Malenella nana Ramírez, 1995a: 376

RAMÍREZ, MARTÍN J., 2003, The Spider Subfamily Amaurobioidinae (Araneae, Anyphaenidae): A Phylogenetic Revision At The Generic Level, Bulletin of the American Museum of Natural History 2003 (277), pp. 1-262 : 39-57

Malenella nana Ramírez, 1995a: 376 .

NEW RECORDS: Región Metropolitana (Santiago): Santiago: Bucalemi, San Antonio, 23–24.X.1994, L. Peña (AMNH). Región IX (Araucanía): Malleco: 16.5 km NE Pucón, 12.I.1951, Ross and Michelbacher, 1♀ (CAS). Cautín: Flor del Lago Ranch, Villarrica, Polo Field, 39°12.300̍S, 72°08.367̍W, 282 m, canopy fogging GT Nothofagus obliqua roble, 13.XII.2001, Arias et al., 1♀ (UCB). Región X (Los Lagos): Chiloe´: Isla de Chiloe´: El Pozuelo, 1.5 km NE Butalcura, 2.II.2001, T. Cekalovic, 1♀ (AMNH).

ANYPHAENINAE BERTKAU Table 4, Figure 9B, C

Anyphaenidae Bertkau, 1878: 358 (type genus Anyphaena Sundevall, 1833 ).

Anyphaeninae Bertkau, 1878: 379. Reviewed at generic level by Brescovit, 1997.

DIAGNOSIS: Tracheal spiracle advanced, at midpoint or closer to epigastrium than to spinnerets. Tegulum without basal notch.

NOTE: In this analysis Anyphaeninae is paraphyletic in terms of Amaurobioidinae , but see comments below.

DESCRIPTION: Recently redescribed by Brescovit (1997).

DISTRIBUTION AND COMPOSITION: Thirtythree genera, mostly from the New World. Anyphaena Sundevall has representatives in Palearctic and Oriental regions; Australaena Berland is known only from Polynesia.

Six representative species of Anyphaeninae are included as outgroups. They were selected for having putatively homologous conditions of important characters diagnostic of higher groups of Amaurobioidinae . Species of Xiruana Brescovit have a structure (‘‘conductor of Anyphaeninae’’, Brescovit, 1997) very similar to the secondary conductor found in amaurobioidines. The prolateral and retrolateral tegular projections in Anyphaena accentuata (Walckenaer) ( Brescovit, 1997: figs. 3, 4) were both tentatively homologized with the secondary conductor. Species of Wulfila O.P.­ Cambridge and Anyphaena accentuata have a putative homolog of a primary conductor (named ‘‘ventral tegular projection’’ by Brescovit, 1997). In A. accentuata , this structure has a shallow canal where the embolus fits, favoring homology with a conductor (fig. 11A). Species of Aysha Keyserling have a complex process at the embolar base comparable to that of Josa . Italaman santamaria Brescovit has an acute retrolateral tibial apophysis similar to that found in Amaurobioides and close relatives. Several conditions present in Malenella nana are also found in Wulfila (green color, minute AME, absence of accessory bulbs in the spermathecae), Italaman (simple male copulatory bulb), and Otoniela Brescovit (many anterior leg spines, absence of accessory bulbs). This selection of representative outgroup characters is clearly insufficient for a sound hypothesis of relationships among anyphaenines. Because only a few characters specific for that subfamily are here included, and because the rather unusual Malenella was used as outer outgroup, it is clear that a more complete analysis is needed to appropriately assess anyphaenine relationships. In this dataset, only three correlated characters would support the monophyly of Anyphaeninae (the advanced tracheal spiracle and associated modifications of the tracheal system, characters 38–40). The effect of the suspicious resolution of Anyphaeninae over the relationships among amaurobioidines, however, is limited: if Anyphaeninae is constrained to be monophyletic, all groups in Amaurobioidinae are the same in the optimal trees, except for the collapsing of clade 95 of Josa . Character optimizations at the basal node of Amaurobioidinae are the same except for characters 15 (state 1 instead of 2), 79 ([012] instead of 1), 123 ([01] instead of 1), 125 (0 instead of [01]), 138 ([01] instead of 0), and 163 ([01] instead of 1).

Brescovit (1997) suggested a monophyletic group of nine genera including Aysha , united by having a process on the male embolar base. I cannot test his hypothesis here, not only because of my modest selection of representatives and specific characters for Anyphaeninae, but because there is not yet a coherent scheme of homologies for the many details of embolar morphology, embracing both anyphaenines and amaurobioidines.

AMAUROBIOIDINAE HICKMAN Table 5

Amaurobioididae Hickman, 1949: 31 (type genus Amaurobioides O.P.­ Cambridge, 1883). Forster, 1970: 165. Synonymized with Anyphaenidae by Platnick , 1974: 211.

Amaurobioidinae : Lehtinen, 1967: 211, 316, 320– 321. Kochalka, 1980: 65. Ramírez, 1995a: 381.

NOTE: Simon (1897a: 99–100, 104, 1903a: 1032, 1903c: 29) used informal names for some Amaurobioidinae that might be confused with family­level names (Tomopisthini, Oxysomini, Oxysomate), which clearly referred to the members of each genus, and are not available at familylevel ( ICZN, 1999: 11 [f]2).

DIAGNOSIS: Distinguished from Anyphaeninae by the male tegulum with a deep notch occupied by the median hematodocha, visible in ventral view as a membranous area at base of the copulatory bulb. Most species have the tracheal spiracle closer to the spinnerets than to the epigastrium.

DESCRIPTION: Body size small (2.50) to medium (22.00). Chelicerae (fig. 12A) with three teeth on promargin (exceptionally four or five), two to seven teeth on retromargin (exceptionally one). Labial apex rounded or slightly notched. Male palp with only one tibial apophysis in retrolateral apical position, or lacking apophysis. Tip of cymbium with ventral canal devoid of setae, often associated with embolus (cymbial conductor). Subtegulum prolateral, visible in unexpanded palp. Tegulum with deep basal notch occupied by median hematodocha. Sperm reser­

TABLE 4 Synapomorphies of Outgroup Clades

TABLE 4 (Continued)

voir in subtegulum, sperm duct running to tegulum through apical margin of bulb (clockwise, left palp ventral view), then parallel through retrolateral border of cymbium, then bordering tegular notch and entering into embolar base. Embolus articulated prolaterally to tegulum. Embolar base usually with one process of diverse shape; terminal portion of embolus filiform or moderately thick, usually fitting into canal on conductor (primary or secondary) and into cymbial conductor. Apical portion of tegulum housing distal hematodocha, where rest of sclerites are inserted. Median apophysis retrolateral, usually articulated on weakly sclerotized area, commonly hook­shaped, sometimes bifid, reduced, even absent in some species. Paramedian apophysis arising from sclerotized plate on distal hematodocha, between median apophysis and retrolateral ventral border of tegulum, with one to several sclerotized cusps. Primary conductor arising between embolus and prolateral ventral bor­ der of tegulum (relictual in Gayennini ), often with canal where embolus fits. Secondary conductor arising distally, associated with apical dorsal tegular stripe where sperm duct runs. Epigyne with paired lateral lobes and median field, copulatory openings usually associated with furrows between median field

TABLE 5 Synapomorphies of Amaurobioidinae and Clade 176

and lateral lobes. Copulatory duct weakly sclerotized from opening to joint with duct of accessory bulb; from that point duct well sclerotized, coming into spermatheca. ‘‘Dictynoid’’ pore conspicuous, close to union of copulatory duct with spermatheca and fertilization duct. Tracheal spiracle variable from close to spinnerets to midpoint between spinnerets and epigastrium (except in some Arachosia and Acanthoceto , slightly closer to epigastrium). Length of lateral tracheae variable according to spiracle position, reaching spinnerets. Trichobothriae in one row on metatarsi, two rows on tarsi (fig. 13A). Anterior lateral spinnerets with two major ampullate gland spigots (figs. 15B, 16B, 17B, 117B), or one plus nubbin ( Ramírez, 1995a: fig. 39), and several unmodified piriform gland spigots. Posterior median spinnerets with two minor ampullate gland spigots (figs. 16C, D, 17C) and several aciniform gland spigots. One PMS minor ampullate detected in Amaurobioides (fig. 15C, contra Ramírez, 1995a), but it is so similar to aciniforms that second one may have been overlooked. Posterior lateral spinnerets only with aciniform gland spigots (figs. 15D, 16E, 17D). Gland spigots not sexually dimorphic, cylindrical gland spigots absent.

DISTRIBUTION: Mainly South America, but Josa extending also to Central America, Arachosia to Central and North America, and the coastal genus Amaurobioides in the seashores of Chile, South Africa, Australia, Tasmania, and New Zealand.

COMPOSITION: Twenty­two genera grouped in two tribes, Amaurobioidini and Gayennini , here defined, plus the genus Josa , sister group of Gayennini , for which I declined to erect a tribe.

TYPES NOT EXAMINED: The following species have been assigned to various amaurobioidine genera, but their placement cannot be clarified without examination of the types: Anyphaena trivittata Bertkau, 1880 ; Anyphaena furcata Keyserling, 1880 ; and Anyphaena vittata Keyserling, 1881 .

NOMINA DUBIA: The types of the following Chilean species have not been found in MHNP, and the original descriptions are very ambiguous: Clubiona lineata Nicolet, 1849 ; Clubiona limbata Nicolet, 1849 ; Clubiona nubes Nicolet, 1849 ; and Clubiona versicol­ or Nicolet, 1849.

MORPHOLOGICAL REMARKS

MALE COPULATORY BULB

The copulatory bulb is united to the cymbial alveolus by the basal hematodocha and a triangular petiole. Because neither the subtegulum nor the tegulum forms a complete ring, the three hematodochae (basal, median, and distal) are continuous (fig. 33E). The subtegulum is compact, partially visible in prolateral view in the unexpanded bulb (figs. 83A, 129E), but mostly distally in Amaurobioidini (fig. 47B). Part of the median hematodocha is visible in ventral view in the unexpanded palp, occupying a basal notch on the tegulum, which is the most conspicuous synapomorphy of Amaurobioidinae . The anterior part of the bulb is occupied by the distal hematodocha, from where the conductors, the median apophysis, and the paramedian apophysis arise.

CONDUCTORS

The homology of the extremely diverse structures accompanying the embolus of entelegyne spiders is a contentious field. Here I tried to establish the homology of the structures across the subfamily, with variable success. I call the primary conductor (C1) a structure arising between the base of the embolus and the prolateral margin of the tegulum. The C1 is most evident in Amaurobioidini , often bearing a long canal where the embolus fits. The apical portion of C1, where the canal ends, usually forms a heavily sclerotized, partially coiled beak (e.g., fig. 50D). The apical portion of C1 may look like a separate sclerite if the basal portion is protruding as well (fig. 33B). In the Gayennini , there is a small sclerite mostly hidden by the other structures arising from the distal hematodocha (figs. 63C, 105C), connected by a sclerotized stripe to the articulation between the embolus and prolateral margin of tegulum, which is a presumed homolog of the C1. In Josa the basal portion of the C1 is fused with the tegulum; the apical portion is no longer identifiable (fig. 60A, B).

I call the secondary conductor (C2) a sclerite also associated with the embolus, arising from the apical­dorsal region of the distal hematodocha. It is closely related, often fused, to a sclerotized stripe of the tegulum where the sperm duct runs. In Amaurobioidini , the main structure leading the embolus is the C1, and the C2 is small, with a poorly defined canal (fig. 26A). In Gayennini this situation is reversed, and the C2 is a conspicuous structure. In Josa , the C2 is hypertrophied and very complex (figs. 56B, 60B).

PARAMEDIAN APOPHYSIS

Hidden behind the retrolateral margin of the tegulum there is a tightly folded section of the distal hematodocha, which becomes inflated during expansion (fig. 50D; see also character 71). A sclerotized plate attached to this section of hematodocha is called the paramedian apophysis (PMA), which, in Amaurobioidini , is distally articulated to the C1 (fig. 50D). The PMA may have one (in Gayennini ) to several cusps (in Amaurobioidini ). The PMA is also connected, often more or less fused, with the median apophysis. A protruding PMA appears in Gayennini (see character 68), together with the anterior pouch on the median epigynal field (APmf, character 104). Because the shape of the PMA seems also to be somewhat correlated with the shape of the APmf (e.g., fig. 124D, H), it is possible that the PMA fits into the APmf during copulation, at least in some species.

MEDIAN APOPHYSIS

The median apophysis (MA) is much more conservative than are the conductors. In most Amaurobioidinae the MA is very simple, small, and hooked, at some extent articulated or fused with the plate of the PMA. Species of Coptoprepes have a larger, complex MA. Davies (1998) doubted the homology of the MA in Amaurobioides , because it is not articulated, but is firmly fused to a complex of paramedian apophysis and part of the conductor. However, when close relatives of Amaurobioides are examined ( Axyracrus , Aysenia , Aysenoides ), it is clear that the distal hematodocha where C1, PMA, and AM are placed underwent variable degrees of sclerotization. Because the morphology of the area is otherwise conservative, the homology of the MA seems clear. The homology of the MA as identified in Josa is somewhat more problematic, because the entire distal region of the bulb is extremely modified. The MA becomes reduced independently in several clades (character 64), but in these cases the general morphology of the area is similar to that of close relatives.

EMBOLUS

Coddington (1990) noted that the embolus of Amaurobioides runs counterclockwise, differing from most entelegynes. Other Amaurobioidini , and to a lesser degree to a Gayennini , also have that embolar course. The change from clockwise to counterclockwise involved only minor modifications of the copulatory bulb, compared to a more generalized conformation. The region of the basal articulation of the embolus has an acute to straight angle, thus changing the course, while the basal part of the sperm duct runs clockwise, as in other entelegynes. Amaurobioidines with a less pronounced angle (e.g., Selknamia , figs. 52, 53A, B) are ambiguous as to the curvature, while in others the curvature looks different from prolateral and ventral views (e.g., Negayan , figs. 47D, 50A, B). In Anyphaeninae there is great variability in embolus morphology, and sometimes it is not clear in what direction a contorted embolus is running.

ALTERNATIVE INTERPRETATIONS OF MALE PALP STRUCTURES

The identification of C1, C2, and cusps of PMA is disputable. The system of homologies adopted here is what seemed more congruent after detailed study of the anatomy and informal examination of alternative codings during the construction of the dataset. Those alternative codings did not produce very different results from those shown here, though. This is because alternative homologies have a coordinated effect on groups of terminals with similar morphology. The C 1 in Gayennini , as interpreted here, is a small piece, compared to the structure found in Amaurobioidini . An alternative coding might consider this small piece the second cusp of the PMA (which is lacking in Gayennini , compared with most Amaurobioidini ). The C2 of Gayennini might be considered a C1 (or conversely, the apical portion of C 1 in some Amaurobioidini , a C2), because in Negayan and Selknamia the apical portion of C1 is similar to the C2 of Gayennini . This coding was used in preliminary datasets, but was later abandoned, because the homologies would be supported by intermediate morphologies that were not intermediate in phylogeny! Settling these problems of unclear homologies may involve the use of alternative codings, keeping those homologies (and the corresponding trees) that imply most parsimonious results ( Rieppel, 1996; Wheeler, 1996).

SPERMATHECAE AND ASSOCIATED DUCTS

There is some disagreement on the terminology for these structures. I mostly followed Sierwald (1989), with modifications. The main difference is interpretation of the copulatory duct. In Sierwald’s view, this duct runs from the copulatory opening to the connection with a duct from the ‘‘head of spermatheca’’, here accessory bulb (AB) after Carico and Holt (1964). The segment from that point to the ‘‘base of spermatheca’’ (here spermatheca) is considered by Sierwald as part of the spermatheca itself, the ‘‘peduncle of spermatheca’’, uniting the head and base. This terminology seems adequate for Lycosoidea ( Griswold, 1993; Diana Silva, personal commun.), but not so much for Amaurobioidinae , at least in functional terms. In Amaurobioidinae , the accessory bulb joins the copulatory duct relatively close to the copulatory opening, and from this point to the origin of the fertilization duct there runs a tube that evidently functions as a duct rath­ er than as a reservoir. In fact, the male embolus runs through that segment ( Ramírez and Kochalka, 1993: fig. 4), and I could not find much justification for considering it part of the spermatheca (a name that indicates storage function). Comparative anatomy does not help much to settle the question, because the evolutionary transformations that lead to the entelegyne female genitalia are not well understood. I am inclined to accept the homology of the accessory bulb with one of the paired receptacles, as found in most Mygalomorphae and many Haplogynae. In this case, the fertilization ducts are homologous to the ducts leading to one pair of receptacles, the entelegyne copulatory openings are homologous to the duct leading to the second pair of receptacles, and the ‘‘peduncles’’ are invaginations of the body wall connecting the two receptacles of each side (see Sierwald, 1989).

The development in Tomopisthes horrendus is in agreement with this hypothesis (fig. 106). The accessory bulb arises anteriorly on a cuticular fold separating the median field from the lateral lobes, the primordium of the copulatory opening leads to the primordium of the accessory bulb, and both primordia are connected by a deep folding corresponding to the copulatory duct. For the sake of descriptive power, I refer as ‘‘spermatheca’’ to the ample chamber immediately connected to the fertilization duct, ‘‘copulatory duct’’ to the tube running from the copulatory opening to spermatheca, and ‘‘accessory bulb’’ to the blind sac bearing conspicuous pores, connected to the copulatory duct by a tube of variable length. The accessory bulb was recently referred to as ‘‘seminal receptacle’’ ( Brescovit, 1997) and ‘‘diverticulum’’ (Hub­ er, 1995). Bonaldo (2000) called it ‘‘secondary spermatheca’’ on account of its function as a sperm reservoir in Corinninae .

LEG SPINES

Goloboff (1995b) used several characters from patterns of spines in Nemesiidae . Bosselaers and Jocqué (2002) used both patterns and individual spines in an analysis of some Corinnidae and Liocranidae . They found those characters informative, with similar levels of homoplasy as for other somatic ones. Here many of the leg spines were cod­ ed as independent characters, homologized according to their position. This is possible since the spination in Anyphaenidae follows a rather conservative pattern.

In most genera the spines on leg I are similarly distributed to those on leg II. Legs III and IV are also similar in spines, which are more numerous than on forelegs. Through the four pairs of legs, most spine positions are conserved, because they are serially homologous. A common pattern is:

Legs I and II, femur d 1–1–1, p 0–1­ (1­ d1), r d1ap; tibia v 2–2–2; metatarsus v 2bas. III, femur d 1–1– 1, p and r 0­ d1­ d1; patella r d1; tibia v 2–2–2, p and r d1–1, d r1bas; metatarsus v 2–2–2, p and r d1–1– 1, d 0­ p1–2. IV, femur d 1–1–1, p 0­ d1­ d1, r d1ap; patella, tibia, and metatarsus = III.

In some groups the anterior legs are almost as spinose as the posterior legs. A common pattern of this type is:

Leg I and II, femur d 1–1–1, p and r 0­ d1­ (1­ d1); tibia v 2–2–2, p and r d1–1, d r1­ 0­ 1­ 0; metatarsus v 2bas, p and r d1–1–1, d 0­ p1–2. III, femur = I; patella r d1; tibia = I; metatarsus = I, but v 2–2–2. IV, femur d 1–1–1, p 0­ d1­ d1, r d1ap; patella, tibia, and metatarsus = III.

Most spine patterns vary between these two examples. In the spinose pattern, spines on anterior and posterior legs differ mostly by the ventrals on metatarsi. There are only a few species with more than two ventral spines on metatarsus I or II, they are not especially spinose on other surfaces, and these spines are not usually sexually dimorphic. Some species have more than three pairs of ventral spines on tibiae I and II, conferring a raptorial appearance (e.g., some Monapia ).

Males are often more spinose than are females. The additional male spines appear after the last ecdysis. Spines of penultimates of both sexes are similar to those of the female. In some rare specimens (but commonly in Sanogasta backhauseni ) there are supernumerary spines, for example, two or three spines where one is expected. Such an anomaly is often asymmetrical.

Bristles (similar to spines but thinner and shorter) seem to be homologous to spines, because some specimens have a bristle where a spine is normally found. Frequent positions for replacement of spines by bristles are the prolaterals and retrolaterals on femora, and the v p1­ x­x of tibia II. In species with spinose males, it is common that the male has a spine where the female has a bristle; common positions are the dorsals of tibiae (r1­ 0­ 1­ 0) and patellae (1–0–1).

Inter­ and intrasexual variabilities in spines were coded without distinction as polymorphisms in the data matrix, with internal steps added accordingly. It is expected that variability is underestimated in species known from a few exemplars, and more drastically if the males are unknown. To estimate the effect of variations in the internal steps for the spine characters, several replications were examined with the internal steps for the spine characters assigned randomly between 0 and 29. These produced virtually the same trees, except for some clades of little support, suggesting that underestimation of variability in spines is not decisive in this analysis.

TRIBE AMAUROBIOIDINI HICKMAN Table 6

Amaurobioididae Hickman, 1949: 31 (type genus Amaurobioides O.P.­ Cambridge, 1883).

DIAGNOSIS: Distinguished from Gayennini and Josa by the male copulatory bulb with an apical dorsal loop on the sperm duct, visible in apical view (figs. 21H, 26C, white arrow; absent in Acanthoceto acupicta group, see character 53), and by the well­developed primary conductor with canal where the embolus fits (fig. 47).

DESCRIPTION: Chelicerae commonly with three or more teeth on retromargin, sometimes only two. Male palp with one retrolateral tibial apophysis (reduced in Gamakia , absent in Coptoprepes campanensis ). Cymbial conductor wide. Sperm duct with loop on apical side of copulatory bulb. Primary conductor with canal where embolus fits; basal portion often weakly sclerotized, close to base of embolus; apical portion, where canal ends, sclerotized, of varied shape, sometimes diverging from basal portion (fig. 33B). Secondary conductor generally small, partially or totally fused to apical tegular

TABLE 6 Synapomorphies of Amaurobioidini and Internal Clades

TABLE 7 Synapomorphies of Amaurobioides and Representatives tralia, Tasmania, and New Zealand, and Sanogasta maculatipes , probably introduced in Eastern Island.

COMPOSITION: Ten genera, four of them newly proposed here: Acanthoceto Mello­Leitão , Amaurobioides O.P.­ Cambridge, Axyracrus Simon , Aysenia Tullgren , Aysenoides , n. gen., Coptoprepes Simon , Ferrieria Tullgren , Gamakia , n. gen., Negayan , n. gen., and Selknamia , n. gen.

stripe where basal part of sperm duct runs, well developed only in some Coptoprepes . Paramedian apophysis not projecting, usually with two conical cusps. Epigyne without anterior pouch, lateral lobes separated from each other. Spermathecae usually of irregular shape, but spherical in Selknamia and Aysenoides .

DISTRIBUTION: South America, except the coastal genus Amaurobioides , which occur in the sea­shores of Chile, South Africa, Aus­

AMAUROBIOIDES O.P.­ CAMBRIDGE

Table 7

Amaurobioides O.P.­ Cambridge, 1883: 356 (type species by monotypy Amaurobioides maritima O.P.­ Cambridge, 1883). Simon, 1897a: 89, 1903a: 1034. Hogg, 1909: 163. Hewitt, 1917: 704. Hickman, 1949: 31, 1951: 1. Forster, 1955: 184. Lehtinen, 1967: 211 (removed from synonymy of Uliodon L. Koch, 1873 ). Coddington, 1990: 10. Ramírez, 1995a: 366, 1997: 178. Davies, 1998: 212.

Cluilius Simon, 1889: 220 (type species by monotypy Clubiona chilensis Nicolet, 1849 ; see Simon, 1904: 100, and Synonymy in Philisca ). NEW SYNONYMY.

SYNONYMY: The type species of Cluilius is here considered a typical Amaurobioides . See also Synonymy in Philisca .

DIAGNOSIS: Distinguished from other genera of Anyphaenidae by having many aciniform gland spigots on posterior median spinnerets ( Ramírez, 1995a: fig. 43).

DESCRIPTION: Redescribed by Forster (1970).

MORPHOLOGICAL REMARKS: The posterior median spinnerets densely covered by spigots might have a correlation with their aquatic habitats, because the retreat silk is much denser than that of other Anyphaenidae . In support of this association, the PMS are similarly enlarged and covered by many aciniform gland spigots in the European aquatic spider Argyroneta aquatica (Clerck) and in species of the intertidal spider genus Desis , from Pacific and South African coasts ( Lehtinen, 1967). Forster (1970) reported three processes on the male palpal conductor of Amaurobioides maritima . One of them is here interpreted as the tip of the primary conductor, bearing the canal, the other two as cusps of the paramedian apophysis. Davies (1998) suggested that the structure that is here identified as secondary conductor may be a primary conductor (C1) instead, because in Amaurobioides isolata Hirst it is a well­defined sclerite, entirely bordered by a membranous area; however, the same is true for the structure here identified as C1.

DISTRIBUTION: Shores of Austral regions: one species from Chile, one from South Africa, all others from Australia, New Zealand, and Tasmania.

COMPOSITION: In addition to the three species detailed below: A. isolata Hirst, 1993 , A. litoralis Hickman, 1949 , A. major Forster, 1970 , A. minor Forster, 1970 , A. pallida Forster, 1970 , A. picuna Forster, 1970 , A. piscator Hogg, 1909 , A. pleta Forster, 1970 , and A. pohara Forster, 1970 .

Amaurobioides maritima O.P.­ Cambridge Figure 18A–D

A. maritima O.P.­ Cambridge, 1883: 356 (male holotype from New Zealand, Otago, Allday Bay, in BMNH, not examined).

A. maritimus : Forster, 1970: 168. Ramírez, 1995a: 28.

DESCRIPTION AND DIAGNOSIS: See Forster (1970). Additional data are provided below.

FEMALE (Otago): Total length 15.45. Carapace length 5.45, width 4.26, wider on legs II–III. Leg III, length of tibia 2.50, metatarsus 2.67; leg IV, length of tibia 3.23. Sternum length 3.50, width 2.07. Abdomen length

10.00, width 6.25, spiracle–epigastrium 4.40, spiracle–spinnerets 1.10. Spines: leg I, femur d 1–1–1, p d1ap; tibia v 2–2–2; metatarsus v 2bas. II, femur = I; tibia v r1–2–2 or 2–2– 2, p 1–1; metatarsus v 2­ (p1­ r1)­ 0 or 2­ r1–0, p 1–0. III, femur d 1–1–1, p and r 0­ d1­ d1; tibia v 0–2–2, p and r 1­ 0­ 1­ 0; metatarsus v 2–2–2, p d1–1–1, r 0–1–1, d 2ap. IV, femur d 1–1–1, p and r d1ap; tibia v p1–2–2 (plus r1 supernumerary), r d1ap or 0­ d1­ d1; metatarsus v 2­ p1–2 or 2­ p1­ r1, p and r 0–1–1, d 2ap. Epigyne: lateral lobes separate, anterior margin elevated. Median field slightly sclerotized. Copulatory ducts irregular, contorted before reaching spermathecae, ducts of accessory bulbs long, thick, diverging (fig. 18A).

MALE (Otago): Spines as in female, except: leg I, femur d 1–1–1, p 0­ d1­ d1, r d1ap; tibia v 2­ 2­ 2­ 2 or 6 spines on each side, irregularly paired, p 1–2–1, r 1–0–1; metatarsus v 2bas, p d1–1–0, r 1, d p1. II, femur = I; tibia v 2–2–2, p and r 1–1–1; metatarsus v 2­ r1–0, p d1–1–0, r 1. III, tibia v 2–2–2, p and r 1­ d1–1. IV, femur = I; tibia = III, but p 0­ d1–1 or 1ap; metatarsus v 2–2–2, p 0–1–1, r d1–1–1. Palp (fig. 18B–D): tibia short, width/legth 0.85, RTA very long, sharp, slightly concave ventrally. Cymbium with retrolateral notch where median apophysis fits, and short basal projection, opposing tibia. Sperm duct thick, suddenly narrowed in front of tegular notch. Embolus with basal process flat, well developed. Median apophysis small, thin, apical. Paramedian apophysis with three separate cusps, apical cusp concave, placed under median apophysis, median cusp conical, heavily sclerotized, ventral cusp flattened, weakly sclerotized. Primary conductor well developed, with conspicuous canal. Secondary conductor compressed, partly fused to anterior dorsal margin of tegulum.

VARIABILITY: Spines, according to Forster (1970). Female: II, tibia p 0–1; metatarsus v 2bas. III, tibia p 1–1–1; metatarsus d p1ap. IV, tibia v p1­ p1–2, p 0; metatarsus r 0–1– 1. Male: I, tibia v 2–2–2, p and r 1–1–1; metatarsus v 2–2–0, p 1–1–0. II, metatarsus v 2–2–0, p and r 1–1–0. III, metatarsus r d1– 1–1, d 0­ p1–2. IV, tibia p 0–1–1, r 1–1–1.

MATERIAL EXAMINED: NEW ZEALAND: Otago, St. Clair Beach, on cliff face, 28.IV.1979, J. Carico, 13 1♀ ( MACN­Ar, identified by R. Forster).

Amaurobioides africana Hewitt Figures 13A–D , 15 , 18E –I

Amaurobioides africanus Hewitt, 1917: 704 (female holotype from South Africa, East London, in BMNH, examined). Lehtinen, 1967: 212. Forster, 1970: 176 (removed from synonymy of A. maritima O.P.­ Cambridge). Lopez, 1974: 902.

DIAGNOSIS: Distinguished from other Amaurobioides by the wide, sclerotized portion of paramedian apophysis with three separate cusps (fig. 18G–I), and by the shape of the elevated margins of epigyne (fig. 18F).

FEMALE ( Namibia): Total length 14.00. Carapace length 4.92, width 3.60, wider on leg II. Length of tibia/metatarsus: I, 2.83/ 2.50; II, 2.67/2.10; III, 2.37/2.10; IV, 2.60/ 2.33. Palpal tarsus length 1.50. Chelicerae very strong, with three teeth on retromargin. Sternum length 2.83, width 1.77. Spines, all short: leg I, femur d 1–1–1, p d1ap; tibia v 2–2–2; metatarsus v 2­ r1–0. II, femur = I; tibia v r1–2–2, p 0–1; metatarsus v 2­ r1–0, p 1. III, femur = I; tibia v p1­ p1–2, p 1­ d1– 1 or 0–1, r d1–1; metatarsus v 2–0–1, p and r 0–1–1, d 2ap. IV, femur d 1–1–1 or d 1– 0–1; tibia v p1­ p1–2, p 0, r d1–1; metatarsus v 2–0–1 or 2­ r1–1, p and r 0–1–1, d 2ap. Abdomen length 9.44, width 5.32. Spiracle– epigastrium 5.19, spiracle–spinnerets 1.07. Epigyne (fig. 18E, F): lateral lobes separate, depressed on anterior margin, prolonged in V­shaped marks on elevated median field. Copulatory ducts irregular, contorted before reaching spermathecae, ducts of accessory bulbs long, thick, converging; some of their gland ducts discharging before expansion of bulb.

MALE ( Namibia): Total length 8.25. Carapace length 3.60, width 2.67. Length of tibia/ metatarsus: I, 2.97/2.83; II, 2.93/2.73; III, 2.47/2.23; IV, 2.50/2.30. Chelicerae smaller than those of female, teeth evenly spaced. Spines as in female, except: leg I, tibia p 1– 0–1. II, tibia v 2–2–2, p 1­ d1–1; metatarsus v 2–2–0. III, tibia v 2–2–2, p and r 1­ d1–1; metatarsus v 2­ r1–1 or 2–2–1, p and r d1–1– 1. IV, femur d 1–1–1, p d1ap; metatarsus v 2­ r1–1, p and r d1–1–1. Palp: tibia short, width/length 1.07, RTA long, sharp, forming subtle angle at base. Cymbium with weak apical notch where median apophysis fits, and short, rounded basal projection opposing tibia. Copulatory bulb (fig. 18G–I): sperm duct thick, suddenly narrowed in front of tegular notch. Embolus with basal process flat, thin, striated longitudinally. Median apophysis long, apical. Paramedian apophysis with flattened portion close to median apophysis, wide, sclerotized, bearing three separate cusps, and weakly sclerotized portion close to primary conductor, projecting as flattened, triangular cusp. Primary conductor with open canal, bifurcating in elongate tips; canal ending on dorsal tip. Secondary conductor compressed, partly fused to anterior dorsal margin of tegulum, its membranous apex with vaguely defined canal.

NATURAL HISTORY: Lamoral (1968) made a detailed ecological and physiological study, comparing A. africana with the sympatrid and also intertidal Desis formidabilis ( Desidae ). He found that A. maritima builds retreats using molusc shells, or only silk, in rock crevices, which endure daily periods of immersion as the tides rise. They resist long periods of immersion, taking oxygen from the water through an air film retained by hydrophobic hairs. They resist at least 12 hours of immersion after the air film has disappeared. The spiders are nocturnal and seem to prey mostly on isopod and amphipod crustaceans. The rhythm of silk nest building remained coordinated with the tides up to a week in the laboratory.

DISTRIBUTION: South Africa.

OTHER MATERIAL EXAMINED: NAMIBIA: Luderitzbucht , intertidal rocks (26°35̍S, 15°10̍E), 8–10.X.1984, C. Griswold and T. Meikle Griswold, 13 3♀ 2 immatures ( CAS). SOUTH AFRICAN REPUBLIC: Western Cape: Cape Peninsula (34°08̍S, 18°20̍E), intertidial rocks, 1966, B. Lamoral, 13 1♀ ( CAS) ; Cape of Good Hope , in rock crevices though white silken tubes, upper edge of average high tide, 6.II.1991, V. and B. Roth, 2♀ ( CAS) ; Kommetjie , 34°9̍S, 18°20̍E, 30 air km S of Cape Town, intertidial zone, under rocks, 13.III.2001, L. Prendini, D. Ubick, 3♀ 3♀ penultimates, 1♀ ( CAS) ; 3 mi S Port Nolloth , 1 m, 5.I.1967, E.S. Ross and K. Lorenzen, 1♀ ( CAS) .

Amaurobioides chilensis (Nicolet) ,

new combination

Figures 19 , 20

Clubiona chilensis Nicolet, 1849: 419 (female presumably holotype from Chile, no specific locality, in MHNP 4231, examined).

Cluilius chilensis : Simon, 1889: 220.

Philisca chilensis : Simon, 1897a: 86, 1904: 48.

Amaurobioides cf. boydi : Ramírez, 1995a: 366 (misidentification).

NOTE: The specimens examined in Ramírez (1995a), misidentified as Amaurobioides cf. boydi (here synonymyzed with Axyracrus elegans ), are provisionally identified here as A. chilensis . The species was not included in the cladistic analysis, because the females are almost identical to those of A. maritima , and the males are unknown. There are some differences in the spermathecae between the type of A. chilensis and the specimens I collected in Chile, and it is possible that more than one species is involved. Opell (American Arachnological Society Annual Meeting, Keene, 2001) reported problems in species delimitation for New Zealand Amaurobioides .

DIAGNOSIS: Very similar to A. maritima , distinguished by the less curved epigynal lateral lobes (fig. 20B).

FEMALE (holotype, fig. 20A, spines from Cruz Grande MACN­Ar 9848): Total length ca. 8.68 (abdomen deteriorated). Carapace length 3.48, width 2.22, wider at leg II. Length of tibia/metatarsus: I, 1.75/1.70; II, 1.70/1.62; III, 1.30/1.40; IV, 1.65/1.65. Chelicerae strong, with three teeth on retromargin. Sternum length 1.91, width 1.11. Spines (six females): femora I–IV d 1–1–1, p d1ap or p 0. Leg I tibia v 2–2–2; metatarsus v 2bas. II, tibia v 2–2–2, p 0–1; metatarsus v 2bas, p 1–0. III, tibia v 0­ p1–2 or p1–2–2, p 1­ d1–1 or d1–1 or 0–1, r d1–1 or 0–1; metatarsus v 2–0–1 or 2­ p1–1, p d1–1–1 or 0­ d1–1, r 0­ d1–1, d 0­ p1–2. IV, tibia v p1­ p1–2, r d1–1 or 0; metatarsus v p1­ p1–1 or 2­ p1–1, p 0­ d1–1 or 1ap, r d1–0–1 or 1ap, d r1ap. Abdomen badly preserved, spiracle– epigastrium 2.77, spiracle–spinnerets 0.43. Color: type with abdomen totally faded. Fresh specimens are very similar to those of A. maritima ( Forster, 1970: fig. 463). Epigyne (fig. 20B, C): lateral lobes separate, elevated at anterior margin. Median field weakly sclerotized. Copulatory ducts irregular, contorted before reaching spermathecae, ducts of accessory bulbs short, thick, diverging.

MALE: Unknown.

NATURAL HISTORY: Collected from retreats made of white, very dense silk in rock crevices at the seashore in the spray zone (fig. 19).

DISTRIBUTION: The type lacks a precise locality. Collected in two widely separate localities at Chilean seashore, probably with a much more extensive distribution.

OTHER MATERIAL EXAMINED: CHILE: Región IV: Elqui : coast 6 km S Cruz Grande, 11.XI.1993, 29°29̍S, 71°19̍W, N. Platnick, K. Catley, M. Ramírez, T. Allen , 8 immatures ( MACN­Ar), 6 immatures 2nd stage ( MACN­Ar), 1♀ ( MACN­Ar 9848 ), 2♀ ( MACN­Ar 9849 , photos MJR 1324–1326), 1♀ ( MHNS), 2♀ ( AMNH) ; same, 9 km S Cruz Grande, 5 immatures ( MACN­Ar). Choapa : 12 km S Los Vilos, Rt. 5, km 213, elev. 5 m, 33°00̍S, 71°31̍W, 13.XI.1993, N. Platnick, K. Catley, M. Ramírez, T. Allen , 1 immature ( MACN­Ar). Región X: Llanquihue: 30 km E Puerto Montt, 41°36̍S ,

TABLE 8 Autapomorphies of Axyracrus elegans

72°42̍W, N. Platnick, K. Catley, M. Ramírez, T. Allen , 1♀ (AMNH).

AXYRACRUS SIMON

Table 8

Axyracrus Simon, 1884: 140 (type species by original designation Axyracrus elegans Simon,

1884), 1887: E23, 1897 a: 90, 96–98. Keyserling, 1891: 83. Ramírez, 1995a: 381, 1997: 178.

Figures

Fig. 9. Female Malenellinae and Anyphaeninae. A. Malenella nana Ramírez (Concepción, Cerro Caracol, photo MJR 137). B. Otoniela adisi Brescovit (Buenos Aires, Atucha, photo MJR 342). C. Xiruana hirsuta (Mello­Leitão, 1938) (Entre Ríos, El Palmar, photo MJR 276).

Fig. 18. Amaurobioides spp. A–D. A. maritima O.P.­ Cambridge (New Zealand, Otago). A. Cleared epigyne, dorsal view. B. Male palp, cymbium and petiolus, bulb dissected, ventral view. C. Same, retrolateral view. D. Male copulatoy bulb, ventral­prolateral view. E–I. A. africana Hewitt (Namibia, Luderitzbucht). E. Cleared epigyne, dorsal view. F. Epigyne, ventral view. G. Male copulatory bulb, ventral view. H. Same, retrolateral­apical view. I. Same, apical­dorsal view. Scale bars = A, E, 0.2 mm; B–D, F–I, 0.5 mm. (C1 = primary conductor; C2 = secondary conductor; E = embolus; MA = median apophysis; PBE = process on base of embolus; PMA = paramedian apophysis.)

Fig. 13. Female left tarsus IV, setae and sensilla. A–D. Amaurobioides africana Hewitt, retrolateral claw tuft shaved (South Africa, Western Cape, Kommetjie). A. Tarsus, dorsal view. B. Same, distal view. C. Same, retrolateral view. D. Same, trichobothria. E–G. Gayenna americana Nicolet (Talca, Alto de Vilches). E. Tarsal organ. F. Trichobothria. G. Claw tuft.

Fig. 15. Amaurobioides africana Hewitt, female spinnerets (South Africa, Western Cape, Kommetjie). A. Spinnerets. B. Left anterior lateral spinneret. C. Left posterior median spinneret. D. Left posterior lateral spinneret. (Ac = aciniform gland spigot; mAmp = minor ampullate gland spigot; MAmp = major ampullate gland spigot; Pi = piriform gland spigots; Tp = tartipore.)

Fig. 19. Amaurobioides chilensis (Nicolet), Elqui, Cruz Grande. A. Habitat. B. Female retreat. C. Female. (Photos MJR 1324–1326.)

Fig. 20. Amaurobioides chilensis (Nicolet). A. Female (holotype). B. Epigyne, ventral view (holotype). C. Cleared epigyne, dorsal view (Elqui, Cruz Grande). Scale bars = A, 2 mm; B, C, 0.2 mm.