2003. The Journal of Arachnology 31:209-245 ON THE USE OF AMPULLATE GLAND SILKS BY WOLF SPIDERS (ARANEAE, LYCOSIDAE) FOR ATTACHING THE EGG SAC TO THE SPINNERETS AND A PROPOSAL FOR DEFINING NUBBINS AND TARTIPORES Mark A. Townley and Edward K. Tillinghast: Department of Zoology, Rudman Hall, University of New Hampshire, Durham, New Hampshire 03824 USA. E=mail: [email protected] ABSTRACT. The means by which female wolf spiders attach an egg sac to their spinnerets was inves- tigated using scanning electron microscopy. In fourPardosa species, we observed that silk fibersemerging from ampullate gland spigots had been affixed to the surface of the egg sac. More specifically, primary (1°) and secondary (2°) major ampullate (MaA) glands and 1° and 2° minor ampullate (MiA) glands all contributed fibers for this purpose. The diameters of the 2° MaA and 2° MiA fibers were greater than those ofthe 1° MaA and 1° MiA fibers and, correspondingly, the widths ofthe 2° ampullate spigots were clearly greater than those of the 1° ampullate spigots. Larger 2° ampullate spigots were also observed in adult females of species from three other lycosid genera. Thus, 2° ampullate glands, which in araneoids function only in juveniles during proecdysis, are not only functional in adult female lycosids (and adult females of several other families), but they appear to play a greater role than the 1° ampullate glands in egg sac attachment. Observations made on the 1° and 2° ampullate spigots of adult females from species belonging to several other families are also presented. Cuticular structures referred to as nubbins and tartipores are present in some spinning fields on spinnerets. A proposal is made for defining these terms by a criterion, namely their different origins, which differs from that applied previously. Keywords: Ampullate silk gland, Pardosa, Hogna, Trochosa, Lycosoidea With some exceptions, those spiders that securing the egg sac, which is positioned be- typically carry their egg sacs using only their low the sternum. At times, lycosids hold the spinnerets belong to one ofthe following three egg sac in a similar attitude; e.g., during the lycosoid taxa: the family Lycosidae, the fam- last phase of egg sac construction, when as- ily Trechaleidae, or the subfamily Rhoicininae sisting in spiderling emergence, or sometimes (with Shinobius considered a member of the when fleeing, after one has attempted to take latter taxon, Yaginuma 1991; Sierwald 1993). the egg sac away from the mother (e.g., Mont- The familial placement of Rhoicininae is un- gomery 1903; Lecaillon 1905). And, con- certain (Sierwald 1993; Carico 1993), but its versely, some pisaurids occasionally “drag the members are currently listed in Trechaleidae sac from the spinnerets alone in the same (see Platnick 2002). The exceptions referred manner as a lycosid” (Bishop 1924:28). to include other lycosoids (sensu Griswold et As described in Carico (1993), Sierwald al. 1999) (e.g., the ctenid Cupiennius, Barth (1990a, 1993), Scheffer (1905) and literature et al. 1991; Silva Davila in press) as well as cited by the first two authors, differences exist non-lycosoids (e.g., the nesticids Nesticus, among lycosids, trechaleids, rhoicinines, and Eidmannella, Nielsen 1932:201; Bristowe pisaurids with regard to egg sac structure and 1958:223; Potzsch 1963:30; the zorids Vor- the maternal care afforded post-emergent spi- aptus, Neoctenus, Lawrence 1964:34; Silva derlings. Typically, the egg sacs of lycosids Davila in press, though in the latter paper and rhoicinines are spherical or lenticular, Neoctenus is transferred to Trechaleidae). The those ofpisaurids are spherical, while those of spinnerets are also involved in carrying the trechaleids are hemispherical. A seamjoining egg sac in at least some genera within the ly- the upper and lower valves of the egg sac is cosoid family Pisauridae (see Discussion), but apparent among lycosids, trechaleids, and here the chelicerae play the principal role in some rhoicinines {Shinobius), but not in some 209 210 THE JOURNAL OF ARACHNOLOGY other rhoicinines (Rhoicinus) or pisaurids, and egg sac to the spinnerets. The impetus for the only in trechaleids is a ‘skirt’ (Carico 1993) study can be explained as follows. Observa- produced at the seam. tions made on spinnerets by a number of Lycosid, trechaleid, and rhoicinine females workers indicated that a certain category of continue to carry their progeny for a number ampullate silk glands, what we call secondary of days after they have emerged from the egg (2°) ampullate silk glands, are functional in sac (lycosids about 2-14 days, e.g., McCook juvenile spiders of most, if not all, entelegyne 1884; Montgomery 1903:72,76,82,90; Engel- taxa. However, in only some entelegynes are hardt 1964:303,387; Trechalea about 17-19 2° ampullate glands also functional in adults days, Carico et al. 1985; Shinobius about 4 (sometimes only in the females, as in lycos- days, Kaihotsu 1988). But while lycosid spi- ids). The only role assigned to these silk derlings climb onto the mother’s abdomen glands that we were aware of when we began during this period, trechaleid spiderlings and this study is to produce silk fibers during one at least Shinobius spiderlings (among rhoici- specific period in the molt-intermolt cycle of nines) climb onto the outside of the egg sac, juveniles (detailed below). The question thus which the female continues to carry. However, arose, what are 2° ampullate silk glands used Yaginuma (1991), specifying twoArctosa and for when they are retained in adults? It oc- one Hygrolycosa species, reports that trans- curred to us that if these silk glands are in- port on the egg sac, rather than on the moth- volved in egg sac attachment in lycosoids, it er’s abdomen, also occurs in some lycosids. In might be a situation in which we could, in those instances in which trechaleid spiderlings effect, catch spiders in the act of drawing fi- have been observed on the mother’s abdomen, bers from these glands and, thus, demonstrate this appears to be due to crowding on the egg their use in at least one specific application in sac with consequent spill-over (Carico 1993), certain adult spiders. At that time we were not just as lycosid spiderlings may spill over onto aware that Carico (1993) had already ob- their mother’s cephalothorax. served certain 2° ampullate gland fibers being Among pisaurids (where known), the egg used for egg sac attachment to the spinnerets sac is carried by the female until shortly be- in trechaleids. As we describe below, trechal- fore the young emerge, or at latest when the eids and the lycosids that we have examined first spiderlings begin to emerge (Gertsch (primarily Pardosa) show similarities and dif- 1979:197). Typically, the egg sac is then se- ferences with respect to egg sac attachment, cured within a nursery web. The mother including in their use of 2° ampullate silks. In builds the network ofsilk fibers that constitute partial answer to the above question, all we the nursery web before and/or after the end of know at this time is that adult females of at the egg-sac-carrying period, often on vegeta- least some lycosid and trechaleid (Carico tion, with the spiderlings adding fibers after 1993) genera use silk from 2° ampullate their emergence. The post-emergent spider- glands to help secure the egg sac to the spin- lings remain within the nursery web, guarded nerets. by the mother (though see Montgomery 1909: To provide a better overall perspective on 556; Forster 1967:84), for a period of time 2° ampullate silk glands and their roles, the that again varies considerably, sometimes following section reviews different categories even within a species (e.g., Dolomedes fiin- ofampullate silk glands. It is followed by four briatus (Clerck 1757) spiderlings may remain sections dealing with nubbins and tartipores, in the nursery web from 3 or4 days (Bristowe protuberances present in some spinning fields. 1958:191) to about 3 weeks (Nielsen 1932: These sections are included because our in- 134)). After this period the spiderlings dis- terpretations of spinneret micrographs ob- perse. Kaihotsu (1988) reports that Shinobius tained during this study rely on and make ref- females, after carrying their young on the out- erence to these protuberances. And because side of the egg sac for a few days, then hang our basis for distinguishing nubbins from tar- the egg sac with spiderlings in a nursery web, tipores differs from that ofearlier authors, and where the spiderlings remain for about one also differs from our own earlier views, the day. first of these four sections explains how we This study is concerned with the specific arrived at our current definitions for nubbins silk glands used by lycosids for attaching the and tartipores. The last three sections review TOWNLEY & TILLINGHAST—EGG SAC ATTACHMENT IN LYCOSIDS 211 some aspects of the occurrence of these pro- ately after ecdysis (even as the spider is hang- tuberances. We hope this overview will be ing by its ‘molting threads’) until about the useful in light of the growing importance of beginning of the following proecdysis (the these structures in phylogenetic studies. In ad- few days preceding ecdysis during which in- dition to presenting observations made on the ternal changes take place in preparation for spinnerets and egg sacs of some female ly- ecdysis), as well as throughout adulthood cosids, this paper contains comparative obser- (Townley et al. 1993). During proecdysis vations made on the spinnerets of male lycos- these glands are remodeled, rendering them ids and the spinnerets of spiders from several temporarily nonfunctional (Townley et al. other families in which adult females retain 1991). The task of producing ampuilate silk apparently functional 2° ampuilate glands. during each proecdysis is taken over by one Categorie—s of ampuilate silk glands and of the two pairs of 2° MaA glands and one of their roles. The ampuilate silk glands of the two pairs of 2° MiA glands. Each pair of spiders within the Orbiculariae and some oth- 2° ampuilate glands cycles through growth er taxa (e.g., Hersiliidae, Kovoor 1984; Se- and regression phases, reaching maximum gestriidae, Clubionidae, Gnaphosidae, Thom- size and accumulation of luminal contents at isidae, Kovoor 1987; Oxyopidae, Kovoor & proecdyses in every other juvenile stadium, Muiioz-Cuevas 1998) can be divided, on the with one pair of 2° MaA/2° MiA glands pro- basis of histochemical differences, into major ducing silk during proecdysis in even-num- ampuilate silk (MaA) glands and minor am- bered stadia and the other pair functioning in p1u9i7l7a,te1s9i8l7k;(aMMlisaAo)AKgolvaonodrs (&rePveiteewresd1i9n88K)o.vTohoer oBedcda-unsuemboenrleydonsetaodifath(eTtowwnolpeayirsetofal.2°1M9a93A)/. ducts of the glands connect to spigots 2° MiA glands produces silk in a given juve- located on the anterior lateral spinnerets (ALS) while MiA gland ducts connectto spig- insiloensltyadoinuem2(°i.eM.,aiAs ‘soppiegno’t, soeneeTaacbhleA1L),Sthaenrde ots on the posterior median spinnerets (PMS). one 2° MiA spigot on each PMS ofjuveniles In some other spiders, however, including the (in addition to the single 1° MaA spigot and Lycosidae, histochemical differences are not single 1° MiA spigot on each ALS and PMS, readily apparent between those ampuilate glands with ducts that empty on the ALS ver- respectively). After the final molt, with no ad- sus those with ducts emptying on the PMS ditional proecdyses to pass through, both pairs of 2° MaA/2° MiA glands degenerate (Seki- (Kovoor 1976, 1987). One may therefore question the validity of recognizing two dif- guchi 1955b; Townley et al. 1991). Thus, 2° ampuilate glands do not function in adults and ferent types of ampuilate glands in such taxa. only nonfunctional vestiges, termed nubbins Nevertheless, for clarity and in keeping with (Coddington 1989; Yu & Coddington 1990), precedents (e.g., Platnick et al. 1991:2), any of 2° ampuilate spigots are present on adult ampuilate glands with ducts attached to the ALS will be called MaA glands and any with ALS and PMS (Sekiguchi 1955b; Peters 1955; ducts attached to the PMS will be called MiA Mikulska 1966; Wilson 1969). This situation glands (Table 1). exists in both sexes. External examinations of In more basal araneoids, including those in spinnerets indicate that this description, out- the families Araneidae and Tetragnathidae lined in Table 1, applies not only to basal ar- (Griswold et al. 1998), both the MaA and aneoids, but to some non-araneoids (e.g., ox- MiA glands can be further subdivided into a yopids, Kovoor & Munoz-Cuevas 1998), and, single pair of primary (1°) MaA/MiA glands whatever their superfamilial placement, to and two pairs of 2° MaA/MiA glands (Table some mimetids (see data forMimetus in Table 1) (Townley et al. 1993; Tillinghast & Town- 3; see also mimetid spinneret micrographs in ley 1994). Observations made on spinnerets Platnick & Shadab 1993). Interestingly, it may (Coddington 1989; Forster et al. 1990; Peters be that the above description applies to some & Kovoor 1991; Hormiga 1994a,b, 2000; species within the mimetid genus Ero, but not Griswold et al. 1998) suggest a tendency for others (cf. figs. 29, 30 with figs. 41, 42 in the more derived araneoids to lack 2° MiA Platnick & Shadab 1993, noting especially the glands. The 1° MaA and 1° MiA glands func- presence of a MiA nubbin and MiA tartipore tion in each juvenile stadium from immedi- (tartipore defined below) in fig. 42 and the 11 212 THE JOURNAL OF ARACHNOLOGY on s ’5S-1 00 0) whether glands 00 Dh «« to ampullateprimarily51fJ^.r,t §p 4^^O^ cbSid) o^« r^^Ss .dd d«po openenviron- minor), .S g that made ag- op 3 A ^ outside vs. 2° s g S s sw m sM § o spigots S y^ 'S the meet (majorwhether observations toS bO OD 5aes 0o0 5 .cdadfl >P4) rtdehuceeciitvrse (a2n°)d),ibaossend T^JMC&«3rt .IPg^^wO_3J ^^-®fP«caia 5'^OOS5’ SwS33M ^3^g0.-gucp4p3d) s< CNi .adSoop5 ."MaPdS osaj|fWSt ^Oh ^Oi ttoohupotee-nenvironment secondary ""mi as o 8 8 § 2 i S that side scheme spinnerets vs. C p ’-* (1°)This s §§ 1 ©a o 1 «“ I S 3 3«o S3 i which 1-1® s I 2 3 (primary M of blocked). 2 g ° S 0 ^ 1^1 _ .E -Oon8 tbaohsines proecdysis(ovps.en.§c1;§«S'§ne,2»! MO MP^'3-Oa 1 •-d§gd3P 3dSP «PdoX d3 categoriesjduuverniilnegsstnaodoirutm ^o-'g^==^<3 .^bwCoJO^^00^P0i0)) ^g^«cSO ^3aS0«} ^ 3I.sa«g2.5 OJ r^3h .dad ^oo topoenenviron- f2 1 i .1 p « ^ 6 that idifnfetreont jiounsrtigianvenSbMO<Pfi -^^aP ^S“P 5^Sw| "ocdpIS« t3Mpd^ PaL©dQfc, ^^^ Op spigotstouthsiede ment glandsadults H s S § ^|c,a.S o& ^ 3 asmipililklatejauvnendilesenvironment .rc«•^3-O:i» S^-UsCO s^oMP^Io £OuHcocPn MM« s -dC IOops33-d o^fai r^W^5'f§mhVi _.‘dOddh ^-U4©o9) ttoohupotee-nenvironment of both extracorporeal ^^5 S 3-Ta 'g"d od that side in 200 S 3oo ^o“ division the 2p s '3 T3 i 3p o d8 s to 4) ad The functional 33 .3 T1ab.l—e glaarnedshouaatvlneet '-h§^iC;O H-idP .^«po§s ^S1poS.'BMs0f) •.gn^iS<w'8Oo4>sd)D> 223Mop ".-|pd"da.a TOWNLEY & TILLINGHAST—EGG SAC ATTACHMENT IN LYCOSIDS 213 absence of these protuberances in fig. 30; see tachments to spigots on the old cuticle during also Schiitt 2000:145). proecdysis, despite the formation of the inter- In contrast to the situationjust described for vening new cuticle (Fig. 1). Thus, while some basal araneoid taxa, examinations of spinner- protuberances do seem to be vestigial spigots, ets from spiders in certain amaurobioid (sensu others have a very different origin, being rem- Griswold et al. 1999) and dionychan (sensu nants of these openings. Coddington & Levi 1991) families, including After realizing that we were actually deal- the Lycosidae (Table 2), reveal a sexual di- ing with two different categories of cuticular morphism wherein males appear to conform protuberances, we made an ill-devised attempt to the above description, but females do not to both retain the original distinction between (Fig. 1). Instead, adult females retain appar- nubbins and tartipores (singulars versus mul- ently functional 2° MaA and 2° MiA spigots, tiples) and distinguish vestigial spigots from one pair of each, indicating that they use 2°, remnants of openings by use of the adjectives as well as 1°, ampullate glands as adults. In ‘vestigial-type’ and ‘non-vestigial-type’, re- trechaleids, Carico (1993) has observed the spectively (Townley et al. 1993). We soon use of 1° and 2° MiA silks by adult females abandoned this approach in favor of another, for securing the egg sac to the spinnerets. not previously published, that is concerned Here we report that adult female Pardosa use only with the two different origins of the pro- 1° and 2° MaA and 1° and 2° MiA silks to tuberances under consideration (forfurtherex- attach their egg sacs to their spinnerets. The planation see Townley 1993:7, 8). The latter 2° ampullate fibers have greater diameters approach, which we will follow in this paper, than the 1° ampullate fibers, indicating a great- retains the terms nubbin and tartipore, but de- er contribution from the 2° ampullate glands fined as follows: Nubbin: a nonfunctional, to the support of—the egg sac. only partially formed, i.e. vestigial, spigot, ei- Terminology. Nubbins and tartipores: ther morphologically singular or multiple. As mentioned, the term ‘nubbin’ has been ap- Tartipore: a cuticular scar, morphologically plied to cuticular protuberances on adult ALS singular or multiple, that results, afterecdysis, and PMS that appear to be nonfunctional ves- from a collared opening forming in the de- tiges of 2° MaA and 2° MiA spigots, respec- veloping exoskeleton during proecdysis; the tively. Othercuticularprotuberances, scattered opening accommodates a silk gland duct, al- within piriform and aciniform spinning fields lowing the duct to remain attached to a spigot (Kovoor 1986; Platnick 1990; Yu & Codding- on the old exoskeleton during proecdysis. By ton 1990), as well as on the PMS and poste- these definitions the protuberances that were rior lateral spinnerets (PLS) of at least some initially called tartipores (those among piri- mygalomorphs (Glatz 1973; Palmer 1990), form and aciniform spigots) are still called tar- have been called ‘tartipores’ (Shear et al. tipores (Fig. 1). However, only some of the 1989; Yu & Coddington 1990). Originally, the structures previously referred to as nubbins distinction between a nubbin and a tartipore are still called nubbins by our definition. For was based on whether the protuberance is a example, as in earlier reports, we identify as morphological singular and can, therefore, be nubbins those nonfunctional protuberances in uniquely designated (nubbin) or if it is one of some adults that occur where functional 2° several, or many, such structures on a single MaA/2° MiA spigots would have formed if spinneret that are designated collectively (tar- the spider had instead molted to yet another & tipore) (Yu Coddington 1990; see also Cod- juvenile instar (see Figs. 1, 13, 15). But there dington 1989:81). Both nubbins and tartipores are other protuberances near ampullate spigots were tentatively interpreted to be vestigial in many adult andjuvenile araneomorphs, pre- spigots. Additional observations, however, re- viously called nubbins (e.g., Yu & Coddington vealed that the protuberances identified as tar- 1990; Townley et al. 1991, 1993; Tillinghast tipores are not vestigial spigots. Instead, they & Townley 1994), that we now identify as are the remains of collared openings that ampullate tartipores, including the “second formed in the cuticle when it was first being nubbin” on the PMS of adult anapids and laid down during proecdysis beneath the old some synotaxids (Griswold et al. 1998:41) and cuticle (Townley et al. 1993). The openings the “second remnant” on the PMS of adult allowed silk gland ducts to maintain their at- oxyopids (Kovoor & Munoz-Cuevas 1998: 214 THE JOURNAL OF ARACHNOLOGY TOWNLEY & TILLINGHAST—EGG SAC ATTACHMENT IN LYCOSIDS 215 136). This is not the first time such protuber- ists on the PMS with the spinning apparatus ances have been referred to as tartipores (Plat- of MiA and aciniform glands, respectively. nick & Forster 1993:7, 9; Griswold et ah Ampullate gland nub—bins versus ampul- 1998:11), but in these earlier instances the dis- late gland tartipores. When examining tinction made between tartipores and nubbins spinnerets, care must be taken if one wishes was not stated. The term tartipore was perhaps to determine whether ampullate nubbins and applied solely because ofthe resemblance be- tartipores are present or not, as well as distin- tween ampullate tartipores and the more well guish the former category from the latter. known tartipores in piriform and aciniform Viewing the spinnerets at various angles and spinning fields, rather than because of recog- from different directions is sometimes re- nition of what tartipores, as here defined, rep- quired. In some adult araneoids, for example, resent. As indicated above, ampullate tarti- the MiA nubbin and MiA tartipore on a PMS pores mark the sites where 2° ampullate gland often occur side by side (e.g., Coddington ducts passed through the cuticle during the 1989:fig 16; Platnick et al, 1991:fig. 271, low- most recent proecdysis, enabling 2° ampullate er black arrow, tartipore on left, nubbin right; glands to function throughout proecdysis (Fig. Townley et al. 1991:fig. 24; lower arrow to 1). Note that Fig. 1 depicts only the ALS from tartipore, upper to nubbin; Hormiga et al. a lycosid and so only spigots, tartipores, and 1995:fig. 16C, nubbin left, tartipore right) and ducts ofMaA and piriform glands are shown. can be interpreted as a single structure if Bear in mind that a comparable situation ex- viewed at too low a magnification or from an — Figure 1. Schematic diagram of the left anterior lateral spinneret (ALS) of a female lycosid during proecdysis, shortly before the ecdysis that yields an adult. The upper ALS diagram represents the cuticle of the penultimate instar which will be cast off at ecdysis. The ALS diagram below this represents the underlying, newly-formed cuticle which will be part of the exoskeleton of the adult. Shown is the entire distal segment (DS) of the ALS, to which the major ampullate (MaA) spigots (labeled as ‘1°’ and ‘2°’) and piriform spigots (P) are restricted, atop the more distal portion ofthe ALS proximal segment (PS).To aid orientation, a much less magnified depiction ofthe same part ofthe left ALS is enclosed by a box in the ventral view ofthe spider, shown at top, in which the spinnerets are presented as ifartificially spread. By late proecdysis, the duct of the primary (1°) MaA gland, previously connected to a spigot on the old cuticle (that labeled ‘1°’)^ hasjust been re-modeled (Townley et al. 1991, 1993) and is now connected to a spigot on the new cuticle (again labeled ‘1°’) (silk gland ducts are indicated by dashed lines). Thus, the MaA 1° gland, nonfunctional during proecdysis, will again be functional immediately after ecdysis. Col- lared openings (tartipore progenitors) form in the new cuticle to accommodate the ducts ofany silk glands thataretoremain functional throughoutproecdysis. The ducts oftwo such silkglands are shownconnected to spigots on the old cuticle. After ecdysis the collapsed forms of these openings (tartipores) will remain evident in the new cuticle. A single MaA tartipore progenitor (MaATP) forms on each ALS of the new cuticle to accommodate the duct of a secondary (2°) MaA gland. Multiple piriform tartipore progenitors (FTP) also form on each ALS, one per piriform gland duct. (For clarity only a few piriform spigots are shown, and ofthose on the old cuticle, the duct connected to only one is shown. In reality, more piriform spigots are present and it appears that each piriform spigot on the old cuticle remains connected to a functioning duct, thus requiring the formation of one FTP on the new cuticle for each piriform spigot on the old cuticle.) The 2° MaA gland identified as ‘open’ will become ‘blocked’ at ecdysis (because its outlet, the spigot, will be lost along with the rest of the old cuticle), and, conversely, that identified as ‘blocked’ will become ‘open’ (since the 2° MaA spigot it is connected to will be open to the outside environment after ecdysis). The portion of the female new cuticle shown within a box differs from the situation in males (depicted at lower right) because 2° ampullate spigots do not form in adult males. Only ampullate nubbins (MaA nubbin (MaAN) on ALS, MiA nubbin on PMS) form in the positions occupied by 2° ampullate spigots in adult females and, thus, all 2° ampullate glands are ‘blocked’ and nonfunctional in adult males. Structures not drawn precisely to scale, BL, book lung; ALS, anterior lateral spinneret; PMS, posterior median spinneret; PLS, posterior lateral spinneret; DS, distal segment of anterior lateral spinneret; PS, proximal segment of anterior lateral spinneret; P, piriform gland spigot; PT, piriform tarti- pore; PTP, piriform tartipore progenitor; 1°, primary major ampullate gland spigot; 2°, secondary major ampullate gland spigot; MaAT, major ampullate tartipore; MaATP, major ampullate tartipore progenitor; MaAN, major ampullate nubbin. 216 THE JOURNAL OF ARACHNOLOGY inopportune angle, or if the MiA nubbin is niles, at least as early as second instars (see especially small. In describing the PMS of Methods for the definition of the first instar adult males of two anapid species, Platnick et used in this paper). We have not seen and are al. (1991:60) noted the presence of “a large not aware of any reports of tartipores in first posterior minor ampullate gland spigot ac- instars. However, given the occurrence of companied by a vestigial remnant bearing a functioning silk glands and spigots in postem- short lobe on its medial side”. The “vestigial bryos of at least some mygalomorph taxa remnant” is a MiA tartipore, the “short lobe” (Bond 1994), the possibility of tartipores in is a MiA nubbin. Even with careful observa- first instars of such taxa cannot be dismissed. tion it can sometimes be difficult, especially But at least for those araneomorphs in which with certain species, to discern a given tarti- functional silk glands and spigots first appear pore or nubbin. We were puzzled for a time in first instars, tartipores do not need to form by our inability to spot a MaA tartipore on the until deposition of the second instar cuticle ALS ofjuvenile and adult Araneus cavaticus begins. Consequently, for such spiders, tarti- (Keyserling 1882) until it became clear that pores would first be seen in second instars. this tartipore occurs at a site where, in this (The presence of one, presumably 1°, MaA species, the cuticle is typically compressed or spigot base per AFS in Nephila (Tetragnathi- overhung by the piriform spinning field and dae) postembryos has been described by Ble- the tartipore is obscured (Townley et al. her (2000), but their ability to produce silk is 1993). In contrast, single or multiple MaA tar- uncertain and only the ducts of 2° ampullate tipores are often clearly visible in many other glands are known to be accommodated by am- araneomorphs as a number of published mi- pullate tartipores.) crographs attest (several were cited in Town- It is of interest, therefore, that protuberanc- MaA ley et al. 1993:36 as “non-vestigial-type es, reminiscent of but recognizably different nubbins”; other examples include Platnick et from tartipores, are sometimes evident on the al. 1991: fig. 16, multiples in Gradungida, fig. AFS and PMS of first instars, in positions 39, a single in Thaida, fig. 277, a round single consistent with those of ampullate tartipores in Pachygnadm next to smaller oblong MaA in later instars. We have seen them near 2° nubbin; Harvey 1995: fig. 1 1, a single in Am- ampullate spigots on the AFS and PMS of hicodamus posterolateral to the 2° MaA spig- first instar Pardosa xerampelina (Keyserling ot; Davies 1998a:fig. 68, a single in Jalkahiir- 1877) (Figs. 2-5) and Octonoba sinensis (Si- ra lateral to the 2° MaA spigot; Platnick 1999: mon 1880) (Uloboridae), and on the PMS of fig. 3, a single in Liocranoides between and first insiav Argiope aurantia Fucas 1833 (Ar- lateral to 1° and 2° MaA spigots; Hormiga aneidae) and Herpyllus ecclesiasticus Hentz 2000:plate 42B, a single in Laminacauda pos- 1832 (Gnaphosidae). We tentatively refer to terolateral to MaA nubbin, larger than the them as ‘pre-tartipores’ (not to be confused multiple piriform tartipores). In this paper, sin- with the tartipore progenitors referred to in MaA gle tartipores can be seen in Figs. 1, 8, Fig. 1). If they truly are precursors of the tar- 9, 12, 13, 16, 18, 22, 24, 26, 28, 30, 32, 34, tipores in later instars, their occurrence sug- 36, 40 & 41 and single MiA tartipores can be gests that the epithelial cells that are capable seen in Figs. 10, 11, 14, 15, 19-21, 23, 25, of forming tartipores, at least ampullate tarti- 27, 29, 31, 33, 35, 37-39 & 42. pores, are already determined by the postem- Given that the occurrence and number of bryo stage. — ampullate “nubbins” (i.e., ampullate nubbins Occurrence of nubbins. In general, nub- and/or tartipores) are being used as characters bins as here defined occur in adults, being in cladistic analyses (Coddington 1990; Hor- more abundant in males (largely since silk miga et al. 1995; Scharff& Coddington 1997; glands used solely orprimarily in prey capture Griswold et al. 1998, 1999; Hormiga 2000), tend to regress in adult males), and are onto- accurately determining the presence/absence genetically vestigial. That is, they are located of ampullate tartipores and nubbins, and mak- in adults in positions where functional spigots ing a clear distinction between the two, can would have formed ifthe spider had remained only aid phylogenetic studies.— a juvenile after its most recent molt. In addi- Occurrence of tartipores. Tartipores can tion to the MaA and MiA nubbins present in occur in the exoskeletons of adults and Juve- a variety of adult male and female araneocla- TOWNLEY & TILLINGHAST—EGG SAC ATTACHMENT IN LYCOSIDS — Figures 2-5. ALS and PMS from a first instar Pardosa xerampelina (removed from dorsum of its mother’s abdomen) showing protuberances, tentatively termed ‘pre-tartipores’, in positions that are held by ampullate tartipores in later instars: 2. Right ALS, entire spinning field shown (two MaA and three piriform spigots), pre-tartipore in box; 3. Higher magnification ofpre-tartipore from Fig. 2; 4. Right PMS, entire spinning field shown (two MiA and three aciniform spigots), pre-tartipore in box; 5. Higher mag- nification ofpre-tartipore from Fig. 4. Posterior at top, lateral at right in all four figures. Scale bars (2, 4) = 5 fxm; (3, 5) = 1 pm. 218 THE JOURNAL OF ARACHNOLOGY — Table 2. Spider species in which adult females are known to have two MaA spigots (1° and 2°) on each ALS and two MiA spigots (1° and 2°) on each PMS while adult males have one MaA spigot (1°) on each ALS and one MiA spigot (1°) on each PMS. The families listed here are almost certainly not the only ones that contain species fitting this description. Note thatNeoramia (Agelenidae) apparently do not fit this description (Griswold et al. 1999); nor do some salticid genera (see ‘Ampullate gland spigot, nubbin, tartiporecomplements’ in Results) orseveralamaurobiidgenera, inclndmgAmaurobius(see 'Com.= parative ampullate gland spigot morphology’ in Discussion). Also, this description may not extend to all Coras species (see ‘Ampullate gland spigot, nubbin, tartipore complements’ in Results). Family Species References Lycosidae Gladicosa gulosa (Walckenaer 1837) this study Pardosa amentata (Clerck 1757) Richter 1970 Pardosa lapidicina Emerton 1885 this study Pardosa lugubris (Walckenaer 1802) W^sowska 1977 Pardosa modica (Blackwall 1846) this study Pardosa moesta Banks 1892 this study Pardosa saxatilis (Hentz 1844) this study Pisauridae Dolomedes scriptus Hentz 1845 this study Pisaurina mira (Walckenaer 1837) this study Agelenidae Agelena labyrinthica (Clerck 1757) Kokocihski 1968 Agelenopsis naevia (Walckenaer 1842) this study Agelenopsispotteri (Blackwall 1846) this study Amaurobiidae Coras aerialis Muma 1946 this study Thomisidae Misumenops asperatus (Hentz 1847) this study Xysticus cristatus (Clerck 1757) W^sowska 1977 Philodromidae Tibellus oblongus (Walckenaer 1802) W^sowska 1967, 1977; this study Clubionidae Clubiona phragmitis C.L, Koch 1843 Mikulska 1969; W^sowska 1969; Wisniewski 1986a,b Miturgidae Cheiracanthium mildei L. Koch 1864 this study Salticidae Salticus scenicus (Clerck 1757) this study dans, flagelliform and aggregate nubbins form 1993:figs. 17D,E). By the interpretation of on the PLS of many adult male araneoids Platnick et ak (1991:51) the latter would be (Sekiguchi 1955a; Peters & Kovoor 1991:fig. MaA nubbins. 3b; Platnick et ak 1991:fig. 275; Townley et Among the examples of exceptions to the ak 1991:fig. 16; Townley 1993:fig. 16; Gris- general rule are nubbins occasionally seen in wold et ak 1998:figs. 25D, 39D, 43D), though early instar Cyrtophora (Araneidae) that sug- a number of males within the ‘reduced piri- gest phylogenetic vestiges of either aggregate form clade’ ofGriswold et al. (1998) (see also or flagelliform spigots. In an examination of Hormiga 2000) and the Micropholcommatidae six first instar Cyrtophora citricola (Forskal (Schiitt 2000) retain the aggregate/flagelliform 1775), on one PLS Peters (1993:figs. 11b, c) spigot triad. Other examples include aciniform nubbins on the PMS (Muller & Westheide observed a single “shaft-like structure'’ on the 1993) and PLS (pers. obs.) of adult male Ar- vestigial plate of the aggregate-flagelliform giope and on the PMS of some adult male triad. (On nine PLS one or two “knobs with uloborids (Kovoor & Peters 1988:53), pseu- pores” were seen on these vestigial plates, but doflagelliform nubbins on the PLS and para- we do not interpret these as nubbins.) Nubbins cribellar nubbins on the PMS of some adult that are also apparently phylogenetic vestiges male deinopoids (Kovoor & Peters 1988; Pe- of aggregate spigots are often retained right ters 1992), paracribellar nubbins on the PMS up to maturity in female Drapetisca socialis of adult male austrochilids (Peters 1983; Plat- (Sundevall 1833) (Linyphiidae); functional nick et al. 1991:fig. 33), and nubbins of un- aggregate spigots are absent throughout the certain gland type on the ALS of adult male ontogeny of these spiders (Schiitt 1995). The Hypochilus, next to the single, large ampullate occurrence of a MaA nubbin on the ALS of spigot (Platnick et al. 1991:fig. 4; Townley penultimate instar female Maiala lubinae Da-