally distributed and a breakdown of temperaturesby instar showed that, within a species,mean valuesvaried no more than l.2oC acrossthe stadia(Table 1). For most speciesand sizeindexes(carapacelength and body weight),maximum force measurements had 12valuesthat were equal to or greaterthan those of resting force (Figs. 3-12). Mean maximum force was always greaterthan mean resting force(Table2). When /-testswere usedto compareregressions of the naturallogs of thesetwo force measurements, using the naturallog of carapacelengthasan index of spidersize,only the Costa Rican Miagrammopes speciesshowed no significant difference(0.10 < p I 0.05) betweenresting and maximum force. In U. glomosus and H. cavatus the y-intercepts of the

regressionlines were not significantly different, althoughthe slopeswere. When weight was used as an index of size, the maximum force of each speciesexceededits resting force, although rn U. glomosus, H. cavatus, and the Costa Rican Miagrammopes species,these differences were significant for the y interceptbut not the slope. Relativeto carapacelength, H. cavatusand Miagrammopes sp. had the greatestdifferentialsbetweenmaximum and resting forces,with maximum force being 2.9 and 3.0 times, respectively, that of restingforce. In the remainingspecies,maximum forcewas 1.8 to 2.1 times that of restingforce. Hyptiotes cavatus exerted greater force relative to both its carapacelengthand weight than the other speciesstudied(Figs.

r032

cAN.J. ZOOL.VOL.65.1987

Tnst-n 2. Carapaceand first femur lengths used to assign specimensto stadiaand the mean (and standarddeviation) weight and force values for each stadium Hyptiotes cavatus Third instar Carapacelength (pm) Femur I length (pm) Weight (mg) Restingforce (N) Maximum force (N) Fourth instar Carapacelength (pm) Femur I length (pm) Weight (mg) Restingforce (N) Maximum force (N) Fifth instar Carapacelength (pm) Femur I length (pm) Weight (mg) Restingforce (N) Maximum force (N) Sixth instar Carapacelength (pm) Femur I length (pm) Weight (mg) Restingforce (N) Maximum force (N)

440-560 0 .3 4(0 .0 9 ) 1 .2 2 1x rc -s 3 .5 0 9x l 0 -5 600-700 0 .9 8(0 .2 9 ) 2 . 7 0 1x l 0 - 5 7 .2 8 3x l 0 -5 800-960 2.4s(0.93) 8 .9 0 5x l 0 -5 2 . 5 6 5x l 0 - 4 1000-I 250 6 ,1 6(3 .0 6 ) 1 . 3 4 2x 1 0 - 4 4 . 5 4 3x l 0 - 4

Uloborus glomosus Miagrammopes sp.

Relativeweight :

Miagrammopesanimotus

780-n20 840-1420 r.t2 (0.s2) 3 . 5 4 tx 1 0 - s 7 . 3 7 3x l 0 - 5

532-980 0 . 3 6( 0 . 2 0 ) x l0-s 1.201 3 . 3 9 1x l 0 - 5

480-920 0 . 3 4( 0 . l 8 ) 1 . 2 8 5x l 0 - 5 2 . 8 7 1 xl O - s

520-1200 0. 54( 0. 33) 1 . 6 7 7x l 0 - 5 3 . 0 3 0x l 0 - 5

1000-l 300 1260-l 800 3 . 3 e( 1 . 0 0 ) 5 . 9 2 8x l 0 * 5 1 . 1 9 9x l 0 - a

9 8 8 -r 1 7 8 1 . 0 5( 0 . 3 2 ) 2 . 8 3 2x l 0 - 5 8 . 7 8 6x l 0 - 5

1000-I 350 0.99(0.40) 2 . 8 1 0x l o - s 7 . 2 4 4x l 0 - s

1240-1640 t . 2 3( 0 . 4 t ) 2 . 8 3 3x l 0 - s 6 . 5 2 5x l 0 - 5

1200-t420 1780-2200 4 . 7 1( 0 . 8 s ) 6 . 8 6 5x l 0 - s | . 6 6 6x t 0 - 4

1254-1520 1 . 5 3( 0 . 4 0 ) 3 . 8 4 9x l 0 - 5 1 . 1 2 5x l 0 - 4

1400-1900 2 . 0 4( 0 . 8 3 ) 4 . 3 6 4x l 0 - 5 | . 1 2 2x l 0 - a

t720-2040 3. 07( 0. e8) 5 . 1 8 3x l 0 5 1 . 2 7 6x l 0 - a

I 300-l 550 1750-2400 9 .9 3(4.6s) 1 . 0 7 2x l 0 - 4 2 . 5 5 5x l 0 - a

r7ro-2094 3 . 5 9( l . l l ) 6 . 6 t 9x l 0 - 5 1. 7 2 0x l 0 - a

1920-2720 3 . s s( l . l l ) 6 . 9 3 8x l 0 - 5 2.020x l 0-a

2160-2840 6. 39( 3. 36) 1 . 0 9 2x l 0 - s 1 . 8 2 5x l O - a

3-6, Table 3). Relativeto carapacelength, U. glomosrsexerted significantly more resting force than M. animotus and M. pinopus,but not thanMiagrammopessp. (Figs. 7-12, Table 3). Relativeto carapacelength, U. glomoszs exerted significantly more maximum force than the three Miagrammopes species. When first femur length was usedas an index of size of the three Miagrammopesspeciesstudied, results support those obtained for carapacelength: Miagrammopes sp. exerted significantly greater(p < 0.05) resting force than both M. pinopus and M. animotus, but there was no significant difference in the maximumforce expressedby the threeMiagrammopesspecies. Relativeto body weight, therewere no significantdifferences betweenthe resting and maximum forces expressedby U. glomosus and M. animotus. Both M. pinopus and Miagrammopessp. expressedsignificantly greater weight-specific restingand maximum forces than did U . glomosus. Becausea large part of a spider's weight resides in its abdomen,abdomenshapeinfluencesspider weight. t.ll

Miagrammopespinopus

Spiderweight (mg) Carapacelength (pm)

x 1000

When relative weight (eq. 1) is usedas an index. U, glomosus has a mban value of 4.30, which is significantlygreater(p < 0.0001)than that of both H. cavatus(2.93) and M. animotus (2.38). In turn, theselatter two specieshave relative weights thataresignificantlygreater(p < 0.0001)than thoseof both M. pinopus(1.23) and Miagrammopessp. (1.25), which do not differ significantly from one another (p > 0.79). The low relativeweight valuesof the latter two speciesresult from their morecylindricalabdomens.The superiorstrengthof H. cavatus is emphasizedby the fact that, even with only 68Vothe relative weightof U. glomosus,it expressessignificantlymore weightspecificresting and maximum force.

Resultsof this studyagreewith thoseof my earlierinvestigation (Opell 1985) by showing that, relative to both carapace length and body weight, H. cavatur expressedsignificantly more resting force than U . glomosus. To avoid problems resultingfrom differencesin relative leg lengths, the current study uses carapacelength rather than first femur length as a linear index of spidersize. However, when regressionsof first femur length (L) and resting force (F) are computed from the current data (eqs. 2 and 3), neither their slopes nor their interceptsdiffer significantly (p < 0.05) from those of the previousstudy when testedwith a two-tailed /-test. I2l

Uloborusglomosus: F : ( 6 . 8 7 8x 1 0 - t tL ) - ( 4 . 5 5 3x 1 0 - s ) N (r' :0.61)

t3l

Hyptiotes cavatus: p : ( 2 . 2 3 3x 1 0 - 7 D - 6 . 9 4 1 x l 0 - 5 ) N (f :0.73)

When regressionsof the natural log of spider weight and the natural log of resting force of these two studiesare compared, neither the slopes nor the intercepts of U. glomosas differ significantly. However, the intercept of H. cavatus differs significantly (p < the earlier study having a greatery-intercept. Mean relative first femur length (eq. 4) and its standard deviation were computed for five mature females of each speciesstudied: t4l

Relativefemur length :

Femur length (pm) Carapacelength (pm)

Listed in order of increasingvalues,the speciesare H. cavatus (0.71 + 0.08),Mi agrammopes sp. (1.29 -1_ 0.1 7) ,M . pinopus

1033

OPELL

Tesl-n 3. Resultsof one-tailedr-testscomparingregressionsof restingand maximum forces

Restingforce Carapacelength lntercept Slope Weight lntercept Slope Maximum force Carapacelength Intercept Slope Weight Intercept Slope

H. cavatusvs. U. glomosus

U. glomosusvs. M. animotus

+ x{0.0005.

(1 .3 1 -r 0. 27) , U. glom o s rz(1 s .3 8 -f 0 .1 3 ), a n d M. a n i motus (1.55 -r O.2l). Although thesevalues do not include the four distalleg articles,they suggestdifferencesin leg leverageand show that even within the genusMiagrammopes,mechanical advantagemay differ. Discussion Resultsof this study supportthe hypothesisthat, within the family Uloboridae,web reductionis accompaniedby changesin strengththat arecompatiblewith web monitoring and operation. As predictedby this hypothesis,the triangle weaver Hyptiotes cavatus expressesmore force than the orb weaver Uloborus glomosus. As a group, the three Miagrammopes species expressedless force than U. glomosus. However, the Miagrammopesspecieswith the most highly modified carapacedid not exert significantlyless resting force relative to carapace lengththan did U. glomosus. Although the resultsof this study agree with those of my earlier comparisonof the resting forces of H. cavatus and U. glomosus(Opell 1985), they show that my interpretationof those results overemphasizedthe role of prosomal changein increasingspiderstrength.If leg reorientationand the carapace modificationthat facilitatesit were the primary mechanismsfor increasinga spider's strength, then Miagrammopes species, with their extremely modified carapaces,should exert the greatestforce. Their expressionof the leastforce indicatesthat, aftereffectiveparaxialleg movementis established,changesin leg musculatureandleg articlelengthsareof greatimportancein altering the force a spider can exert. The relatively flatter prosomasof Miagrammopesspeciesmay result in their having smaller intrinsic leg muscles. However, as these muscles operateonly the coxa and trochanter,the differencesin their massesprobably do not explain the large force differences betweenHyptiotes and Miagrammopes. BecauseH. cavatusexpresses the greatestforce, mechanisms for reorientingits legs should be at least as well developedas thosern Miagrammopesspecies.Therefore,it is likely that the more extremecarapacemodificationsof this latter genusserve an additionalfunction. I haveneverobservedmuch retrolateral movementof H. cavatus first legs, although I have seen an individual of M. animotus move its extended first les both

almost directly dorsal to and almost directly lateral to its carapace to pull on oneof its capturethreads.Ratherthansimply enhancingparaxial movementof the first legs, the more extreme carapacemodifications of Miagrammopes may facilitate this greaterrangeof leg movements. Likewise, the greater carapace modification that distinguishesMiagrammopes sp. from the other two Miagrammopes of this study may not be responsible for the greater force expressedby this species.Other studies (Opell and Cushing 1986; B. D. Opell and A. D. Ware, manuscriptsubmittedfor publication) show that the larger eye tubercles of Miagrammope.r sp. contributeto its having a more extensivevisual surveillancethan the less highly modified M. animofus.This, ratherthan strengthenhancement,may be the function of larger eye tubercles.Changesin the lengthsof leg articlesand in leg musculaturemay contributeto the differencesin force observed within the genusMiagrammopes. Although all of the uloborids studiedused their first legs to monitorand adjustthe final tensionsof their restingthreads,11. cavatuswas the only specieswhose memberstypically used their fourth legs in the initial stagesof thread tensing. These observationsindicate that the forces measured in the other speciesof this study represent first leg strength and that the fourth legs of H. cavatus are probably also strongerthan those of the other speciesstudied. There is no clear anatomicalor behavioralevidencefor a "locking" or "catchmechanism"in the legs of any of the uloborids studied. In fact, the leg flexions which they exhibit in an apparenteffort to evaluateweb tension and vibration seem inconsistentwith the existenceof such a locking mechanism. "Resting force" probably approximatesthe sustainedwebmonitoring force of a spider, whereas"maximum force" -ay representthe limit of a spider's strength,expressedfor only short periods during prey capture. During this study, spiders seldommaintainedmaximum force for more than about 10 or l5 s afterthey wereproddedwith a brush,althoughimmediately after reducing this force, they often establisheda force that exceededtheir previouslymeasuredresting force. There are no definitive studies to evaluate the relative importance of uloborid web reduction in protecting these spidersfrom visually huntingpredators,making their websless

t034

cAN. J. ZOOL. VOL. 65. 1987

conspicuousto potential prey, and reducing the cost of constructinga prey-capturedevice. Lubin and Dorugl's (1982) study suggeststhat reduced webs may be more economical prey-capturedevicesthan orb webs. However, they found that, unlessa struggling prey is quickly subdued, it more easily escapesfrom a reduced web. Selection for such active and immediateweb manipulationwould favor the anatomicaland strengthchangesthat characterizereduced-webuloborids. The appearancesand web-monitoring postures of these spiders serve, even if only secondarily,to make them more cryptic (Figs. I and 2). Indeed, the principal adaptivevalue of some features,suchasthe first leg setaltufts and greencolor of some Miagrammopesspecies(8. D. Opell 1986), seems to be enhancedprotectiveresemblance.

L U B I N ,Y . D . , E s p R H e R oW , . G., andMoNrcolrsny, G. G. 1978. Webs of Miagrammopes(Araneae:Uloboridae)in the Neotropics. P s y c h e , 8 5l:- 2 3 . MUMA, M. H., and GeRrscs, W. J. 1964. The spider family Uloboridaein North America north of Mexico. Am. Mus. Novit. N o . 2 1 9 6 .p p . l - 4 3 . OpEt-I-,B. D. 1979. Revision of the genera and tropical American speciesof the spider family Uloboridae. Bull. Mus. Comp. Zool. 148:433-549. 1982a. Post-hatchingdevelopmentand web production of Hyptiotescavatus(Hentz) (Araneae:Uloboridae).J. Arachnol. l0: 1 8 5 -t 9 l . 1982b. Cribellum, calamistrum,and ventral comb ontogeny (Hentz) (Araneae: Uloboridae). Bull. Br. in H-vptiolesca\.,etLts Arachnol.Soc. 5: 338-343. 1984a. Comparison of the carapacefeatures in the family Uloboridae(Araneae).J. Arachnol.12: 105-114. Acknowledgments 1984b. Phylogeneticreview of the genus Miagrammopes Comments by William G. Eberhard helped improve this sensulato (Araneae:Uloboridae).J. Arachnol. 12:229-240. paper.National ScienceFoundationgrant No. BSR-8401919 1985. Web-monitoring forces exerted by orb-web and awardedto the author supportedthis study. triangle-webspidersof the family Uloboridae. Can. J. Zool. 63: 580-583. AKERMAN, C. 1932.On the spiderMiagrammopes sp. which con1986. The choice of web-monitoring sites by a green structsa single-line snare.Ann. NatalMus. 7: 131-143. Miagrammopes(Araneae:Uloboridae).Psyche.In press. EnnRHeRo, W. G. 1972.The web of Uloborusdiversus(Araneae: Oprlr, B. D., and CusHtNG,P. E. 1986. Visual fields of orb-web Uloboridae). J. Zool. 166:4ll-465. and single-line-web spiders of the family Uloboridae. Zoomor1976.The websof newly emergedUloborusdiversusand a phology (Berlin), 106: 199-204. maleUlobor&.f sp.(Araneae: Uloboridae). J. Arachnol.4:201-206. Opsll, B. D., and EssRHnRo,W. G. 1983. Resting posturesof LUBIN,Y. D. 1986.Web buildingand prey capturein Uloboridae. orb-weavinguloborid .spiders.J. Arachnol. ll: 369-316. 1nSpiders: webs,behavior,andevolution.Editedby W. A. Shear. PETERS, H. M. 1938. Uber das Netz der Drieckspinne, Hyptiotes Stanford UniversityPress,Stanford,California.pp. 132-171. paradoxus. Zool. Anz. l2l(3-4): 49-59. LuntN, Y. D., and DoRucr-,S. 1982. Effectiveness of single- WILDER,B. G. 1875.The trianglespider.Pop. Sci. Monthly, 1875 threadwebsasinsecttraps:stickytrapmodels.Bull. Br. Arachnol. ( A p r i l ) :1 - 1 5 .

Soc.5: 399-401.