The velvet spiders: an atlas of the Eresidae (Arachnida, Araneae)

Abstract The family Eresidae C. L. Koch, 1850 is reviewed at the genus level. The family comprises nine genera including one new genus. They are: Adonea Simon, 1873, Dorceus C. L. Koch, 1846, Dresserus Simon, 1876, Eresus Walckenaer, 1805, Gandanameno Lehtinen, 1967, Loureedia gen. n., ParadoneaLawrence, 1968, Seothyra Purcell, 1903, and Stegodyphus Simon, 1873. A key to all genera and major lineages is provided along with corresponding diagnoses, as well as descriptions of selected species. These are documented with collections of photographs, scanning electron micrographs, and illustrations. A new phylogeny of Eresidae based on molecular sequence data expands on a previously published analysis. A species of the genus Paradonea Lawrence, 1968 is sequenced and placed phylogenetically for the first time. New sequences from twenty Gandanameno Lehtinen, 1967 specimens were added to investigate species limits within the genus. The genus Loureedia gen. n. is proposed to accommodate Eresus annulipes Lucas, 1857. Two species, Eresus semicanus Simon, 1908 and Eresus jerbae El-Hennawy, 2005, are synonymized with Loureedia annulipes comb. n. One new species, Paradonea presleyi sp. n. is described. Eresus algericus El-Hennawy, 2004 is transferred to Adonea Simon, 1873. The female of Dorceus fastuosus C. L. Koch, 1846 is described for the first time. The first figures depicting Paradonea splendens (Lawrence, 1936) are presented.


Introduction
The Eresidae, commonly known as velvet spiders (Dippenaar-Schoeman and Jocqué 1997; Dippenaar-Schoeman and Van den Berg 1988), comprise nearly 100 known species organized into nine genera, one of them newly described. Most genera are found principally in arid areas of Africa and Eurasia although some species are found in rainforests of the Afrotropical and Neotropical regions (Platnick 2011). Many eresids are cryptic sit-and-wait predators in deserts (Dippenaar-Schoeman 1990;El-Hennawy 2002). Members of the genus Stegodyphus Simon, 1873 typically build silken nests in vegetation ( Fig. 4J-L) while other eresids typically live in silk tubes under objects (e.g., bark, stones) or underground ( Fig. 4A-I). Stegodyphus exhibit varying degrees of solitary and subsocial behavior, and at least three species have independently evolved quasisocial behavior (Johannesen et al. 2007; Kraus and Kraus 1988;Seibt and Wickler 1988). Quasisocial means that these spiders live in groups throughout their lives and are non-territorial within the colony (Avilés 1997). Quasisocial Stegodyphus colonies may contain hundreds of closely related individuals that participate in dramatic mass attacks on prey ( Fig. 3E-F) and form biotic islands with a characteristic fauna of kleptoparasites, parasites and inquilines (Fig. 3E, H;Griswold and Meikle-Griswold 1987;Griswold and Meikle 1990;Henschel et al. 1996). The bright red and black males of European species of Eresus Walckenaer, 1805, colloquially known as ladybird spiders, are among the most beautiful spiders in Europe, if not the world (Fig. 2B, D). In spite of their superficial resemblance to jumping spiders (Salticidae) and palp-footed spiders (Palpimanidae), or perhaps because of it, the phylogenetic placement of eresids has long been problematic. Although some genera have been revised in recent decades (Dippenaar-Schoeman 1990;El-Hennawy 2002;Kraus and Kraus 1988), the limits and distinguishing features of most genera are not well understood. In this paper we review the taxonomic and phylogenetic history of eresids, briefly summarize the biology of the family, redescribe the known genera and describe one new genus, provide diagnoses for these, provide the first key to the genera of Eresidae since Simon (1892), and present a new phylogenetic hypothesis based on molecular sequence data for all genera.

Taxonomic and phylogenetic history of Eresidae
Taxonomic research on Eresidae goes back to 1778, when Martini and Goeze described a male from Bavaria, Germany, and named it Aranea sandaliata (Lister 1778). In the following decade, the spectacular Eresus males were described four times under different names without reference to the already described species (Olivier 1789;Petagna 1792;Rossi 1790;Villers 1789). Hahn (1821), Brullé (1832) and Koch (1836) were the first to distinguish more than one Eresus species. However, they used coloration as the only character for discrimination and elevated some infraspecific varieties to the specific level (e.g., Koch 1836;1846). Due to the remarkable sexual dimorphism, males and females of the same species were given different names (Brullé 1832;Koch 1846). Koch (1850) even established a new genus, Erythrophorus, for Eresus males. Unfortunately, insufficient descriptions and lost type material during this early phase of taxonomical study brought about considerable confusion in the nomenclature of eresids. Řezáč et al. (2008) believe that Eresus kollari Rossi, 1846 is the earliest identifiable name for the widespread species of this genus. Over the next century additional eresid genera were discovered: Dorceus C. L. Koch, 1846, Adonea Simon, 1873, Stegodyphus Simon, 1873, Dresserus Simon, 1876, Penestomus Simon, 1902, Seothyra Purcell, 1903, Magunia Lehtinen, 1967, Wajane Lehtinen, 1967and Paradonea Lawrence, 1968 Eresidae as a family-level taxon was established by Koch, who called it Eresides (Koch 1850: 70). Suffixes of family group names (e.g., -idae) were not standardized until the publication of the Règles Internationales de la Nomenclature Zoologique (International Commission on Zoological Nomenclature 1905), although it is recommended at least as far back as the non-binding Strickland Code (Strickland et al. 1843). There is some confusion over the correct date of the publication establishing Eresidae. Bonnet (1945Bonnet ( , 1956 gives the date of this publication as 1851, in contradiction with the date on the frontispiece. No supporting evidence is presented to justify the later date. By contrast, Roewer (1942Roewer ( , 1954 accepted the date of Koch's publication as 1850 in his catalog. Determining the correct publication date of certain classic works in zoology can be problematic. In a paper investigating the publication dates of some works in arachnology, Brignoli (1985) argued that the evidence for the 1850 date was stronger than the alternative.
Eresidae was historically divided into two subfamilies: Eresinae and Penestominae (Simon 1903). Penestomines have a controversial history (see Miller et al. 2010a;Miller et al. 2010b). Based on the results of a molecular phylogenetic analysis, Miller et al. (2010a) removed Penestominae from Eresidae. As currently circumscribed, Eresidae are three-clawed, cribellate spiders characterized by a subrectangular carapace, median eyes grouped together and subtended by a clypeal hood, ALE placed near the anterior lateral corners of the carapace, and a strongly recurved PER (a key to anatomical abbreviations is given in the Methods section, below). The male palp has a conductor that interacts with the embolus, but lacks a median apophysis and retrolateral tibial apophysis.
Nearly all eresids are distributed in Europe, Africa, and Asia, but records from Brazil also exist. In the original description of Eresus annulipes Lucas, 1857 (transferred  Platnick et al. (1991: 68, fig. 311) B simplified cladogram from Griswold et al. (1999: 58, fig. 1). Terminals merged into family level (normal type) or higher level (all capital type) taxa.
herein to a new genus), the author gives "Rio-Janeiro" Brazil as the locality. However, the vial with the type specimen (examined by MR) includes a label indicating that the locality is unknown ("patria ignota"). Subsequent collections indicate that this species comes from arid parts of the Mediterranean. A second Brazilian eresid, Stegodyphus manaus Kraus and Kraus, 1992 appears to be a legitimate eresid not known from any other part of the world (Kraus and Kraus 1992).  Griswold et al.(2005: 316, fig. 219) B simplified Bayesian tree from Miller et al. (2010a: 792, fig. 3). Terminals merged into family level (normal type) or higher level (all capital type) taxa.

Phylogenetic placement and arrangement of Eresidae
Presumably because of the broad carapace and ocular area and widely dispersed eyes, Eresidae was traditionally associated with families such as Palpimanidae and Salticidae. Koch (1837) placed Eresus in his Attides, and later selected this as the type genus of the family level group Eresides (Koch 1850). Koch (1850) included the genus Palpimanus, Figure 7. A-B Historical phylogenetic hypotheses of relationships among Eresidae. A tree from Lehtinen (1967: 387, fig. 13). Magunia was synonymized with Stegodyphus by Kraus and Kraus (1988); Wajane was synonymized with Penestomus and removed from Eresidae by Miller et al. (2010a;2010b) B relationships among Eresidae based on molecular phylogenetic analysis of Miller et al. (2010a: 792, modified fig. 3). "Stegodyphus" annulipes relabeled Loureedia annulipes to reflect a nomenclatural change proposed in this work and Gandanameno species epithets removed to reflect increasing uncertainty about species limits and identity in this genus. along with Dorceus, Eresus, and Erythrophorus (representing Eresus males), in his original circumscription of the family. Placement continued to vary: Doleschall (1852) and Blackwall (1861) placed Eresus in Salticidae and O. Pickard-Cambridge (1872) placed the genus in Dictynides. Eresinae was used as a subfamily by Simon (1903) and this usage as a valid family has remained stable since.
In a landmark paper in spider systematics, Lehtinen (1967) examined all known genera of eresids, described two new genera (Gandanameno and Magunia, the latter synonymized with Stegodyphus by Kraus and Kraus 1988), produced a comparative table for a wide variety of somatic and genitalic characters (tables 46 and 47), and offered a branching diagram depicting his hypothesis of phylogenetic relationships among eresid genera (Fig. 7A;Lehtinen 1967: fig. 13). Notable were his association of Dresserus with Gandanameno, genera with modified PMS, and of Dorceus with Seothyra, sand dwelling genera with telescoping ALS. Lehtinen placed the Eresidae in his higher group Zodariides, along with the Thomisoidea and Salticoidea and the diverse Zodarioidea, this latter group including taxa traditionally associated with eresids such as Palpimanidae and Zodariidae. Lehtinen (1967: 385) admitted that "limitation of this group remains rather vague" and that "it might include several polyphyletic groups." Lehtinen's Zodariides did not gain widespread acceptance, and subsequent phylogenetic analyses corroborated Lehtinen's initial misgivings about the naturalness of this group. In contrast, Levi (1982) placed Eresidae in the Eresoidea, including Eresidae, Hersiliidae and Oecobiidae, a suggestion followed by Coddington and Levi (1991) and corroborated by the analyses of Griswold et al. (1999) and Griswold et al. (2005). Wunderlich (2004) proposed a considerably different concept of Eresoidea excluding Hersiliidae and Oecobiidae and including Archaeidae (including Mecysmaucheniidae), Huttoniidae, and Palpimanidae (including Stenochilidae), plus the extinct Lagonomegopidae and Spatiatoridae. Although Wunderlich did not present a matrix-based phylogenetic analysis, he did exhibit tree-based thinking in the organization of his character evidence. Wunderlich (2004: 761) specified the following characters as apomorphies of his Eresoidea: a large raised cephalic region, rugose and heavily sclerotized prosoma, strong front legs, wide eye field, widely spaced median and lateral eyes, loss of dorsal and lateral leg bristles, reduced cheliceral teeth, and small male and female palpi. He also noted that the conformation of the palpal bulb is basically similar in Eresidae, Palpimanidae, and Spatiatoridae, characterized by a protruding (sub)tegulum, an embolus and conductor typically originating in a distal position and directed to the tip of the cymbium, and the absence of a median apophysis. For Eresidae (including Penestomidae), Wunderlich (2004: 761) specified the following apomorphic characters: entelegyne female genitalia with only one pair of spermathecae, a short clypeus, median cheliceral keel, maternal feeding of offspring, molting of adults at least in females, and (excluding Penestomidae) sexual size dimorphism.
After the widespread adoption of cladistic reasoning in Arachnology and advent of matrix based comparative data and computer assisted analyses, eresid placement and circumscription continued to evolve. Coddington (1990a;see also Coddington 1990b) became the first to include an eresid exemplar in a quantitative analysis. This analysis, designed primarily to test the hypothesis of orb-weaver monophyly, included the eresid Stegodyphus. This genus fell as outgroup to a clade comprising the Orbiculariae plus RTA clade, i.e., an amaurobiid, dictynid, psechrid and titanoecid. Oecobius, also included in the analysis, was not sister to Stegodyphus, in effect challenging Levi's Eresoidea. In contrast, Platnick et al. (1991) corroborated Levi's hypothesis: this study was the first to associate Eresidae (Stegodyphus) and Oecobiidae (Oecobius) in a quantitative cladistic analysis (Fig. 5A); synapomorphies were loss of paracribellum, MAP spigots dispersed with the PI field, and transverse ridges on the hood of the trichobothrial base. Griswold et al. (1999), using an expanded exemplar set of eresids and potential sister groups, associated Eresidae (Stegodyphus and Eresus) and Oecobiidae (Oecobius and Uroctea) as Eresoidea (Fig. 5B); Eresoidea synapomorphies were loss of paracribellum, multiple MAP spigots dispersed with the PI field, transverse ridges on the hood of the trichobothrial base, and divided cribellum. For the first time they also suggested a suite of synapomorphies for the Eresidae: a square to rectangular carapace, clypeal hood, presence of a cheliceral boss, loss of the tapetum, recurved PER, MAP shaft cuticle papillate or scaly, and loss of the MA on the palp. In keeping with Kovoor and Lopez' (1979) studies of eresid silk glands, the PMS were interpreted as having multiple mAP and CY spigots and lacking AC spigots; these features also optimized as eresid synapomorphies. Schütt (2002) examined the limits of the orb building spiders (Araneoidea) and their kin. Her study had a sparse, but broad, sampling of taxa, including Eresus as eresid exemplar. She reinterpreted eresid spinning organs and, contra Kovoor and Lopez, recognized eresids as having a brush of AC spigots. Her analysis associated Eresidae with Palpimanoidea (Palpimanus, Archaea and Afrarchaea); her taxon sampling did not allow testing of Eresoidea. Griswold et al. (2005) produced a new, expanded analysis of entelegyne relationships, and also reinterpreted many of the spigot and silk gland characters used in previous studies. Using ontogeny as the prime criterion for recognizing spigot types, they coded eresids as having AC, mAP and CY spigots. Analyses under equal weights and implied weights (Fig. 6A) both supported an Eresoidea comprising Eresidae (Eresus and Stegodyphus) and Oecobiidae (Oecobius and Uroctea). Miller et al. (2010a) carried out a study relevant to eresid phylogeny and placement that was novel in three ways: a dense sampling of eresid genera, a broad array of entelegyne taxa, and data from 4 molecular markers (28S rDNA, 18s rDNA, H3 and CO1). This analysis included representatives of all Eresoid families (Eresidae, Oecobiidae and Hersiliidae), seven of the eight eresid genera (Paradonea was unavailable), and an additional 54 exemplars from across the Entelegynae. Notable results were that: 1) Eresoidea was never corroborated: eresids grouped with the orbicularian Zygiella and the nicodamids Megadictya and Oncodamus, 2) Penestominae never grouped with Eresinae, but with zodariids, leading to the proposal of the new family Penestomidae (Fig. 6B), and 3) a phylogeny for all eresid genera except Paradonea was proposed (Fig.  7B). Association of Eresidae with Nicodamidae and Orbiculariae was anticipated and corroborated by the molecular studies of Spagna and Gillespie (2008) and of Spagna et al. (2010). Eresidae was divided into two major clades: Seothyra, Dresserus, and Gandanameno form a southern and eastern African clade. Seothyra is exclusively southern African, whereas the sister genera Dresserus, and Gandanameno occur in southern and eastern Africa. The other major clade comprises Stegodyphus, Eresus, Adonea, and Dor-ceus; Eresus, Adonea, and Dorceus form a Palearctic/Mediterranean clade; Stegodyphus is found in Africa, the paleotropics, and the Amazon.

Spinneret spigot morphology
The silk glands and spinnerets of Eresidae have been discussed previously but their interpretation remains controversial. Kovoor and Lopez (1979) studied the silk glands of the eresids Eresus kollari (as E. niger) and Stegodyphus dufouri. Peters (1992b) studied the spigots of two species of Stegodyphus and traced the origin of the fibers that composed the cribellate strands. Eresid spinnerets have been studied with scanning electron microscopy and coded in matrices by Coddington (1990a), Platnick et al. (1991), Griswold et al. (1999), Schütt (2002) and Griswold et al. (2005). These studies relied upon the position, number, fine structure and ontogeny of the spigots for their classification, and these were named assuming the glands that they served, a departure from the gland-based studies of Kovoor and Lopez. Kovoor and Lopez (1979) identified pseudoflagelliform glands in eresids. Griswold et al. (1999) and Griswold et al. (2005) asserted homology between the pseudoflagelliform gland spigots of Deinopoidea and a unique spigot on the PLS of females, which they termed the "modified spigot" (MS). These spigots were overlooked in eresids by Griswold et al. (1999), but were recognized by Griswold et al. (2005). Eresid MS are unique in being anterobasal on the PLS, far removed from the rest of the spinning field, though the MS may or may not be accompanied by flanking spigots (Figs 30D, F, 36E, 39D, 57D, 58C, 60D, 61B-C, 66D, 67D, 74B, 75F, 77D, 78D, 87D, 88D, 95D). Recognizing eresid MS spigots is therefore unproblematic. Eresids are unique in that their ampullate gland spigot (MAP, mAP) shafts have small papillae or imbricate protrusions, rendering these easy to recognize on the ALS and PMS (Figs 61A, D, 67A, C). Kovoor and Lopez (1979) further asserted that eresids have numerous ampullate and cylindrical glands but lack aciniform glands. Previous phylogenetic studies (Griswold et al. 1999) relied upon these gland data to code eresids as having numerous MAP, mAP, and CY and lacking AC, but Schütt (2002) coded eresids as having a brush of AC spigots and Griswold et al. (2005) relied upon ontogenetic data to recognize AC spigots as present. Nevertheless, distinguishing AC and CY spigots remains problematic. Whereas in "higher" entelegynes, e.g., Orbiculariae (Griswold et al. 1998), Gnaphosoidea (Platnick 1990) and the "austral cribellates" (Griswold et al. 2005) classes of spigots are both distinct and uniform, in the "lower entelegynes," e.g., Oecobiidae, Eresidae, individual spigots vary in size and form such that, even with ontogenetic information, distinguishing AC and CY remains difficult in most cases. The field of small spigots with stout shafts located on the posterior lobe of the PMS of females (not males) of Dresserus and Gandanameno seems made of obvious CY spigots (Figs 36C-D, 57C, 58E). The situation in other eresids is more puzzling. Examining the PMS of Stegodyphus, Griswold et al. (2005) found three classes of spigots in females (large, medium and small) but only the large and small in males: the large were clearly mAP, the small likely AC (occurring in males and females) and the intermediate class, found only in females, were classified as CY. This ambiguity prevails in the other genera examined here, with the exception of Dresserus and Gandanameno, and our classification of AC and CY on the posterior spinnerets must remain provisional.

Scanning electron microscopy
Specimens were critical point dried, then mounted on stubs or round-headed rivets using a combination of white glue, nail polish, and adhesive copper or aluminum tape. They were sputter coated with platinum-palladium and scanned with one or more of the following: a JEOL JSM-6335F field emission scanning electron microscope, a JEOL JSM-840A scanning electron microscope, and a FEI Inspect scanning electron microscope, all at the Natural History Museum of Denmark, or a LEO 1450VP at the California Academy of Sciences. Electron micrographs of internal female reproductive structures were accomplished by first dissolving soft tissue from dissected epigyna in pancreatin P1750 enzyme (Álverez-Padilla and Hormiga 2008).

Light microscopy
Specimens were photographed in dishes of alcohol or temporary slide mounts (Coddington 1983). Where necessary, positioning specimens was aided by the use of sand or commercial Purell Hand Sanitizer (http://www.purell.com/). Photographs were made using digital cameras mounted on microscopes, either a BK+ Imaging System fromVisionary Digital (http://www.visionarydigital.com/) based on a Canon 7D digital camera body and a K2 Infinity microscope fitted with various Infinity lenses and Nikon metallurgical objectives at the Zoological Museum, Copenhagen, a Leica M205A stereoscopic microscope equipped with a Leica DFC420 camera and Leica Applications Suite software at the Zoological Museum, Copenhagen, a Nikon DS-Ri1 driven by Nikon NIS Elements software mounted on a Leica M165 C stereomicroscope at the Netherlands Centre for Biodiversity Naturalis, Leiden, or a Leica DFC500 digital camera driven by Leica Applications Suite software mounted on a Leica MZ16A stereomicroscope at the California Academy of Sciences, San Francisco. Stacks of images from multiple focal planes were combined and edited using either Helicon Focus Professional MP or Auto-Montage Pro software version 5.03, and further processed in Photoshop CS5 to adjust color, brightness, and contrast, and remove blemishes. Female reproductive structures were cleared for images of internal structures using methyl salicylate (Holm 1979). In some cases, pancreatin was also used first to digest away soft tissue. To investigate expansion of the male palp, these were removed, boiled for 2-3 minutes in a bath of hot concentrated (92%) lactic acid solution (SIGMA-ALDRICH, Inc., St. Louis, USA), then transferred to warm distilled water where expansion took place. Expanded palpi were photographed in water and positioned using a temporary slide mount (Coddington 1983).

Measurements and conventions
PLE position is expressed as a ratio of the distance from the anterior margin of the carapace to the anterior margin of the PLE divided by total carapace length. Eye diameter measured to the outside margin (analogous to measuring the cornea rather than the iris). Carapace length measured to the straight anterior margin excluding the clypeal hood.
We use the terms horizontal axis and vertical axis to describe the configuration of the median eyes. Overlapping on the horizontal axis means that if one drew a line connecting the ventral limits of the PMEs, it would pass through the AMEs (Fig. 10G); separated on the horizontal axis means that the line would not pass through the AME (Fig. 11A). Similarly, if one drew a perpendicular line tangential to the mesal limit of a PME and it passed through the corresponding AME, we call this overlapping on the vertical axis (Fig. 11E); separated on the vertical axis means that the line would not pass through the AME (Fig. 8I). Arrangement of the eyes and eye rows is depicted schematically in Figs 8-11.
Specimen collection data are given in Appendix A. Latitude-longitude coordinate pairs inferred from labels are given in square brackets; coordinates explicitly given on labels are not in square brackets. Some coordinate pairs are taken from the Iziko South African Museum (Cape Town) collections database are so indicated.
Descriptions were made mostly based on alcohol-preserved museum specimens. Some color information can be lost from such specimens.
The following anatomical abbreviations were used in the text and figures: AC: aciniform gland spigot; AER: anterior eye row; AL: anterior lobe [applied to epigynum of Loureedia gen. n.]; ALS: anterior lateral spinneret; ALE: anterior lateral eye; AME: anterior median eye; BH: basal haematodocha; C: conductor; E: embolus; MAP: major ampullate gland spigot; mAP: minor ampullate gland spigot; MH: median haematodocha; ML: median lobe (of epigynum); MS: modified spigot; PER: posterior eye row; PI: piriform gland spigot; PLS: posterior lateral spinneret; PLE: posterior lateral eye; PME: posterior median eye; PMS: posterior median spinneret; S: spermatheca; SH: spermathecal head; ST: subtegulum; T: tegulum. References to figures published elsewhere are listed in lowercase type ( fig.); references to figures in this paper are listed with an initial capital ( Fig.).    Some specimens used for this research are deposited in the personal collection of Milan Řezáč (MR). Specimen record codes often incorporate collection information, but this is not always the case and can be misleading. For this reason, specimens are referenced by both the record code (if available) and collection abbreviation.

Molecular methods
PCR products were generated either in the DNA Markerpoint Lab at the University of Leiden using standard methods (see Miller et al. 2010a) or the NCB Naturalis DNA Barcoding Facility. Sequencing was performed by Macrogen (http://www.macrogen. com). DNA sequence data was added to the manual alignment described in Miller et al. (2010a). GenBank accession numbers linked to online records for all new sequence data generated for this study are given in Table 1.
The data were analyzed using MrBayes version 3.1 (Huelsenbeck and Ronquist 2001;Ronquist and Huelsenbeck 2003) under the conditions described in Miller et al. (2010a, i.e., mixed model analysis with eight data partitions, gaps treated as missing) on the Cyberinfrastructure for Phylogenetic Research (CIPRES) portal (http:// www.phylo.org/). Analysis proceeded until the standard deviation of split frequencies fell below 0.01 (after approximately 13,500,000 of 25,000,000 generations). The first 10% of generations was discarded as burnin based on evaluation in Tracer version 1.5 (Rambaut and Drummond 2007).

Data resources
We used the Pensoft IPT Data Hosting Center to expose specimen occurrence records to the Global Biodiversity Information Facility (GBIF; http://ipt.pensoft.net/ipt/resource. do?r=specimen_occurrence_data). A KML (Keyhole Markup Language) file for viewing these same specimen occurrence records interactively in Google Earth (http://earth. google.com/) plus links to species pages on the Encyclopedia of Life (http://www.eol. org/) is available as part of a Dryad data package (doi: 10.5061/dryad.qj8t7r0q).
The alignment of the molecular sequence data used for the phylogenetic analysis is available on Dryad (doi: 10.5061/dryad.qj8t7r0q). Figures showing the full phylogenetic tree (Fig. S1) and images of some specimens newly sequenced for this study (Figs S2,S3) are available as an electronic document via Dryad (doi: 10.5061/dryad. qj8t7r0q).       Phylogeny. Our phylogenetic analysis is a modest expansion of Miller et al. (2010a) and the topology is congruent with the earlier study. The additions to the new analysis are two specimens of Paradonea variegata and twenty more specimens of Gandanameno. As reported previously, Eresidae is divided into two major clades: one consisting of Seothyra, Dresserus, and Gandanameno, the other containing the remaining genera including Paradonea (Figs 51, S1). In our topology, Paradonea sits on a long branch sister to a clade consisting of Eresus, Adonea, Loureedia gen. n., and Dorceus; Stegodyphus is sister to this five-genus clade. Note that our exemplar for Paradonea is not the type species and the monophyly of this genus is uncertain. Our focus on sequencing Gandanameno was designed to elucidate species limits within the genus, in combination with morphological data (Figs 50, S2, S3). These results are discussed further in the section on Gandanameno, below.

Key to genera of Eresidae
(note: females of Paradonea striatipes Lawrence, 1968, P. splendens (Lawrence, 1936), P. parva (Tucker, 1920), and P. presleyi sp. n. are unknown) 1a Median eyes small, subequal in size, and no more than slightly overlapping on vertical axis ( Male chelicerae strongly excavated mesally (Fig. 69C). Embolic division of male palp shorter than tegular division ( Note. Adonea contains one recognized species, A. fimbriata Simon, 1873, from the Mediterranean. In addition, Eresus algericus El-Hennawy, 2004 is transferred to Adonea and may be a junior synonym of A. fimbriata. We examined syntype specimens from Algeria and Tunisia, and additional specimens from the Algeria-Morocco border and Israel. Diagnosis. Male distinguished from other eresids except Paradonea splendens by the profile of the carapace, which has the posterior part of the cephalic region overhanging the anterior part of the thoracic region (Fig. 19D); distinguished from P. splendens by several characters including the subtriangular shape of the cephalic region that is rounded posteriorly ( Fig. 19A; trapezoidal in P. splendens and straight posteri-orly, Fig. 68D) and by the mesally contiguous chelicerae ( Fig. 19C; mesally excavated in P. splendens, Fig. 68F).
Female distinguished from other eresids except Loureedia gen. n., Eresus walckenaeri Brullé, 1832, and some Paradonea species by the relatively large PME (AME/ PME ca. 0.4, Fig. 19G); distinguished from Loureedia gen. n. by the longer than wide cephalic region (wider than long in Loureedia gen. n.); from E. walckenaeri by the presence of a glabrous median lobe between the copulatory openings ( Fig. 22A; hirsute cuticle between the copulatory openings in E. walckenaeri, Fig. 42B); and from Paradonea variegata by the nearly vertical posterior margin of the cephalic region (Figs 1A, 19H; cephalic region only moderately raised in P. variegata); females of other Paradonea species are unknown. The proportions of the epigynum in Adonea, which is more than two times wider than long, further separates it from most eresids (Figs 16A, 22A).
Natural history. Known from Loess desert habitat with low shrubs, often in wadis. They build a simple vertical or inclined burrow lined by silk, often on the edge of stones. The opening is covered by a silken flap camouflaged from above by debris. Signaling threads radiate out from the edges of this roof. Prey include various epigaeic arthropods, especially beetles from the family Tenebrionidae. Prey remnants are incorporated into the roof of the burrow. Males take approximately 2-3 years to mature, females one year longer (Martin Forman, personal observation). Description. Male (Algeria-Morocco, MR012, MR): Carapace with band of white setae around margin of thoracic region and scattered patches elsewhere; cephalic region subtriangular, longer than wide, strongly raised with rounded posterior margin overhanging thoracic region; AME distinctly smaller than PME (AME/PME 0.48), median eyes slightly overlapping on horizontal and vertical axes, PME somewhat sunken into carapace; ALE tubercles present; PER slightly narrower than AER (PER/AER 0.88), PLE position on carapace 0.35; clypeal hood forms acute angle; fovea indistinct. Chelicerae contiguous mesally, with lateral boss. Legs with bands of white setae; with row of distal ventral macrosetae on metatarsus I-IV and scattered short ventral macrosetae on tibia, metatarsus and tarsus I-IV. Abdomen dark gray, nearly encircled by a band of white setae, with numerous patches of white setae dorsally, especially around sigilla (Figs 8A, B, 19A-D). Male palp with proximal-distal axis, tegulum moderately elongate, subtrapezoidal; second loop of sperm duct curves proximally away from then back to distal margin of tegulum in retrolateral view (Figs 12B,19J); conductor and embolus together form apical complex making one helical turn; conductor tapers to point; tegular division longer than embolic division; cymbium with one retrolateral and several prolateral macrosetae (Figs 12A-C, 19I, J, 20A-F).

Adonea fimbriata
Female (Wadi Mashash, Israel, MR013, HUJ): Carapace with scattered white setae; cephalic region subtriangular, longer than wide, so strongly raised as to be nearly vertical (Figs 1A,19H); AME distinctly smaller than PME (AME/PME 0.37), median eyes slightly overlapping on horizontal and vertical axes; PME somewhat sunken into      Note. Dorceus contains five recognized species previously recorded only from North Africa. The genus was revised by El-Hennawy (2002). We examined specimens of D. fastuosus C. L. Koch, 1846 from Israel and Senegal and the Tunisian holotype of D. viberti Simon, 1910, which is a junior synonym of this species. The holotype of D. fastuosus is a dry pinned specimen and was not examined. But we examined material that El-Hennawy examined for his 2002 revision (9126, AR5404, NMHN and 1237, AR 5405, NMNH) and compared to the holotype. Diagnosis. Distinguished from other eresid genera except Seothyra by the small median eyes subequal in size (Fig. 26C, G; AME/PME > 0.7), and the long, extensible ALS contrasting with reduced PLS ( Fig. 32A; although ALS may be retracted and therefore not look so long); distinguished from other eresid genera except Loureedia gen. n., some Dresserus, and Paradonea splendens by the cephalic region, which is wider than long. Male distinguished from Seothyra by the subequal legs I and II ( Fig. 26A; leg I enlarged in Seothyra, Figs 72A, 74A) and by the form of the conductor, which is a simple spiral or L-shaped hook shorter than the tegulum (Fig.  sal in Dresserus with the embolus encircling the ventral part, Figs 33I-K, 34A-D); from P. splendens by the subrectangular shape of the cephalic region, which does not overhang the thoracic region posteriorly and the mesally contiguous chelicerae (Figs 8F, 26A, C; subtrapezoidal, slightly overhanging the thoracic region, chelicerae mesally excavated in P. splendens, Fig. 68D, F). Female distinguished from Seothyra by the median lobe of the epigynum, which is as wide as long or wider with more or less straight, converging lateral margins (Figs 16B, 29C; clearly longer than wide with a central constriction in Seothyra; see Dippenaar-Schoeman 1990); from Loureedia gen. n. by the small eyes subequal in size (Fig. 26C, G) and details of the female genitalia.
Distinguishing species. Dorceus was revised by El-Hennawy in 2002. However, the quantity and quality of specimens available to him for several species was limited.
Phylogenetic affinities. Past morphological studies have placed Dorceus and Seothyra as close relatives ( Fig. 8A; Lehtinen 1967). A recent molecular study contradicted this hypothesis ( Fig. 7B; Miller et al. 2010a). Although some of us were involved in that molecular study, we do not consider the question resolved. Morphological similarities, including features of the eyes and spinnerets, remain compelling. On the other hand, Peters (1992a) pointed out morphological characteristics shared (through parallel evolution) by Seothyra and the distantly related sparasid Leucorchestris arenicola exclusive of the eresid Stegodyphus, particularly the large ALS, which are extensible and retractable. This morphology is apparently linked to burrowing in loose sand, which Dorceus does as well. Whether these attributes ultimately prove to be the result of shared ancestry or convergence in Eresidae remains fertile ground for future study. Natural history. Known from sand dunes in deserts with very sparse shrub, grass, and annual herb patches. Juveniles feed on their mother's corpse before dispersing (El-Hennawy 1998; cf. Fig. 3D). Males take approximately 3 years to mature, females one year longer (Martin Forman, personal observation).  Description. Male (Mashabin Sand Dunes, Israel, MR006, HUJ): Carapace with few white setae; cephalic region subrectangular, wider than long, strongly raised, with sil-very patches around some eyes; AME slightly smaller than PME (AME/PME 0.95), median eyes adjacent on horizontal axis, slightly overlapping on vertical axis; ALE tubercles absent; PER as wide as AER (PER/AER 0.99), PLE position on carapace 0.45; clypeal hood forms a nearly 90° angle; fovea moderately deep (Figs 8E, F, 26A-D, 28A). Chelicerae slightly excavated mesally, with lateral boss (Figs 26C, 28B). Legs with bands of white setae; with row of distal ventral macrosetae on metatarsus I-IV, one subdistal ventral macroseta on tibia IV, and scattered ventral macrosetae on metatarsus and tarsus I-IV, strongest and most numerous on metatarsus and tarsus IV. Abdomen gray, white dorsally with large dark heart mark (Fig. 26A).
Epigynum with curved, converging slit-like atria occupying ca. the posterior half, anterior-lateral margin a curved ridge with median septum leading to subtrapezoidal median lobe ( Distinguishing species. The taxonomic literature on Dresserus is fragmentary and rarely comparative. This genus is ripe for revision. We have been unable to confidently assign species names to the specimens used in this study. Lehtinen (1967: 231) indicated that he was working on a taxonomic revision, but none was ever published. According to him, all but one of the described species (including the type species) had been checked and verified as congeneric based on primary types and other material.
Natural history. Known from Savanna, stony semidesert, and forest habitats. They build a silken tube under stones. Prey is mainly beetles (Milan Řezáč, personal observa-  Description. Male (Manga Forest Reserve, Tanzania, ZMUC): Carapace with few white setae, mostly in thoracic region; cephalic region subrectangular, wider than long, moderately raised; AME distinctly smaller than PME (AME/PME 0.71), median eyes overlapping on horizontal axis, separated on vertical axis; ALE placed on pointed apophyses, PER slightly wider than AER (PER/AER 1.04), PLE position on carapace 0.27, clypeal hood forms obtuse angle, fovea deep. Chelicerae slightly excavated mesally, with lateral boss. Legs with white setae, with row of distal ventral macrosetae on metatarsus II-IV. Abdomen dark gray, nearly encircled by a band of white setae (Figs 8I, J, 33A-D).
Male palp with dorsal-ventral axis; tegulum disc-shaped; conductor arises on broad membranous stalk from center of tegulum, with arching distal and proximal arms covering much of mesal portion of palpal bulb, distal arm larger than proximal, fringe Female (Mazumbai, Tanzania, CASENT 9025747, CAS): Carapace without conspicuous white setae; cephalic region subrectangular, about as wide as long, moderately raised; AME distinctly smaller than PME (AME/PME 0.41), median eyes overlapping  Epigynum with pair of longer-than-wide atria on posterior margin separated by hirsute cuticle (Figs 16C, 37D). Vulva with copulatory ducts making one loop leading to anterior complex of spermatheca and spermathecal head. Fertilization duct runs posteriorly through the copulatory duct loop (Figs 16F, 37E).  39B). Female PMS longitudinally elongate, transversely bilobed, with 2 anterior mAP, between these 2-3 AC, posterior to this on anterior and posterior lobes a dense field of more than 105 short, squat, conical CY spigots (Figs 36C, D, 37B, C); male PMS small, oval, with 2 anterior mAP and 3 AC (Fig. 39C). Female PLS with anterobasal MS without accompanying spigot and distal field of 9 AC (Fig. 36E); male same ( Fig.  39D, E). Male cribellar plate with no sign of spigots; epiandrous gland spigots present (Fig. 39F). Note. Eresus contains 23 recognized species group names (including 6 subspecies) from the Mediterranean and temperate latitudes of Europe and Asia. We examined specimens of several species representing the two major morphological groups within the genus. Our first exemplar, E. walckenaeri, is a cohesive species, based on both morphological and molecular data (Johannesen et al. 2005). Our specimens were from Bulgaria and Greece (Kresna, Bulgaria, MR; Pieria, Greece, MR020, MR; Lakonia, Greece, ZMUC 00012903, ZMUC). The primary types of E. walckenaeri are probably lost, but the original description and drawings, and our use in part of nearly topotypic material, allow us to identify this species with confidence. The second major group within the genus is a complex of closely related species including E. kollari (including the junior synonyms E. cinnaberinus and E. niger) and E. sandaliatus (Martini & Goeze, 1778); we refer to this assemblage collectively as the Eresus sandaliatus group. The somatic description was based on specimens of E. kollari from Czechia (Srbsko, MR016, MR; Prague, MR007, MR); scanning electron micrographs of the male palp are from a specimen of E. kollari from Hungary (Remete Mountain, CASENT 9037134, CAS); scanning electron micrographs and some photographs of the female genitalia are from a specimen of E. sandaliatus from SE of Silkeborg, Denmark (CASENT 9039243, CAS). The spinneret spigot morphology of E. sandaliatus group species was described and scanned in Griswold et al. (2005: 24-27, figs 31-32, 33A-F, 34A-C). Řezáč et al. (2008) examined copious material from the Eresus sandaliatus group including one of the two syntypes of E. kollari.

Eresus
We found no characters that clearly and simultaneously separate Eresus females from other eresid genera, despite support from molecular data for a monophyletic Eresus (Miller et al. 2010a). Separate diagnoses for females of the two species groups appear in the following section.
Diagnosis. Male Eresus are usually recognized by their distinctive dorsal abdominal pattern, which features two pairs of large round dark patches surrounding the first and second sigilla on a field of red setae, sometimes with a pair of smaller dark patches surrounding the third sigilla (Figs 2B, D, 40A, 43A); if with a black abdomen, then recognized by the distinctive notch on the embolic conductor (cf. Fig. 43J; note that black-abdomened form known as E. tristis Kroneberg, 1875 is currently cataloged as a synonym of E. kollari, but this is not accepted by all workers, e.g., Řezáč et al. 2008: 264, who noted that it is an "obviously different species"). Females of E. walckenaeri distinguished from other eresids except Dresserus and Gandanameno by the copulatory openings, which are broadly separated by hirsute cuticle (Figs 16G, 42B; separated by a glabrous median lobe in other eresids, e.g., Figs 16B, I, 29C, 45A), distinguished from Dresserus and Gandanameno by the more posterior position of the PLE (<0.28 in Gandanameno and Dresserus, > 0.33 in E. walckenaeri); females of the E. sandaliatus group distinguished from other eresids except Paradonea variegata by the large, bulbous spermathecal head (Figs 16K, L, 45B-D), distinguished from P. variegata by the smaller size difference between the AME and PME (AME/PME > 0.5 in E. sandaliatus group, Fig. 9C; < 0.4 in P. variegata, Fig. 10G) and by the overall darker color; thick stalks leading to the spermathecal head in some Stegodyphus species can be mistaken for large, bulbous spermathecal heads (Fig. 18J) although they are in fact compact sinuous ducts (Fig. 82B); E. sandaliatus group further distinguished from Stegodyphus by the more posterior position of the PLE (> 0.33 in E. sandaliatus group; < 0.29 in Stegodyphus).
Distinguishing species. Male of E. walckenaeri distinguished from those of the E. sandaliatus group by the ribbon-like conductor with a blunt, rounded tip (Fig. 41C, D; notched in E. sandaliatus group, Fig. 44D, E) and the 1.5 helical turns of the embolus (ca. 1 helical turn in E. sandaliatus group). Various species within the E. sandaliatus group distinguished primarily by coloration and details of the conductor shape and vulva (Fig. 16H, I, K, L; see also Řezáč et al. 2008).
Natural history. Known from various non-forest warm and dry habitats. Some species build a simple vertical burrow lined with silk. The opening is covered by silken sheet camouflaged from above by debris. Signaling threads radiate out from the edges of this roof. Some species (e.g., E. walckenaeri, E. crassitibialis Wunderlich) do not dig burrows; their silken tubes lie just under stones or on (Milan Řezáč, personal observation). Description. Male (Prague, Czechia, MR007, MR): Carapace with scattered white setae; cephalic region subtriangular, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.63), median eyes adjacent on horizontal axis, slightly overlapping on vertical axis; ALE tubercles absent; PER slightly narrower than AER (PER/AER 0.89), PLE position on carapace 0.33; clypeal hood forms acute angle; fovea shallow. Chelicerae contiguous mesally, with lateral boss. Legs with clusters of white setae; with single distal ventral macroseta on metatarsus I, row of distal ventral macrosetae on metatarsus II-IV, scattered ventral macrosetae on metatarsus and tarsus II-IV, strongest and most numerous on metatarsus and tarsus IV, and retrolateral mac- Male palp with proximal-distal axis; tegulum subtrapezoidal; conductor and embolus together form apical complex making 1.5 helical turns; conductor ribbon-like with blunt, rounded tip; tegular division longer than embolic division (Figs 12J-L, 40I, J, 41A-D, F); cymbium with several prolateral macrosetae (Fig. 41E).  Female (Srbsko, Czechia, MR016, MR): Carapace with few scattered white setae; cephalic region subtriangular, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.63), median eyes separated on horizontal axis, adjacent on vertical axis; ALE tubercles absent; PER slightly narrower than AER (PER/ AER 0.87), PLE position on carapace 0.34; clypeal hood forms acute angle; fovea shallow. Chelicerae contiguous mesally, with lateral boss. Legs without conspicuous white setae; legs with row of distal ventral macrosetae on metatarsus I-IV plus additional ventral macrosetae on metatarsus and tarsus I-IV. Abdomen without conspicuous white setae (Fig. 40E-H Epigynum with pair of wide atria on posterior margin separated by hirsute cuticle (Figs 16G, 42B). Vulva with spermathecal heads on long sinuous stalks gradually transitioning to sinuous multilobed spermathecae (Figs 16J, 42D-F). Description. Male (Prague, Czechia, MR007, MR): Carapace with scattered white setae; cephalic region subrectangular with broadly rounded posterior margin, longer than wide, moderately raised; AME distinctly smaller than PME (AME/ PME 0.63), median eyes slightly separated on horizontal axis, slightly overlapping on vertical axis; ALE tubercles absent, PER slightly narrower than AER (PER/ AER 0.89), PLE position on carapace 0.39; clypeal hood forms acute angle; fovea moderately deep. Chelicerae contiguous mesally, with lateral boss. Legs with bands of white setae; with row of distal ventral macrosetae on metatarsus I-IV plus additional ventral macrosetae on tibia, metatarsus and tarsus I-IV. Abdomen red dorsally with large dark patches surrounding anterior two pairs of sigilla (Figs 2B, 9A, B, 43A-D).
Female (Srbsko, Czechia, MR016, MR): Carapace with scattered white setae; cephalic region subrectangular with broadly rounded posterior margin, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.63), median eyes slightly overlapping on horizontal and vertical axes; ALE tubercles absent; PER slightly narrower than AER (PER/AER 0.89), PLE position on carapace 0.39; clypeal hood forms acute angle; fovea shallow. Chelicerae contiguous mesally, with lateral boss. Legs with scattered white setae, with row of distal ventral macrosetae on metatarsus I-IV plus additional ventral macrosetae on metatarsus and tarsus III-IV. Abdomen without conspicuous white setae (  male with at 5 MAP and spinning field of more than 40 PI. Female PMS with anterior mAP spigots, with posterior field of more 40 small spigots of varying size and shape; male PMS with 4 mAP and 6 AC, suggesting the female may have AC and CY spigots. Female PLS with anterobasal MS with 2 accompanying spigots and distal field of about 55 AC; male MS MS with 2 accompanying spigot nubbins, with 9 AC. Male cribellar plate with no sign of spigots; numerous epiandrous gland spigots present. Note. Gandanameno contains five recognized species from eastern and southern Africa. We examined the collection holdings of several museums and most primary type speci-mens. The oldest available name, Eresus fumosus C. L. Koch, 1837, apparently lacks any type specimen (Lehtinen 1967: 235). We evaluated morphological variation and analyzed DNA sequences from 24 individuals based on previously published and new data. Descriptions are based on a female specimen from Tanzania (ZMUC 19970530, ZMUC) and a male from Zimbabwe (AcAT 2005/123, NCA) supplemented by collections mostly from South Africa.
Diagnosis. Distinguished from other eresids except Dresserus by the position of the PLE which are both advanced (< 0.28) and widely spaced (PER/AER > 0.95; Fig. 9F, H); other eresids with advanced PLE (e.g., Stegodyphus, some Paradonea) have them closer together (PER/AER < 0.90) than the ALE (e.g., Figs 10B, 11F). Male further     Tucker, 1920; type locality of Eresus fumosus C. L. Koch, 1837 is reported simply as "Afrika" and no type specimen is known (Lehtinen 1967: 235). Localities of males illustrated in Fig. 48: square A Fig. 48A-C square D Fig. 48D square E Fig. 48E square F Fig. 48F. Ellipsoids indicate regions for size chart Fig. 50, region names are for convenience only. Distinguishing species. Gandanameno are variable in several conspicuous characteristics. In the male, these include the presence or absence of a cheliceral boss (Fig.  56C-F), palp size, details of the conductor shape, and the curvature and position of the tegular sperm duct (Figs 48B, D-F, S2A-J). In the female, these include the height of the cephalic region, and the presence of cuspules of various weights on the carapace, sternum, and basal segments of legs I, II, and sometimes III (Figs 50, 52A-F, 53A-F, 54A-F), also details of the epigynum shape (Figs 17A-C, S2K, L, S3A-F, H-L). It has been noted previously that multiple forms may occur sympatrically and that intermediate combinations make species determination problematic (Tucker 1920). Indeed, in most regions where several specimens are known, a wide range of variation can be Figure 50. Carapace height plotted against carapace length for adult female Gandanameno specimens from eight regions: Bloemfontein, Cape Town, Gauteng, Hanover, Namaqualand, Namibia, and Port Elizabeth. Regions circumscribed in Fig. 49; sample size given in parentheses. Symbol shape indicates region while symbol darkness indicates presence and strength of cuspules. Specimens were scored as having cuspules absent, having medium to strong cuspules only on the legs, having a mixture of medium and strong cuspules on the prosoma, sternum, and/or legs, and having exclusively strong cuspules on the prosoma, sternum, and legs. As reflected in the legend, not all degrees of spinulation were observed in all regions. found (Fig. 50). We attempted to sequence DNA from several museum specimens. Unfortunately, most of our successes were from specimens at one end of the range of variation, i.e., specimens with a moderately to strongly raised cephalic region and abundant, heavy cuspules on the carapace, sternum, and anterior legs. The three con-trasting female specimens (13-10: KwaZulu-Natal; 18-04: AcAT 2002/181; 14517) did not group together. Instead, relationships appear to be more strongly tied to geography than to morphology (Fig. 51). The genitalia of 23 out of 24 Gandanameno  We looked for patterns of variation in adult female morphology. Adult female specimens are much more abundant in collections than males. We defined eight geo-graphic regions based on two criteria: 1) concentration of specimens available to us and/or 2) type localities of Gandanameno species (Figs 49, 50). We assessed a series of characters: carapace length, carapace width, presence and strength of cuspules on the ventral surface of femur I and II, prosoma, and sternum for all specimens available  from these regions. When cuspules are present on the prosoma, they always occur on the femurs, but the reverse is not always true. So the set of specimens with cuspules on the femurs is larger than and contains the set of specimens with cuspules on the prosoma. There appears to be a general trend towards greater spinulation and a higher carapace with increasing size (carapace length), but the spinulation data are fairly noisy (Fig. 50). No morphological pattern emerged to segregate specimens by region. Not all degrees of spinulation were observed in every geographic region.
Based on the combination of genetic, morphometric, and spinulation data, we see no evidence that the characters traditionally used to discriminate Gandanameno species are valid. We speculate that ontogenetic factors are responsible for the morphological variation in this genus (perhaps juvenile nutrition or post adult molting). However, we judge that the synonymy of all Gandanameno species is premature, particularly in light of the phylogeographic signal suggested by the molecular data and the limited avail-  ability of specimens from beyond the Republic of South Africa. We therefore identify our specimens simply as Gandanameno sp. We encourage additional investigations that might further elucidate the systematics of Gandanameno.
Natural history. Associated with crevices in or under rocks or tree bark in savanna, parks, and gardens. They build a silken tube with a widened entrance; the tube of an adult female under bark can be up to 1 m long (Martin Forman, personal observation). Also found under tree bark well above the ground in dense aggregations (Nikolaj Scharff, Jere-my Miller, Iringa, Tanzania). Prey remnants are placed at the bottom of the tube. Clutches consist of tens of juveniles (Martin Forman, personal observation). Juveniles do not feed on their mother's corpse and adult females can produce a number of sequential clutches. Males mature in less than one year, females take longer (Martin Forman, personal observation).  Male carapace with scattered white setae, cephalic region subrectangular, about as long as wide, slightly raised; AME distinctly smaller than PME (AME/PME 0.63), median eyes slightly  Male palp with dorsal-ventral axis; tegulum disc-shaped; conductor arises on stalk from center of tegulum, with arching prolateral and retrolateral arms covering much of anterior portion of palpal bulb, retrolateral arm with fringed posterior margin; embolus makes ca. three loops, long and flexible, fits into groove originating on prolateral arm of conductor; cymbium without distinct macrosetae (Figs 13D-F, 48A-F, 55A-E, S2A-J).

Loureedia
Etymology. Named for Lou Reed, leader of the rock band The Velvet Underground from 1965-1970; the gender is feminine.
Diagnosis. Distinguished from other eresid genera except Dorceus, some Dresserus, and Paradonea splendens by the cephalic region, which is wider than long (Fig. 9J, L); distinguished from Dorceus by the median eye group, which have the PME clearly larger than the AME (AME/PME ca. 0.5; Fig. 9I, K); median eyes small with the PME only slightly larger than the AME in Dorceus (Fig. 8E), AME/PME > 0.8; Figs 8E, G, 28A, 29A); distinguished from Dresserus by the lack of prominent tubercles bearing the ALE and the palpal conformation, which has a proximal-ventral axis with the helical embolus encircling the distal part ( Fig. 63B; obliquely ventral-dorsal in Dresserus with the embolus encircling the ventral part, Figs 33I-K, 34A-D); distinguished from P. splendens by the subrectangular shape of the cephalic region, which does not overhang the thoracic region posteriorly (Figs 9J, 62A, D; subtrapezoidal, slightly overhanging the thoracic region in P. splendens, Fig. 68D). Male further distinguished from other eresids except Stegodyphus dumicola, S. tentoriicola, and Paradonea striatipes by the strongly bifid conductor (Fig. 63D, E); separated from these species by several characters including details of the conductor shape, the shape of the cephalic region, the lack of ALE tubercles, and a striking abdominal pattern of white and red patches on a black field (Fig. 1G, H). Female further distinguished by the epigynum with its unique anterior depression and by the compact configuration of the reproductive duct system (Figs 18A, D, 65A, B).  Natural history. Known from Loess desert habitat with low shrubs, often in wadis. They build a simple vertical or inclined burrow lined by silk. The opening is covered by silken sheet camouflaged from above by debris. Signaling threads radiate out from the edges of this roof. Mating occurs in late autumn. Prey remnants are incorporated into the roof of the burrow. Juveniles feed on their mother's corpse before dispersing (cf. Fig. 3D). Males take approximately 3 years to mature, females one year longer (Martin Forman, personal observation).   Eresus annulipes Lucas, 1857: 21. Eresus semicanus Simon, 1908: 83;1910: 294, fig. 5;El-Hennawy 2004: 28, figs 2A-B, 3A-C, 4A-B. Syn. n. Stegodyphus annulipes (Lucas, 1857). Kraus 1992: 15, 19. Eresus jerbae El-Hennawy, 2005: 88, figs 1-4. Syn. n.
Description. Male (Nitzanna village, Israel, MR018, HUJ): Carapace with scattered white setae; cephalic region subrectangular with broadly rounded posterior margin, wider than long, strongly raised; AME distinctly smaller than PME (AME/PME 0.52), median eyes slightly overlapping on horizontal and vertical axes; ALE tubercles absent; PER as wide as AER (PER/AER 1.01), PLE position on carapace 0.42; clypeal hood forms a nearly 90° angle; fovea moderately deep. Chelicerae slightly excavated mesally, with lateral boss. Legs relatively long with bands of white setae; legs with row of distal ventral macrosetae on metatarsus I-IV, and scattered ventral macrosetae on tibia III-IV and metatarsus and tarsus II-IV, strongest and most numerous on metatarsus and tarsus III-IV. Abdomen dark with pattern of white spots surrounding sigilla and two longitudinal yellow stripes running through sigilla (Figs 1G, H, 9I, J, 62A-D).
Spinneret spigot morphology. Our female preparation is in poor condition but intact; the male preparation is in very poor condition making it impossible to provide accurate counts, especially on the PMS. Female ALS with at least 5 MAP within and on inner edge of spinning field of more than 23 PI (Figs 66B, 67A); male with MAP and PI, but number can't be determined. Female PMS with 4 anterior mAP spigots and posterior field of 22 spigots of varying size and shape (Figs 66C, 67B, C); male seems to have fewer spigots than female, suggesting the female may have AC and CY spigots. Female PLS with anterobasal MS with two accompanying AC spigots and distal field of at least 19 AC (Figs 66D, 67D); male appears to have MS and flanking AC, with field of more than 8 AC. Male cribellar plate with no sign of spigots; numerous epiandrous gland spigots present (Fig. 65F).

Paradonea
Lawrence http://species-id.net/wiki/Paradonea Paradonea Lawrence, 1968: 116. Type species Paradonea striatipes  Note. Paradonea contains four recognized species from southern Africa (Fig. 71). We examined specimens (including the primary types) of all four described species and propose one new species. Of these five, only one (P. variegata) is known from the female as well as the male. Two species are particularly remarkable in their morphological distinctness: the type species, P. striatipes, is large with distinct markings, an enlarged first leg with tibial brush, relatively small palpi with a distinctly bifid conductor, widely spaced median eyes, and PLE set near the front of the carapace; P. splendens has a strongly raised cephalic region that occupies most of the cephalothorax and slightly overhangs the thoracic region, but the palpal morphology is fairly typical of Eresidae. Despite our skepticism about the monophyly of this group, we are not proposing nomenclatural changes at this time. We hope that this work may help catalyze new activity in this group, including the discovery of females and the collection of fresh specimens for DNA sequencing. Our phylogenetic analysis includes P. variegata, the first Paradonea species to be sequenced and the first time Paradonea has been placed phylogenetically (Figs 51, S1). We do not offer a diagnosis of the genus but distinguish each species independently in the key and species diagnostic sections.
Spinneret spigot morphology was not investigated for any species in the genus. ( Fig. 68H). Distinguished from other eresids except P. parva, P. presleyi sp. n., Seothyra, and some Stegodyphus by the enlarged leg I (Fig. 68A, B), distinguished from P. parva, P. presleyi sp. n., and Seothyra by the presence of a dense brush of setae, especially on the tibia (Fig. 68A, B); distinguished from P. presleyi sp. n. and Stegodyphus by the separation of the median eyes on the vertical axis ( Fig. 10A; broadly overlapping in P. presleyi sp. n. and Stegodyphus, Figs 11E, 70I). Paradonea striatipes has the PLE in a more advanced position (ca. 0.25) than most other eresids ( Fig. 11B; Dresserus, Gandanameno, and Stegodyphus may also have the PLE around 0.25). The markings, especially the distribution of white setae, are unique ( Fig. 68A-C). The palpi are relatively small proportional to body size compared to other eresids (Fig. 68B). Description. Male (Outjo Namibia, NMBA05700, BMSA): Carapace with broad band of white setae around margin and between AME; cephalic region subtriangular, longer than wide, strongly raised; AME distinctly smaller than PME (AME/PME 0.33), median eyes widely separated on horizontal axis, adjacent on vertical axis; ALE on distinct tubercles; PER much narrower than AER (PER/AER 0.82), PLE position on carapace 0.25; clypeal hood forms acute angle; fovea shallow. Chelicerae with lateral boss, basal three quarters covered in white setae, contiguous mesally. Legs with bands of white setae, especially dorsally along the length of most segments; leg I somewhat thickened and elongated, tibia I with brush of dark setae; with scattered ventral macrosetae on tibia II-IV and metatarsus and tarsus I-IV. Abdomen black with series of irregular transverse stripes formed by thick patches of white setae (Figs 10A, B, 68A-C).

Paradonea striatipes
Male palp with proximal-distal axis; tegulum moderately elongate, subtrapezoidal; second loop of sperm duct follows complicated path featuring multiple switchbacks; conductor and embolus together form apical complex making 1.5 helical turns; conductor with conspicuous bifid apophysis arising from retrolateral side, consisting of a spine-like dorsal branch curving distally for nearly 180° and a broad, flattened ventral branch with several small sharp processes along the ventral and distal margins; tegular division longer than embolic division; cymbium with a few mesosetae (only slightly thicker than normal setae) over dorsal to prolateral surface (Figs 12J, K, 68G, H).

Diagnosis.
Distinguished from other eresids except Dorceus, some Dresserus, and Loureedia gen. n. by the wider than long cephalic region (Figs 10D, 68D), distinguished from these by the subtrapezoidal shape of the cephalic region, much wider anteriorly than posteriorly (Figs 10D, 68D; subrectangular in Dorceus, Dresserus, and Loureedia gen. n. with little difference in width anteriorly to posteriorly, e.g., Figs 8F, 9J); distinguished from other eresids except the male of Adonea by the posterior margin of the cephalic region, which slightly overhangs the thoracic region; distinguished from many eresids including Adonea, Dorceus, Dresserus, and Loureedia gen. n. by the mesally excavated chelicerae ( Fig. 68F; contiguous or only slightly excavated in Adonea, Dorceus, Dresserus, and Loureedia gen. n.). Description. Male (Sunnyside, South Africa, C1076, SAM): Carapace with band of white setae around margin of thoracic region, additional white setae form several patches including one on the posterior part of the cephalic region; cephalic region subtrapezoidal, wider than long anteriorly, posterior margin straight, strongly raised, slightly overhanging thoracic region posteriorly; AME distinctly smaller than PME (AME/PME 0.40), median eyes separated on horizontal axis, some overlap on vertical axis; ALE tubercles absent; PER narrower than AER (PER/AER 0.86), PLE position on carapace 0.44; clypeal hood forms acute angle; fovea moderately deep. Chelicerae with lateral boss, excavated mesally. Legs with bands of white setae, especially dorsally along the length of most segments; with row of distal ventral macrosetae on metatarsus I-IV and scattered ventral macrosetae on tibia, metatarsus and tarsus I-IV. Abdomen black with longitudinal stripe with uneven margins formed by a thick concentration of white setae (Figs 10C, D, 68D-F; note description of pattern of setae and color on carapace, legs, and abdomen based in part on specimens other than C1076, which is missing most setae; see Appendix A) Male palp with proximal-distal axis; tegulum bulbous; conductor and embolus together form apical complex making one helical turn; conductor moderately sclerotized, tip a blunt point; tegular division longer than embolic division; cymbium with several macrosetae over dorsal to prolateral surface (Figs 13L, 14A, 68I, J).

Diagnosis.
Male distinguished from other eresids except P. presleyi sp. n. by the conductor, which bears a helical ridge fringed with distinct papillae (Figs 14B, C, 69G, H); distinguished from P. presleyi sp. n. by the mesally excavated chelicerae ( Fig. 69C; contiguous in P. presleyi sp. n., Fig. 70I), the dorsal abdominal pattern (Fig. 69A), the median eye group, which has only slight overlap on the vertical axis (Figs 10E, 69C; significant overlap in P. presleyi sp. n., Fig. 70I), and the relatively more narrow and long shape of the conductor (Fig. 14B, C; compare with Fig. 14J, L). Female distinguished from other eresids except members of the Eresus sandaliatus group by the large, bulbous spermathecal head (Fig. 18E); distinguished from the E. sandaliatus group by the larger size difference between the AME and PME (AME/PME ca. 0.5 in P. variegata, Fig. 69F; >0.6 in E. sandaliatus group, Fig. 9C) and by the overall lighter color (Fig. 69D; compare with Fig.  43E). Note that interpretation of the female reproductive duct characters for P. variegata is based on light microscopy, not SEM, and must therefore be regarded as tentative.
Description. Male (Breekkierie Dunes, South Africa, C1062, SAM): Carapace with some white setae, especially around margin of thoracic region, cephalic region subtriangular, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.52), median eyes adjacent on horizontal axis, some overlap on vertical axis; ALE tubercles absent; PER much narrower than AER (PER/AER 0.86), PLE position on carapace 0.31, clypeal hood forms a slightly less than 90° angle; fovea shallow. Chelicerae with lateral boss, excavated mesally. Legs with some white setae; with row of distal ventral macrosetae on metatarsus I-IV and scattered ventral macrosetae Male palp with proximal-distal axis; tegulum bulbous; conductor and embolus together form apical complex running more or less distally; conductor moderately sclerotized with helical ridge fringed with distinct papillae, hooked distally; tegular division longer than embolic division; cymbium with several prolateral macrosetae (Figs 14B, C, 69G, H).
Female (Steinkopf, South Africa, ZMB 26964, ZMHB): Carapace with patches of white setae; cephalic region subtrapezoidal, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.48), median eyes with some overlap on horizontal axis, some overlap on vertical axis; ALE on small tubercles; PER slightly narrower than AER (PER/AER 0.93), PLE position on carapace 0.31; clypeal hood forms acute angle; fovea moderately deep. Chelicerae contiguous mesally, with lateral boss. Legs without conspicuous white setae; legs with row of distal ventral macrosetae on metatarsus I-IV and numerous ventral macrosetae on metatarsus and tarsus II-IV. Abdomen with numerous white setae (Figs 10G, H, 69D-F).
Natural history. Known from savanna and semiarid desert. They build silken tubes under stones or under shrubs. Sometimes, spiders build a round web approximately 10 cm in diameter that may be covered with sand and herbal debris. Juveniles feed on their mother's corpse before dispersing (cf. Fig. 3D). Adults appear around December, juveniles disperse in October (Martin Forman, personal observation).
Note. The holotype specimen is bleached and damaged (Figs 14D-F, 70A-C). The description is based mostly on better preserved, putatively conspecific individuals. However, there are morphological differences including carapace shape (compare Fig.  70C with Fig. 70F) and the shape of the palpal conductor (compare Fig. 14E with Fig.  14H), which could be taken to mean that there is more than one species in this complex. But for now, we attribute these differences to preservation artifacts. The discovery and examination of more fresh specimens would be helpful in resolving this question. For now, we present photographs of both the holotype (Figs 14D-F, 70A-C) and a more well-preserved specimen (Figs 14G-I, 70D-F).
Description. Male (4 km N of Hopetown, South Africa, AcAT 97/988, NCA): Carapace with white setae concentrated in thoracic region and posterior of cephalic region, and forming two longitudinal lines connecting the lateral eyes; cephalic region subtriangular, longer than wide, slightly raised; AME distinctly smaller than PME (AME/ PME 0.45), median eyes slightly overlapping on horizontal and vertical axes; ALE tubercles absent; PER slightly narrower than AER (PER/AER 0.94), PLE position on carapace 0.35; clypeal hood forms a slightly less than 90° angle; fovea indistinct. Chelicerae Figure 71. Distribution of Paradonea species. Circles, P. striatipes; squares, P. splendens; inverted triangles, P. variegata; diamonds, P. parva; triangles, P. presleyi sp. n. Primary type localities with all black symbols, those for other localities with white center.
slightly excavated mesally, with lateral boss. Legs with patches and bands of white setae; leg I slightly thickened; with row of distal ventral macrosetae on metatarsus I-IV, a few scattered ventral macrosetae on tarsus I-IV and metatarsus III-IV; leg I slightly thickened. Abdomen with two longitudinal stripes of white hairs ectal to sigilla (Fig. 70D-F).
Female: Unknown. Diagnosis. Distinguished from other eresids except Stegodyphus by the median eye group, which has large, subequal eyes (AME/PME > 0.6) with clear separation on the horizontal axis and significant overlap on the vertical axis (Fig. 70I); distinguished from Stegodyphus by the short, slightly obtuse clypeal hood (long and acute in Stegodyphus, Fig.  11A, E, I). Male distinguished from other eresids by the distinctive dorsal abdominal pattern (Fig. 70G). Further distinguished from other eresids except P. variegata by the conductor, which bears a helical ridge fringed with distinct papillae (Fig. 14J, L); distinguished from P. variegata by the contiguous chelicerae ( Fig. 70I; mesally excavated in P. variegata, Fig. 69C), by the median eye group, which has significant overlap on the vertical axis ( Fig.  70I; slight overlap in P. variegata, Fig. 69C), and by the proportions of the conductor, which are wider and shorter than in P. variegata (Fig. 14J, L; compare with 14B, C).
Description. Male (Kruger National Park, South Africa, AcAT 2008/4480, NCA): Carapace with white setae concentrated in thoracic region and eye region; cephalic region semicircular, wider than long, moderately raised; AME smaller than PME (0.57), median eyes slightly separated on horizontal axis, significant overlap on vertical axis, ALE tubercles absent, PER width narrower than AER width (0.87), PLE position on carapace 0.29; clypeal hood forms a slightly more than 90° angle; fovea shallow. Chelicerae contiguous mesally, with lateral boss. Legs with patches and longitudinal bands of white setae; femur I slightly thickened with thick brush of dark setae; with row of distal ventral macrosetae on metatarsus I-IV, a few scattered ventral macrosetae on tarsus I-IV and metatarsus II-IV. Dorsum of abdomen with two longitudinal stripes of white hairs more or less parallel anteriorly, diverging posteriorly, then connected by transverse portion, median part medium brown, ectal and posterior part dark brown (Fig. 70G-I).

Seothyra
Purcell http://species-id.net/wiki/Seothyra Seothyra Purcell, 1903: 31;Lehtinen 1967: 264;Dippenaar-Schoeman 1990: 136. Type species Seothyra schreineri Purcell, 1903. Note. Seothyra contains 13 recognized species from southern Africa and was revised by Dippenaar-Schoeman (1990). We examined specimens of S. henscheli Dippenaar-Schoeman, 1991 from Namibia. Our species identification was based on the recent revision; we did not seek to examine type material. See Dorceus for note on phylogenetic relationships. Diagnosis. Distinguished from other eresid genera except Dorceus by the small eyes subequal in size (AME/PME > 0.7, Figs 10I, K, 72C, G), and the long, extensible ALS contrasting with reduced PLS (Fig. 72D, although ALS may be retracted and therefore not look so long). Male distinguished from Dorceus by the enlarged leg I (Figs 72A,B,74A;legs I and II subequal in Dorceus,Fig. 26A), and by the form of the conductor, which is highly variable and elaborate and usually longer than the tegular division (Figs 15A, B, 72I, J, 73A-F; a simple spiral or L-shape hook and shorter than the tegular division in Dorceus, Figs 12D, E, 26I, J; see El-Hennawy 2002). Female distinguished by the median lobe of the epigynum, which is clearly longer than wide with a central constriction (Figs 18C; 76A; wider than long with more or less straight, converging lateral margins in Dorceus, Figs 16B, 29C) and by the lack of ampullate gland spigots on the ALS (Fig. 77B).
Distinguishing species. Seothyra was revised by Dippenaar-Schoeman in 1990. Natural history. Known from semi-stabilized sand dunes in semiarid desert (Dippenaar-Schoeman 1990). They build a simple straight to curved 5-15 cm deep burrow (Fig. 4H). The opening is covered by two or four lobed silk flaps covered with sand, resembling a hoof print (Fig. 4G;Lubin 1989;Peters 1992a). Prey remnants are placed at the bottom of the burrow. The spider positions itself upside down under the burrow cover (Filmer 1995). Prey are mainly ants (Dippenaar-Schoeman 1990). Eggs hatch at the beginning of summer. Juveniles feed on their mother's corpse before dispersing (cf. Fig. 3D). Mating occurs in April and May. Males are active during the day, mimicking Camponotus ants and mutillid wasps in behavior and appearance (Henschel 1989). Seothyra henscheli Dippenaar-Schoeman, 1991: 156, figs 15, 26, 37, 67-70. Description. Male (Gobabeb Station, Namibia, SMN 40828, NMN): Carapace without conspicuous white setae; cephalic region subrectangular, longer than wide, strongly raised; AME slightly larger than PME (AME/PME 1.08), median eyes adjacent on horizontal axis, some overlap on vertical axis; ALE tubercles absent; PER slightly wider than AER (PER/AER 1.18), PLE position on carapace 0.35; clypeal hood forms a slightly less than 90° angle; fovea deep. Chelicerae with small lateral boss, excavated mesally, with tooth bearing a row of four denticles adjacent to the base of the fang. Legs without conspicuous white setae; femur, patella and tibia I conspicuously thickened; with row of distal ventral macrosetae on metatarsus II-IV plus a distal ventral macroseta on tibia III and scattered ventral macrosetae on metatarsus and tarsus III-IV and tibia IV. Abdomen with many white setae dorsally (Fig. 72A-D).

Seothyra henscheli Dippenaar-Schoeman
Male palp with proximal-distal axis; tegulum bulbous; conductor and embolus together form apical complex making about two helical turns; conductor increasingly sclerotized distally terminating in a recurved hook; embolic division longer than tegular division; cymbium with several prolateral and retrolateral macrosetae, some macrosetae arising from retrolateral modified with subbasal enlargement (Figs 15A-C, 72I, J, 73A-F). Female (Gobabeb,Namibia,SMN 46627,NMN): Carapace with many stiff dark setae in the cephalic region and scattered white setae, especially in the thoracic region; cephalic region subrectangular, longer than wide, moderately raised; AME slightly larger than PME (AME/PME 1.05), median eyes adjacent on horizontal axis, some overlap on vertical axis; ALE tubercles absent; PER slightly wider than AER (PER/AER 1.09), PLE position on carapace 0.38; clypeal hood forms a slightly obtuse angle; fovea moderately deep. Chelicerae contiguous mesally, with small lateral boss. Legs without conspicuous white setae; legs with row of distal ventral macrosetae on metatarsus II-IV, a few scattered ventral macrosetae on tarsus II and numerous ventral macrosetae on metatarsus and tarsus III-IV. Abdomen with white setae mostly around the margin (Figs 72E-H, 75A, B).
Epigynum with sinuous slit-like atria occupying ca. half the total length, median lobe margin defined anteriorly and laterally by a ridge (Figs 18G, 76A). Vulva with tightly-wound sinuous stalk leading to spermathecal head at anterior end, multilobed spermathecae at posterior end (Figs 18F,. Spinneret spigot morphology. Female ALS with spinning field of more than 50 PI (Figs 75E, 77B), male with more than 30 PI (Fig. 78B); both sexes lack MAP on ALS. Female PMS with 2 anterior mAP spigots, a large posteromedian spigot (probably CY), and 8 smaller spigots (Fig. 77C); male PMS with 1 central mAP and 3 AC  (Fig. 74C, D).  established to accommodate S. tentoriicola Purcell, 1904 andS. dumicola Pocock, 1898, both of which have a bifurcated conductor including a hook-shaped sclerotized branch (Kraus and Kraus 1988: figs 245, 248). Kraus and Kraus (1988) rejected this genus on the grounds that it would require dividing Stegodyphus sensu lato into at least three genera, which they considered needless. Molecular analyses have corroborated the monophyly of Stegodyphus sensu lato and nested taxa representing Magunia within (not sister to) a monophyletic Stegodyphus (Johannesen et al. 2007;Miller et al. 2010a). We therefore concur with Kraus and Kraus (1988) in treating Magunia as a subjective junior synonym of Stegodyphus.

Stegodyphus
We studied specimens representing three Stegodyphus selected to cover a wide range of the variation present in the genus. Our exemplars were S. lineatus, the type species for the genus, based on specimens from Afghanistan, Israel, and Turkey, S. mimosarum based on specimens from Madagascar and Malawi, and S. sarasinorum based on specimens from Myanmar. Our identification was based on the revision of Kraus and Kraus (1988); we did not seek to examine type material.
Diagnosis. Distinguished from other eresids except Paradonea presleyi sp. n. by the median eye group, which has large, subequal eyes (AME/PME > 0.6) with clear separation on the horizontal axis and significant overlap on the vertical axis (Fig. 11A, C, E, G, I, K); distinguished from P. presleyi sp. n. by the long, acute clypeal hood (short, slightly obtuse in P. presleyi sp. n., Fig. 70I). Classically diagnosed by having the PER noticeably more narrow (65-90%) than the AER (Kraus and Kraus 1988;Simon 1892); however, Adonea, and some Eresus and Paradonea species approach or even overlap with some Stegodyphus species. Further distinguished from other eresids except Dresserus, Gandanameno, and some Paradonea by the advanced position of the PLE (< 0.29; > 0.31 in other eresids). Distinguishing species. The revisionary monograph by Kraus and Kraus (1988) remains the key resource for identifying Stegodyphus species. Kraus and Kraus (1988) divided Stegodyphus into three species groups, each with one non-territorial permanently social species. In a molecular phylogenetic study, Johannesen et al. (2007) suggested that S. lineatus is distinctive enough to warrant its own monotypic species group, but otherwise their findings were congruent with the earlier morphological study of  Kraus and Kraus (1988) and supported the hypothesis that sociality in Stegodyphus has been independently derived multiple times. The Stegodyphus africanus group contains seven species including our exemplar, the social species S. mimosarum. Males of the S. africanus group have the palpal conductor as a distally-projecting complex comprising an outer leaf and a longer, less sclerotized, inner leaf. The embolic division is longer than the tegular division. The epigynum has the posterior part of median lobe raised, membranous, and wider than long. The median eyes in both sexes are relatively heterogenous (AME/PME ca. 0.6-0.8; data from Kraus and Kraus 1988). The Stegodyphus dufouri group contains four or five species (depending on whether S. pacificus and S. dufouri are considered distinct) including our exemplar, the social species S. sarasinorum. Males of the S. dufouri group have a spiral conductor that lacks a free inner leaf and is shorter than the tegular division. The median eyes in both sexes are relatively homogenous (AME/PME usually ca. 0.7-1.0, rarely ca. 0.6; data from Kraus and Kraus 1988). The Stegodyphus mirandus group contains six species including our exemplar, S. lineatus, the type species of the genus (but see Johannesen et al. 2007). Males of the S. mirandus group have complex conductors with a membranous inner lobe similar to that found in the S. africanus group, although it is characteristically more spiral (as opposed to distally-projecting) and the embolic division is shorter than the tegular division. In some species (S. tentoriicola and S. dumicola), a sclerotized hook-like apophysis arises from the inner lobe of the conductor. The median eyes in both sexes are relatively homogenous (AME/PME ca. 0.8-0.9) except in S. tibialis (AME/PME ca. 0.5-0.6; data from Kraus and Kraus 1988). The epigynum of some S. mirandus group species is more or less rotated into vertical position (Kraus and Kraus 1988).
Male palp with proximal-distal axis; tegulum subtrapezoidal; conductor and embolus together form apical complex making one helical turn; conductor with more or less membranous and papilliated inner layer extending beyond moderately sclerotized outer layer; tegular division slightly longer than embolic division; cymbium with several prolateral macrosetae (Figs 15D-F, 79I, J, 80A-D).
Female (Belkis, Turkey, MR015, MR): Carapace with numerous white setae, cephalic region subtriangular, longer than wide, moderately raised; AME nearly as large as PME (AME/PME 0.87), median eyes separated on horizontal axis, largely overlapping on vertical axis; ALE on small tubercles; PER much narrower than AER (PER/ AER 0.68), PLE position on carapace 0.21; clypeal hood forms acute angle; fovea shallow. Chelicerae contiguous mesally, with lateral boss. Legs with numerous white setae, with pair of distal ventral macrosetae on tibia I-IV and row of distal ventral macrosetae on metatarsus I-IV plus scattered ventral macrosetae on metatarsus and tarsus I-IV. Abdomen with numerous white setae (Figs 11C, D, 79E-H, 81A-F).
Description. Male (Forêt d'Analalava, Madagascar, CASENT 9005869, CAS): Carapace with band of white setae around margin, longitudinal line in cephalic region and patches near PLE; cephalic region subtriangular, longer than wide, moderately raised; AME distinctly smaller than PME (AME/PME 0.62), median eyes separated on horizontal axis, largely overlapping on vertical axis; ALE on small tubercles; PER much narrower than AER (PER/AER 0.76), PLE position on carapace 0.32; clypeal hood forms acute angle; fovea shallow. Chelicerae with lateral boss, slightly excavated mesally. Legs with patches and longitudinal bands of white setae; leg I thickened with thick brush of dark setae on femur and especially tibia; with row of distal ventral macrosetae on metatarsus I-IV, a few scattered ventral macrosetae on tarsus I-IV and metatarsus II-IV. Dorsum of abdomen with median longitudinal stripe and posterior patch of white setae (Figs 11E, F, 84A-D).
Male palp with proximal-distal axis; tegulum subtrapezoidal; conductor and embolus together form apical complex making one helical turn; conductor with more or less membranous and papilliated inner layer extending beyond moderately sclerotized outer layer; embolic division longer than tegular division; cymbium with several prolateral macrosetae (Figs 15G-I, 84I, J, 85A-D).
Spinneret spigot morphology. Female ALS with at least 7 MAP within and on inner edge of spinning field of more the 90 PI and many small tartipores (Fig. 94B); male with at least 1 MAP on inner margin and 3 more within spinning field of about 50 PI (Fig. 95B). Female PMS with 1 central mAP spigot, a large anterior spigot (probably CY) flanked by 2 large tartipores, and 30 smaller spigots scattered from anterior to posterior (Fig. 94C); male PMS with 1 central mAP and about 9 AC, suggesting that the additional spigots on the female may comprise AC and CY spigots (Fig.  95C). Female PLS with anterobasal MS and 1 accompanying spigot and distal field of more than 35 AC (Fig. 94D); male MS with 2 flanking spigots, also with about 35 AC (Fig. 95D, E). Male cribellar plate with no sign of spigots (Fig. 95F); numerous epiandrous gland spigots present (Fig. 93F).