Linyphioid Systematics and Phylogeny

The linyphioid clade comprise the families Pimoidae and Linyphiidae, both members of the Araneoidea, the ecribellate orb weavers (Griswold et al., 1998; Hormiga and Griswold, 2014). The typical orb web architecture consists of a frame holding radii that support a spiral adhesive thread. Although some of the most emblematic araneoids build orb webs, such as the majority of Araneidae and Tetragnathidae, other araneoid groups build distinctly different webs, such as the cobwebs of theridiids or the sheet webs of linyphioids, or have abandoned foraging webs altogether, such as the pirate spiders (Mimetidae). With close to five thousand described species, linyphioids are the most species-rich lineage of Araneoidea. The vast majority of those species are members of the Linyphiidae (over 4,700 species in 625 genera), the second largest spider family (World Spider Catalog, 2022). Pimoids comprise only two genera and 85 species (all but one in the genus Pimoa). The monophyly of linyphioids is supported by five synapomorphies: cheliceral stridulatory striae; patella–tibia autospasy; an enlargement of the peripheral cylindrical spigot base on the posterior lateral spinnerets; a 9+0 axonemal pattern in the sperm; and an ectal cymbial process in the male palp (Wunderlich, 1986; Hormiga, 1993, 1994a,b; Michalik and Hormiga, 2010; Hormiga et al., 2021). The monophyly of Pimoidae is supported by four synapomorphies: a modified macroseta on a dorsoectal cymbial process; a retrolateral cymbial sclerite (pimoid cymbial sclerite); an alveolar sclerite; and the absence of aciniform silk gland spigots on the female posterior median and lateral spinnerets (Hormiga, 1994b; Hormiga et al., 2005, 2021). Linyphiid synapomorphies include several genital traits: an intersegmental paracymbium, suprategulum, a distal suprategular apophysis, and a column and radix (Hormiga et al., 2021). The phylogenetic structure of the Linyphiidae tree is for the most part poorly understood.

In contrast with most araneoid families (e.g., Theridiidae or Araneidae), the somatic morphology of linyphioids is rather conservative (Fig. 1), except for the prosomal modifications (lobes, sulci and pits, setae, etc.) of many erigonine males (Hormiga, 2000), which are probably evolving under sexual selection (Michalik and Uhl, 2011; Lin et al, 2021). The paucity of variation in the somatic characters of linyphiids contrasts with the remarkable diversity of their genitalic structures, which are probably also under sexual selection (Eberhard, 1985). We argue below that several aspects of this linyphioid conservatism in morphology, in particular their relatively long thin legs, their moderately robust chelicerae with teeth on both margins of the basal segment (Draney and Buckle, 2017; Wolff et al., 2022), their moderately long cheliceral fangs (Pimoa is an exception) (Wolff et al., 2022), and their reduced number of aciniform silk gland spigots may be associated with their web architecture, which facilitate rapid attacks but are unable to retain prey for long periods because of their relatively limited adhesion to prey. Attacks on prey were very rapid in Linyphia triangularis and Microlinyphia pusilla (Sundevall, 1830) (Bristowe, 1941; Benjamin et al., 2002). In recordings (at 30 frames/s) of 12 attacks by Neriene coosa (Gertsch, 1951) on live Drosophila prey that were dropped onto the sheet a median of 13.5 body lengths from the spider, the spider delayed only a median of 0.03 s before first responding and ran toward the prey at a median speed of 57 body lengths/s; spiders reached the prey in a median of 0.37 s (minimum 0.17 s) (Eberhard, 2021, unpublished data).

Figure 1.

Habitus images of linyphioid spiders. (A) Pimoidae and (B–I) Linyphiidae habitus photographs. (A) Pimoa cthulhu Hormiga, 1994), female from California (DSC_5065). (B) Putaoa seediq Hormiga & Dimitrov, 2017 (Stemonyphantinae), male from Taiwan. (C) Dubiaranea sp. (MPME clade), female from Ecuador. (D) Laminacauda rubens Millidge, 1991 (Erigoninae), female from Juan Fernandez Islands (DSC_1027). (E) Orsonwelles malus Hormiga, 2002, male from Kauai. (F) Novafroneta sp. (Mynogleninae), female from New Zealand (DSC_2505). (G) Labulla thoracica (Wider, 1834), female from Zealand, Denmark. (H) Laetesia raveni Hormiga & Scharff, 2014, female from New South Wales, Australia (DSC_2950). (I) Pityohyphantes costatus (Hentz, 1850), female from Maryland.

The first quantitative cladistic analyses of linyphiids were based on small taxonomic samples but offered explicit phylogenies based on some of the main morphological features of the subfamilies, including spinneret spigot morphology (Hormiga, 1994a,b). Subsequent cladistic work focused primarily on Erigoninae, the largest linyphiid subfamily (e.g., Hormiga, 2000; Miller and Hormiga, 2004; Frick et al., 2010; Tu and Hormiga, 2011); Mynogleninae (Frick and Scharff, 2014); and newly described pimoid genera (later transferred to stemonyphantine Linyphiidae) (Hormiga, 2003; Hormiga et al., 2005; Hormiga and Tu, 2008). Arnedo et al. (2009) carried out the first linyphiid analysis that used multigene nucleotide sequence data (from the mitochondrial markers cytochrome c oxidase subunit I and 16S and the nuclear markers 28S, 18S, and histone H3), combined with morphological data. Their character matrix included 35 linyphiids representing the subfamilies Stemonyphantinae, Mynogleninae, Erigoninae, and Linyphiinae and 12 species representing nine other araneoid families. Their analyses supported the monophyly of linyphioids, Pimoidae, Linyphiidae, Erigoninae, and Mynogleninae and indicated that the stemonyphantines are the lineage sister to all other Linyphiidae. The Micronetinae (a large group with many northern hemisphere species that includes the genera Microneta, Agyneta, and Lepthyphantes) was shown to be paraphyletic with respect to Erigoninae.

Subsequent molecular work that used the same genes but with a much larger taxon sample (Dimitrov et al. 2012, 2017; Wang et al., 2015; Tian et al., 2022) has corroborated many of those groups but also suggested a much more complex picture of the phylogeny of Linyphiidae. All more recent analyses have confirmed the hypothesis of Arnedo et al. (2009) that micronetines are paraphyletic. The Mynogleninae, a southern hemisphere group with a disjunct distribution pattern (tropical Africa, New Zealand, and Australia), have been consistently found to be monophyletic in both morphological (Hormiga, 1994b; Frick and Scharff, 2014) and molecular analyses (Arnedo et al., 2009; Dimitrov et al., 2012; Wang et al., 2015). More recent morphological cladistic analyses have helped delimit poorly studied groups, such as the mounded posterior median eyes (MPME) clade and the Pocobletus clade (Silva-Moreira and Hormiga, 2021), but the broad picture of the linyphiid tree remains poorly defined. Hormiga et al. (2021) recently addressed the family limits of both Pimoidae and Linyphiidae with Sanger sequences and recircumscribed the former to only include Pimoa and Nanoa. The genera Weintrauboa and Putaoa (formerly in Pimoidae) and Pecado were placed in an expanded linyphiid subfamily, Stemonyphantinae, the sister lineage of a clade that includes all other linyphiids. The most comprehensive hypothesis regarding the higher level phylogeny of linyphioids (Silva-Moreira et al., in preparation), on the basis of combined morphological and target sequencing data, is summarized in Figure 2.

Figure 2.

Phylogenetic tree of linyphioids. Phylogenetic relationships of linyphioids. This cladogram is a simplified version of the optimal maximum likelihood tree resulting from a combined analysis of morphological and molecular data (cytochrome c oxidase subunit I, 16S, 18S, and histone H3) of Silva-Moreira, Kulkarni and Hormiga (in prep). The genera represented in the various clades and grades are listed on the right.

Figure 2.

Continued.

To summarize our current knowledge of linyphioid phylogeny, the monophyly of linyphioids Pimoidae and Linyphiidae is well supported. The subfamilies Stemonyphantinae, Mynogleninae, and Erigoninae also appear to be monophyletic, but all other subfamilial groups (Linyphiinae, Micronetinae, Dubiaraneinae, and Ipainae; see Tanasevitch, 2022) are not.

What are the closest relatives of linyphioids? The answer to this question can help elucidate the origin of linyphioid sheet webs. The difficulties of resolving this problem derive from the challenges of addressing the more inclusive and recalcitrant question of the phylogenetic relationships of the araneoid families. An early cladogram based on morphological and behavioral characters (Griswold et al., 1998) suggested that linyphioids were the sister group of a lineage that included theridiids, nesticids, synotaxids, and cyatholipids, but this hypothesis has not been supported by any subsequent molecular phylogenetic analysis. Target gene analyses with Sanger sequencing data have resulted in multiple incongruent araneoid topologies (e.g., Blackledge et al., 2009; Dimitrov et al., 2012, 2017; Wheeler et al., 2017) and different linyphioid sister groups. High throughput sequencing methods subsequently offered still different answers on the basis of hundreds of markers, first with the use of transcriptomic data (e.g., the analysis of Kallal et al. (2021), which included representatives of all araneoid families) followed by analyses of ultraconserved elements (UCEs; Kulkarni et al., 2021). The results of these transcriptomic analyses were not congruent, however, with those of the UCE data; and paradoxically, the incongruent placements of groups were strongly supported in both hypotheses. Both transcriptomes and UCEs place theridiids as the sister group of a lineage that includes all other araneoid families, but they disagree on the sister group of linyphioids. Transcriptomes suggest that linyphioids are sister to a lineage that includes cyatholipids and synaphrids (Kallal et al., 2021: fig. 3). UCEs place linyphiids and pimoids in a clade along with cyatholipids, physoglenids, and nesticids (Kulkarni et al., 2021: fig. 1), but with cyatholipids nested between pimoids and linyphiids. Both hypotheses place linyphiids and pimoids in a lineage with sheet and irregular web builders. The aforementioned synapomorphies of linyphioids suggest that the placement of cyatholipids between pimoids and linyphiids might be incorrect. Figure 2 provides a summary of the most recent phylogenetic hypothesis of linyphioid relationships from morphological and DNA sequence data. Considering the phylogeny in Figure 2, the similarities between the webs of cyatholipids and some physoglenids (Dimitrov et al., 2017; see also Figs. 3–10) compared with linyphioid webs, it appears that the common ancestor of these three groups probably built sheet webs of some sort.

Figure 3.

Cyatholipidae webs (1). (A) Forstera sp., female. Australia, Queensland, Atherton Tablelands (DSC_8144.NEF). (B) Forstera sp., female. Australia, Queensland, Atherton Tablelands (DSC_8129.NEF). (C) Matilda sp., female. Western Australia, Walpole-Nornalup National Park (DSC_0218.NEF). (D–F) Forstera sp., female. Australia, Queensland, Atherton Tablelands, Danbulla State Forest (GH020421_R08_32_AUS.TIF, GH020421_R08_35_AUS.TIF, Forstera_GH020421_R08_37_AUS.TIF).

Figure 4.

Cyatholipidae webs (2). (A, B) Forstera sp., female. Australia, New South Wales, Border Ranges National Park; note spider hanging under the sheet (DSC_3143. NEF, DSC_3142. NEF). (C) Teemenaarus sp., female, male. Australia, Queensland, Danbulla State Forest (GH020421_R08_30_AUS.TIF). (D–F) Teemenaarus sp., female, male. Australia, Queensland, Tamborine National Park (GH020417_R04_23_AUS.TIF, GH020417_R04_19_AUS.TIF, GH020417_R04_24_AUS.TIF).

Figure 5.

Cyatholipidae webs (3). (A, B) Forstera sp., female. Australia, New South Wales, Border Ranges National Park; note in panel B the egg sac covered in debris (DSC_3132.NEF, DSC_3135.NEF). (C–E) Teemenaarus sp. Australia, Queensland, Kuranda Environmental Park (GH920827_R05_28_AUS.TIF, GH920827_R05_31_AUS.TIF, GH920827_R05_33_AUS.TIF).

Figure 6.

Cyatholipidae webs (4). (A) Forstera sp., female. Australia, Queensland, Atherton Tablelands (GH020420_R07_17_AUS.TIF). (B–D) Forstera sp., female with spiderlings. Australia, Queensland, Danbulla State Forest (GH020421_R08_18_AUS.TIF, GH020421_R08_17_AUS.TIF, GH020421_R08_20_AUS.TIF). (E–G) Ulwembua sp., female. South Africa, Sodwana Bay National Park (GH010406_R07_07_SAF.TIF, GH010406_R07_08_SAF.TIF, GH010406_R07_12_SAF.TIF).

Figure 7.

Cyatholipidae webs (5), Wanzia fako Griswold, 1998, Cameroon, Mount Cameroon. (A) Male (GH920126_R05_28_CAM.TIF). (B, C) Male (GH920127_R06_07_CAM.TIF, GH920127_R06_27_CAM.TIF). (D) Female, male (GH920122_R04_07_CAM.TIF). (E–G) Male (GH920124_R04_14_CAM.TIF, GH920124_R04_17_CAM.TIF, GH920124_R04_19_CAM.TIF).

Figure 8.

Physoglenidae webs (1). (A, C) Chileotaxus sans Platnick, 1990, male. Chile, Puyehue National Park (GH000730_R03_37_CHI.tif, GH001230_R03_04_CHI.tif). (B) Chileotaxus sans. Sex? Chile (DSC_2028.NEF). (D) Mangua medialis Forster, 1990, male. New Zealand (DSC_7925.NEF). (E) Pahora sp. New Zealand; photo courtesy of Charles E. Griswold (CASENT9062577_CRW_0363_CEG).

Figure 9.

Physoglenidae webs (2). (A, B) Physoglenes puyehue Platnick, 1990, female. Chile, Puyehue National Park (GH001230_R03_14_CHI.tif, GH001230_R03_18_CHI.tif). (C) Tupua sp., female. Australia, Tasmania, Cradle Mountain-Lake St. Clair National Park (DSC_0264.NEF). (D) Paratupua sp., female. Australia, Victoria, nr. Warburton, O'Shannassy Aqueduct Trail, G. Hormiga (DSC_1392.NEF). (E) Paratupua sp., male. Australia, Victoria, Acheron Way, NE of Warburton (DSC_1532.NEF).

Figure 10.

Physoglenidae webs (3). (A–E) Runga sp. New Zealand. (A) Runga sp., juvenile. New Zealand, South Island; Fox Glacier, Westland Tai Poutini National Park (DSC_7972.NEF). (B, C) Runga sp., male. New Zealand, South Island, Slab Hut Creek campground, nr. Reefton (note Stiphidiidae sheet web) (DSC_0064.NEF, DSC_0069.NEF). (D) Runga sp., female. New Zealand, New Zealand, South Island, Victoria Forest Park (DSC_0071.NEF). (E) Runga sp., female. New Zealand, South Island, Arthur's Pass National Park (DSC_7998.NEF).

The Spinneret and Spigot Morphology of Linyphioids

One of the defining synapomorphies of Araneoidea is the presence of a triplet (or triad) of silk gland spigots in the posterior lateral spinnerets. The araneoid triad comprises two spigots that are connected to the aggregate silk glands and one that is fed by the flagelliform gland. The former glands produce the gluey, viscose material of the sticky silk threads, and the latter produces the axial fibers (Kovoor, 1977; Coddington, 1989). In general, this triad is present in all araneoids regardless of their web architecture (Griswold et al., 1998; Hormiga and Griswold, 2014) but is reduced or absent (presumably lost) in lineages that have abandoned foraging webs (e.g., malkarids, mimetids, the anapid Holarchaea). In adult araneoid males the triad is usually reduced to nonfunctional nubbins; erigonine linyphiids and cyatholipids offer an exception to this pattern: the adult males retain the araneoid triad (Hormiga, 2000; Griswold, 2001; Hormiga and Griswold, 2014).

Despite the enormous species diversity of the linyphioid lineage, their spinning fields share common morphological patterns. Early scanning electron microscopy studies of the linyphiid spinnerets go back to the works of Coddington (1989) for Frontinella pyramitela (Walckenaer, 1841) and Peters and Kovoor (1991) for Linyphia triangularis. Hormiga (1994a,b) followed with a survey of the spigot morphology in Pimoa (nine species) and in Linyphiidae (nine species in nine genera). Schütt (1995) studied spigot morphology of Drapetisca socialis. The studies of Hormiga and Miller (Hormiga, 2000; Miller and Hormiga, 2004; Miller, 2007) expanded the sample of erigonines. Hormiga (2002) surveyed the spigot morphology of 10 of the 13 known species of the Hawaiian genus Orsonwelles. Linyphioids exhibit the typical araneoid spigot morphology (Hormiga, 1994b; Griswold et al., 1998), except for the relatively small number of aciniform spigots in the posterior median spinnerets and posterior lateral spinnerets, a condition taken to an extreme in some species of Pimoa; Nanoa enana Hormiga, Buckle, & Scharff, 2005; and Stemonyphantes, in which adults lack aciniform spigots entirely (Hormiga, 1994a,b; Hormiga et al., 2005). This reduction is probably associated with a failure to attack prey by wrapping them in silk and reduced prey wrapping in general (Eberhard, 1967). Another feature of the linyphioid spinning fields is the enlargement of the base of the peripheral cylindrical gland spigot on the posterior lateral spinnerets relative to the sizes other cylindrical spigots (a synapomorphy of this lineage).

Linyphioids have the araneoid triad responsible for spinning sticky silk, with a few exceptions (e.g., Drapetisca socialis has only the flagelliform spigot; Schütt, 1995). As in most araneoids, the triad is “lost” after the last molt in males. The triad is usually present in the adult males of Erigoninae; in the mynoglenines Haplinis diloris (Urquhart, 1886) and Novafroneta vulgaris Blest, 1979; and in Stemonyphantes blauveltae Gertsch, 1951 (Hormiga, 2000). The near ubiquity of triads in linyphiids suggests that, contrary to some previous authors, linyphiid webs generally have sticky lines.

Study Goals

The present paper is largely descriptive rather than synthetic. It is based on a collection of field photographs made by G.H. over three decades in scattered parts of the world while carrying out other studies. The major goal is to expand taxonomic coverage: we triple the number of linyphioid species and more than triple the number of genera whose webs are documented. In most cases our descriptions of web traits are based on one or only a few webs rather than large samples of conspecific webs, so our samples are not adequate to document intraspecific variation; undoubtedly and unwillingly most if not all descriptions are overly typological. The considerable variation in web design in four species that have been studied in more detail (Hirschheimer and Suter, 1985; Rybak, 2007; Eberhard, 2022) suggests that substantial intraspecific variation is common in linyphiids.

The organization of this paper is adjusted to our small sample sizes. We do not try to characterize the web designs for any given species. Instead, the Results section first presents photographs of the webs of 112 species of linyphioids (in 39 genera) and then describes differences in the details of web designs, with photographs to illustrate. Additionally, we document webs of cyatholipids (eight species in five genera) and physoglenids (seven species in seven genera). In the Discussion we examine several evolutionary patterns by contrasting the ranges of variation in web designs in different genera, present a preliminary discussion of the possible functions of different web designs to speculate on why these diverse web forms have evolved, and note a suite of consistent linyphioid traits that we propose are associated with especially rapid attacks on prey. One ultimate intention is to discover web traits that may eventually provide useful information for grouping species and for tracing the evolution of linyphioid webs.

Websites

We recognized several types of web sites. Webs on the substrate, as in Mermessus tridentatus (Emerton, 1882) (Fig. 47F), were attached at multiple points to the upper tips of moss plants near the surface of the ground and lacked clear frame lines. These webs appeared not to have a clearly defined sheet; the web's upper surface was bumpy and formed by the lines that connected projecting tips of the substrate. Erigone dentigera O. Pickard-Cambridge, 1874 (Emerton, 1902), and Erigone atra (Nielsen, 1932) built similar webs lacking clear sheets on the substrate; Alderweireldt (1994) found that the webs of Erigone atra and E. dentipalpis (Wider, 1834) in cultivated fields were nearly always anchored to bare ground.

Figure 47.

Mecynidis and Mermessus webs. (A–C) Mecynidis sp., female. Cameroon, Mount Cameroon, north of Mapanja (GH920127_R05_25_CAM_Mecyn_sp.TIF, GH920127_R05_20_CAM_Mecyn_sp.TIF, GH920127_R05_26_CAM_Mecyn_sp.TIF). (D, E) Mecynidis sp., female. Cameroon, Mount Cameroon, north of Mapanja (GH920127_R06_15_CAM_Mecyn_sp.TIF, GH920127_R06_16_CAM_Mecyn_sp.TIF). (F) Mermessus tridentatus (Emerton, 1882), female. USA, Maryland, Patuxent Wildlife Research Center (GH940519_R02_07_USA_Eperig_trident.TIF).

Webs near the substrate were just above the upper tips of the moss or objects in the leaf litter and were much more regular, usually forming flat sheets, as in Tapinopa longidens (Wider, 1834) (Fig. 76A) and Neriene variabilis (Banks, 1892) (Figs. 53A, C–F). We classified a sheet as near the substrate when the estimated distance between a sheet and the substrate below was less than about a sheet diameter.

Figure 53.

Neriene webs (4). (A, B) Neriene sp. GH02 (GH1397), female. Taiwan, Jianshi Township, Hsinchu County (DSC_1108.NEF, DSC_1104.NEF). (C, D) Neriene variabilis (Banks, 1892), female. USA, Maryland, Patuxent Wildlife Research Center (GH930701_R00_12_USA_Neriene_variab.TIF, GH930701_R00_13_USA_Neriene_variab.TIF). (E, F) Neriene variabilis, female. USA, Maryland, Patuxent Wildlife Research Center (GH930629_R00_29_USA_Neriene_variab.TIF, GH930629_R00_33_USA_Neriene_variab.TIF). (G) Neriene litigiosa (Keyserling, 1886), female. USA, California. Grizzly Creek Redwoods State Park (GH900718_R00_18_USA_Neriene_litig.TIF).

Figure 76.

Tapinopa webs. (A) Tapinopa longidens (Wider, 1834), female (spider is at the top center of image). Sweden, Tullbotorp (GH940815_R00_14_SWE_Tapin_longi.tif). (B, C) Tapinopa vara Locket, 1982, female. Thailand, Naratiwat, Hala Bala (GH031013_R06_05_THA_ED.TIF, GH031013_R06_03_THA_ED.TIF). (D) Tapinopa bilineata Banks, 1893, juvenile. USA, Maryland, Patuxent Wildlife Research Center (GH940621_R00_27_USA_Tapinopa_ED.TIF). (E, F) Tapinopa bilineata Banks, 1893, juvenile. USA, Maryland, Patuxent Wildlife Research Center (GH940621_R00_23_USA_Tapinopa_ED, GH940621_R00_24_USA_Tapinopa_ED).

Webs that were higher above the substrate were classified as elevated. We distinguished several types of elevated webs. In an elevated leaf web, one edge of the sheet abutted against the edge of a leaf. In some cases, as in Agyneta spp. (Figs. 11C, D, 12A–C), the density of lines in the sheet was higher at the edge of the leaf, indicating that the spider apparently rested under the leaf; in others, such as Laminacauda sp. “chdes” (Fig. 38F), the open mesh of the sheet near the leaf suggested that the spider instead rested under the sheet. In some, such as Laminacauda malkini Millidge, 1991 (Fig. 37A), the spider's likely resting place was uncertain.

Figure 11.

Agyneta webs (1). (A, B) Agyneta “CR01,” female. Costa Rica, Estación Biológica La Selva (GH930404_R00_02_CRI_Meioneta_.tif, GH930404_R00_07_CRI_Meioneta_.tif). (C, D) Agyneta “CR01,” female. Costa Rica, Estación Biológica La Selva (GH930402_R00_05_CRI_.tif, GH930402_R00_10_CRI_.tif). (E, F) Agyneta semipallida (Chamberlin & Ivie, 1944), female. USA, Maryland, Patuxent Wildlife Research Center (GH940519_R00_14_USA_Meioneta_sp7.tif, GH940519_R00_15_USA_Meioneta_.tif). (G, H) Agyneta sp., subadult male. Colombia, Valle del Cauca, Farallones de Cali (GH980213_R03_12_COL_Meioneta.tif, GH980213_R03_14_COL_Meioneta.tif).

Figure 12.

Agyneta webs (2). (A–C) Agyneta sp. HI02, female. USA, Hawaii, Kauai (GH950816_R06_13_HAW_Meioneta.tif, GH950816_R06_15_HAW_Meioneta.tif, GH950816_R06_18_HAW_Meioneta.tif). (D–F) Agyneta nr. luctuosa, male. Guyana, Gunn's Landing Strip (GH990710_R04_01_GUY_Meioneta.tif, GH990710_R04_05_GUY_Meioneta.tif, GH990710_R04_07_GUY_Meioneta.tif). (G, H) Agyneta micaria (Emerton, 1882), female. USA, Maryland, Patuxent Wildlife Research Center (GH940602_R00_28_USA_Meioneta_sp2.tif, GH940602_R00_29_USA_Meioneta_sp2.tif).

Figure 37.

Laminacauda webs (2). (A) Laminacauda malkini Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0709.NEF). (B, C) Laminacauda malkini, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0584.NEF, DSC_0587). (D) Laminacauda malkini, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0712.NEF). (E, F) Laminacauda malkini, subadult female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0575.NEF, DSC_0578).

Figure 38.

Laminacauda webs (3). (A) Laminacauda rubens Millidge, 1991, male. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0046.NEF). (B, C) Laminacauda rubens, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0020.NEF, DSC_0026.NEF). (D) Laminacauda rubens, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0001.NEF). (E) Laminacauda rubens, male. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0318.NEF). (F) Laminacauda sp. (“chdes”), male. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0299.NEF).

In sheltered webs, one side was less than approximately one sheet diameter from a large nearby, more or less vertical object (e.g., a tree trunk, a dirt bank, etc.), as in Labulla thoracica (Fig. 31F). At the opposite extreme, no sheltering object was on any side of the web, as in Novafrontina sp. Brazil (Fig. 57C).

Figure 31.

Labulla webs. (A–H) Labulla thoracica (Wider, 1834), Hestehaven Forest, Denmark. (A) Female, GH940830_R04_07_DEN_Labull_thorac.tif. (B) Male, female, GH940829_R04_33_DEN_Labull_thorac.tif. (C) Male, female, GH940830_R04_05_DEN_Labull_thorac.tif. (D) Male, female, GH940831_R04_34_DEN_Labull_thorac.tif. (E) Female, GH940901_R04_24_DEN_Labull_thorac.tif. (F–H) Male, female, GH940901_R04_06_DEN_Labull_thorac.tif; female, GH940901_R04_07_DEN_Labull_thorac.tif; male, female, GH940901_R04_08_DEN_Labull_thorac.tif.

Figure 57.

Novafrontina webs (1). (A, B) Novafrontina sp., female. Brazil, Rio Jufari (DSC_9231.NEF, DSC_9233.NEF). (C) Novafrontina sp., female. Brazil, Rio Jufari (DSC_9313.NEF). (D, E) Novafrontina sp., female. Brazil, Rio Jufari (DSC_9407.NEF, DSC_9413.NEF). (F) Novafrontina sp., female. Brazil, Rio Jufari (DSC_9318.NEF).

Often the sheltered side of the sheet had a denser mesh and was attached directly to the large object, as in Laminacauda parvipalpis Millidge, 1985 (Fig. 44E) and Laetesia sp. 1 (Figs. 33D, E). In some cases the density of the sheet did not vary dramatically, and its curvature or the positions of the tangles suggested that the spider rested on the sheet rather than the substrate; the sheet was simply attached on one side to a large object such as a tree trunk or a rock, as in Neriene clathrata (e.g., Figs. 51A, E), Laminacauda magna Millidge, 1991 (Fig. 40A) and L. rubens Millidge, 1991 (Figs. 38B–D). In contrast, the leaf shelter webs were attached to the underside of a large object such as a leaf or a branch, as in Laetesia raveni Hormiga & Scharff, 2014 (Figs. 32A–C, 33A, B). This species rested upside down under a leaf near the center of the sheet at the apex of the dome. When disturbed, the spider flattened its body against the leaf surface. Some species had webs with traits of both leaf webs and sheltered webs, as in Laminacauda parvipalpis (Fig. 44E).

Figure 33.

Laetesia webs (2). (A, B) Laetesia raveni Hormiga & Scharff, 2014, female. Australia, Victoria (DSC_1325.NEF, DSC_1312.NEF). (C, D) Laetesia GH01, female. Australia, Queensland, Cape Tribulation National Park (GH920726_R05_02_AUS_Laetesia.tif, GH920726_R05_04_AUS_Laetesia_sp.tif). (E) Laetesia GH01, female. Australia, Queensland, Cape Tribulation National Park (GH020425_R12_02_AUS.TIF). (F) Laetesia GH01, female. Australia, Queensland, Cape Tribulation National Park (GH920726_R05_34_AUS_Laetesia_sp.tif).

Figure 40.

Laminacauda webs (5). (A–C) Laminacauda magna Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island. (B) Undusted, detail showing glistening silk (DSC_0674.NEF, DSC_0672.NEF, DSC_0678.NEF). (D, E) Laminacauda magna, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island. (F) Undusted, detail showing glistening silk (DSC_0682.NEF, DSC_085.NEF, DSC_0680.NEF).

Figure 44.

Laminacauda webs (9). (A) Laminacauda parvipalpis Millidge, 1985, subadult male, female. Chile, Zapallar (DSC_2267.NEF). (B) Laminacauda parvipalpis, juvenile (not certain it built this web). Chile, Zapallar (DSC_2270.NEF). (C) Laminacauda parvipalpis, female. Chile, Zapallar (DSC_2307.NEF). (D) Laminacauda parvipalpis, female. Chile, Zapallar (DSC_2322.NEF). (E) Laminacauda sp., juvenile. Chile, Zapallar (DSC_2264.NEF).

Figure 51.

Neriene webs (2). (A–C) Neriene clathrata (Sundevall, 1830), female. USA, Maryland, Patuxent Wildlife Research Center. (C) Detail of understructure after removal of the main sheet; denser areas are probably remains of an older sheet, suggesting that the understructure already existed (at least, partially) when the new main sheet was built (GH930701_R00_24_USA_Neriene_clathr.TIF, GH930701_R00_26_USA_Neriene_clathr.TIF, GH930701_R00_29_USA_Neriene_clathr.TIF). (D) Neriene clathrata, female. USA, Maryland, Patuxent Wildlife Research Center (GH930701_R00_34_USA_Neriene_clathr.TIF). (E, F) Neriene clathrata, female. USA, Maryland, Patuxent Wildlife Research Center (GH930701_R00_32_USA_Neriene_clathr.TIF, GH930701_R00_33_USA_Neriene_clathr.TIF). (G–I) Neriene clathrata, female. USA, Maryland, Patuxent Wildlife Research Center (GH940510_R00_29_USA_Neriene_clathr.TIF, GH940510_R00_30_USA_Neriene_clathr.TIF, GH940510_R00_34_USA_Neriene_clathr.TIF).

The species of the plant may have been an important web site variable for the Australian plant-specific species Laetesia raveni. All but two webs in a series of 48 were built on two thorny plant species: most were on wait-a-while vines (Calamus muelleri Wendland, 1875, Arecaceae) (Figs. 32B–E, 34A–C); others were on Gin's Whiskers (Solanum inaequilaterum Domin, 1929, Solanaceae) (Fig. 32A). The two exceptions, in Binna Burra (Lamington National Park), were webs built on other plant species that were in physical contact with one of the two thorny species (see Hormiga & Scharff, 2014). In a locality in Victoria, however, we found webs on non-thorny plants (Figs. 33A, B).

Figure 32.

Laetesia webs (1). (A–H) Laetesia raveni Hormiga & Scharff, 2014, female. Australia, New South Wales, Dorrigo National Park. (A) DSC_2834.NEF. (B, C) DSC_2808.NEF, DSC_2804.NEF. (D, E) DSC_2824.NEF, DSC_2828.NEF.

Figure 34.

Laetesia webs (3). (A, C) Laetesia raveni Hormiga & Scharff, 2014, male. Australia, Queensland, Bina Burra (GH020418_R06_05_AUS.tif, GH020418_R06_07_AUS.tif). (B) Laetesia raveni Hormiga & Scharff, 2014, female. Australia, Queensland, Bina Burra (GH020418_R05_33_AUS.tif). (D, E) Laetesia GH02, male. Australia, Queensland, Atherton Tablelands (GH020420_R07_03_AUS.tif, GH020420_R06_35_AUS.tif).

Most elevated webs were not associated with large objects, as, for instance, Frontinella spp. (Figs. 27A–D). In highly elevated webs the sheet was more than two sheet diameters above the ground or leaf litter, and was suspended from the leaves and branches of plants. An extreme case was Pocobletus versicolor (Millidge, 1991) (Fig. 73E), in which the sheet was far from any substrate and was anchored by long lines at only four corners; the spider rested in the central portion of this sheet protected by the lower tangle (Fig. 73F).

Figure 27.

Frontinella webs. (A, B) Frontinella sp., female. Dominican Republic, Cachote, Sierra de Barohuco (GH050409_R02_20_DR_Frontinella_.jpg, GH050409_R02_28_DR_Frontinella_.jpg). (C, D) Frontinella pyramitela (Walckenaer, 1841), female. USA, Maryland (GH940511_R02_30_USA_Frontin_pyram.tif, GH940511_R02_32_USA_Frontin_pyram.tif).

Figure 73.

Pocobletus webs (8). (A, B) Pocobletus versicolor (Millidge, 1991), male. Ecuador, Parque Nacional Yasuní (GH960621_R07_20_ECU_Exechop_versic_ED.TIF, GH960621_R07_21_ECU_Exechop_versic_ED.TIF). (C, D) Pocobletus versicolor (Millidge, 1991), female. Costa Rica, Estación Biológica La Selva (GH930405_R08_11_CRI_Exechop_versic_ED.TIF, GH930405_R08_14_CRI_Exechop_versic_ED.TIF). (E, F) Pocobletus versicolor (Millidge, 1991), female (E, sheet web seen from above; F, lateral view of the same web). Costa Rica, Estación Biológica La Selva (GH930405_R08_21_CRI_Exechop_versic_ED.TIF, GH930405_R08_25_CRI_Exechop_versic_ED).

Only very seldom did a web have any object protruding through either the sheet or the tangles of a web. In other words, web sites almost always had open spaces at least the size of a web. One exception was a Juanfernandezia melanocephala Millidge, 1991, sheet with a stem protruding through it (Fig. 29C) (an obstructed sheet). The only linyphiid species we know of in which objects (thin leaves) regularly protrude through the sheet is Neriene coosa (Eberhard, 2022). Twigs and leaves that penetrate sheet webs are not uncommon in species in other families such as Lycosidae, Zoropsidae (Eberhard and Hazzi, 2017), and Hahniidae (Eberhard, 2019). In sum, linyphiids consistently chose web sites with sufficient objects nearby to which to attach their webs and that had a completely open space in which the web would be built.

Figure 29.

Juanfernandezia webs (1). (A–E) Juanfernandezia melanocephala Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island. (A, B) DSC_2502.NEF, DSC_2500.NEF. (C) DSC_2512.NEF. (D) DSC_2542.NEF. (E) DSC_2553.NEF.

Sheet Form

All of the webs described here except those on the substrate had more or less horizontal sheets that were at least moderately dense. A “sheet” was an approximately planar array of lines (sometimes strongly curved upward or downward) where coplanar lines appeared to be denser than elsewhere; only occasionally was it difficult to distinguish a tangle from a sheet (e.g., some of the webs of Laminacauda ansoni Millidge, 1991, Figs. 36A–F). The sheets were “dense” compared with the open mesh sheets in other families such as Psechridae (Robinson and Lubin, 1979; Eberhard, 1987) and Austrochilidae (Lopardo et al., 2004), often having lines so dense that they were difficult to distinguish from each other in the photos. One linyphiid, Bathyphantes eumenis (= simillimus), builds sparse webs that lack sheets (Rybak, 2007). These webs, which were discovered in captivity, were so sparse and irregular in form that it is not clear that they would have been recognized as prey capture webs in the field. The lack of such webs in our study may thus be an artifact of the difficulty of distinguishing such sparse webs from incidental, sparse accumulations of drag lines in the field.

Figure 36.

Laminacauda webs (1). (A) Laminacauda ansoni Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0035.NEF). (B) Laminacauda ansoni Millidge, 1991, male. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0020.NEF). (C) Laminacauda ansoni, male. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0052.NEF). (D, E) Laminacauda ansoni, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0443.NEF, DSC_0438.NEF). (F) Laminacauda sp. (“chps”). Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0304.NEF).

Cup-shaped. The outer edges of cup-shaped sheets sloped upward more or less symmetrically on all sides from a more or less central low area. The angles of the slopes varied. In weakly cup-shaped sheets, the upward slant was barely perceptible, as in Laminacauda tuberosa Millidge, 1991 (Fig. 39H), Acroterius sp. (Fig. 79F), and Neriene clathrata (Figs. 51A, E); weak slants verged on flat, as in Acroterius sp. (Fig. 79E). In moderately cup-shaped sheets the angle was clear but not strong, as in Novafrontina sp. (Fig. 57B). The angle of curvature was more acute in strongly cup-shaped sheets, as in Pocobletus sp. GH02 (Figs. 74C, D). In asymmetrical cup-shaped sheets, one edge of the sheet was much higher than the others and formed a nearly vertical wall, as in Neriene sp. GH03 (Figs. 52A, B) and Acroterius spp. (Figs. 78, 79B); in other tilted asymmetrical cup-shaped sheets the entire cup was tilted to the side, as in Laminacauda sp. “fPC” (Fig. 43A).

Figure 39.

Laminacauda webs (4). (A) Laminacauda propinqua Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2495.NEF). (B) Laminacauda propinqua, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2519.NEF). (C) Laminacauda propinqua (no sex info). Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2496.NEF). (D) Laminacauda propinqua, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2532.NEF). (E) Laminacauda propinqua, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2537.NEF). (F) Laminacauda propinqua, female. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2529.NEF). (G) Laminacauda propinqua, male. Chile, Juan Fernandez Archipelago, Alejandro Selkirk Island (DSC_2535.NEF). (H) Laminacauda tuberosa Millidge, 1991. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0299.NEF).

Figure 43.

Laminacauda webs (8). (A–C) Laminacauda sp. (“fPC”), female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0409.NEF, DSC_0405.NEF, DSC_0413.NEF). (D, E) Laminacauda sp. (“fChi”), female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0073.NEF, DSC_0077.NEF).

Figure 52.

Neriene webs (3). (A–C) Neriene sp. GH03 (GH1398), female. Taiwan, Jianshi Township, Hsinchu County (DSC_1076.NEF, DSC_1077.NEF, DSC_1081.NEF). The main sheet of this web is vertically oriented. (D, E) Neriene sp. GH03 (GH1398), male and female. Taiwan, Jianshi Township, Hsinchu County (DSC_1043.NEF, DSC_1036.NEF).

Figure 74.

Pocobletus webs (9). (A, B) Pocobletus sp. exoGH01, juvenile. Costa Rica, Cerro Pittier, Parque Internacional La Amistad (GH950612_R00_14_CRI_Exocora_sp1_ED.tif, GH950612_R00_15_CRI_Exocora_sp1_ED.tif). (C, D) Pocobletus sp. exoGH02, female. Costa Rica, Cerro Pittier, Parque Internacional La Amistad (GH950612_R00_17_CRI_Exocora_sp2_ED.tif, GH950612_R00_22_CRI_Exocora_sp2_ED.tif). (E, F) Pocobletus sp. exoGH02, female. Costa Rica, Cerro Pittier, Parque Internacional La Amistad (GH950613_R00_30_CRI_Exocora_sp2_ED.tif, GH950613_R00_32_CRI_Exocora_sp2_ED.tif).

Figure 78.

Acroterius webs (1). Acroterius GH02, subadult male. Taiwan, forest north of Dungyuan (DSC_0964.NEF).

Figure 79.

Acroterius webs (2). (A) Acroterius GH02, juvenile. Taiwan, forest north of Dungyuan (DSC_0975.NEF). (B) Acroterius GH02, juvenile. Taiwan, forest north of Dungyuan (DSC_0959.NEF). (C) Acroterius GH02, female. Taiwan, Forest N of Dungyuan (DSC_0938.NEF). (D, E) Acroterius GH01, male. Thailand, Chiang Mai, Doi Inthanon (GH031004_R03_21_THA_ED.TIF, GH031004_R03_22_THA_ED.TIF). (F) Acroterius GH01, subadult female. Thailand, Chiang Mai, Doi Inthanon (GH031004_R03_25_THA_ED.TIF). (G) Acroterius GH01, subadult female. Thailand, Chiang Mai, Doi Inthanon (GH031004_R03_29_THA_ED.TIF).

Double Sheets. Some webs with a cup-shaped principal sheet had a second, lower sheet just below that was also cupped-shaped. Some webs barely had enough space between the two sheets for the spider to move, as in Acroterius sp. (Figs. 78, 79C); the space was much greater in others, such as Mecynidis sp. (Figs. 47A–C) and Frontinella spp. (Figs. 27A–D). The distinction between a sparsely meshed lower sheet and a lower tangle was not always clear, as in Laminacauda “fCh” (Fig. 43E). Flat sheets also occasionally had a second sheet-like structure in the lower tangle, as in Mecynidis sp. (Figs. 47A–C). In contrast, no web with a dome-shaped sheet had a second sheet. In one web of Mecynidis sp. it appeared that a leaf replaced the lower sheet (Figs. 47D, E).

Lower sheets were always less densely meshed than the principal sheet. Some lower sheets were relatively small and covered only the lowermost portion of the principal sheet, as in Acroterius sp. (Fig. 78), and Neriene digna (Keyserling, 1886) (Figs. 50G, H)). Other lower sheets were nearly as wide as the principal sheet, as in Frontinella spp. (Figs. 27A–D).

Figure 50.

Neriene webs (1). (A, B) Neriene albolimbata (Karsch, 1879), subadult female. Taiwan, Jianshi Township, Hsinchu County (DSC_1058.NEF, DSC_1059.NEF). (C, D) Neriene albolimbata, female. Taiwan, Jianshi Township, Hsinchu County (DSC_1085.NEF, DSC_1093.NEF). (E) Neriene albolimbata, female. Taiwan, Jianshi Township, Hsinchu County (DSC_1050.NEF). (F) Neriene sp. GH01 (GH1401), male and female. Taiwan, Jianshi Township, Hsinchu County (DSC_1100.NEF). (G) Neriene digna (Keyserling, 1886), female. USA, Oregon, Siuslaw National Forest (GH900714_R00_21_USA_Neriene_digna.TIF). (H) Neriene digna, female. USA, Oregon, Siuslaw National Forest (GH900714_R00_18_USA_Neriene_digna.TIF).

Trough-shaped. A male of Agyneta sp. Guyana built a unique trough-shaped sheet just above the upper surface of a longitudinally curved leaf on the forest floor. The elongate sheet, about 8 cm long, approximated the lower half of a cylinder laid on its side just above the surface of the leaf (Fig. 12F). The sheet's lateral edges slanted upward, but not its ends. The upper tangle of this web spanned the space between the lateral edges of the leaf (Figs. 12D, E), with a much smaller lower sheet under the main sheet.

Dome-shaped. The outer edges of dome-shaped sheets sloped downward more or less symmetrically from a more or less central area. The angle of the slope varied widely. In weakly dome-shaped sheets, the slant was barely perceptible, as in Diplothyron dianae Silva-Moreira & Hormiga, 2022 (Fig. 20D), and Laminacauda propinqua Millidge, 1991 (Fig. 39G), and verged on flat, as in Neriene variabilis (Figs. 53D–F). In moderately dome-shaped sheets, the angle was clear but more obtuse, as in Laetesia raveni (Fig. 33B). The angle of curvature was more acute in strongly dome-shaped sheets, as in Neriene albolimbata (Karsch, 1879) (Figs. 50A–D), and N. litigiosa (Keyserling, 1886) (Figs. 53G, 56). Only rarely a sheet was dome-shaped in one portion and saddle-shaped in another, as in Neriene helsdingeni (Locket, 1968) (Figs. 55C, D). The surfaces of domes were relatively smooth with small upward-projecting dimples, as in Diplothyron sp. (Fig. 20D). In asymmetrical dome-shaped sheets one side of the dome extended much farther downward than the others, as in Putaoa seediq Hormiga & Dimitrov, 2017 (Figs. 75C, D), and Laetesia sp. GH02 (Figs. 34D, E); in extreme cases the extended side formed a vertical wall, as in Pimoa cthulhu Hormiga, 1994 (Figs. 64A, B).

Figure 20.

Bathyphantes and Diplothyron webs. (A, B) Bathyphantes pallidus (Banks, 1892), subadult female. USA, Maryland, Patuxent Wildlife Research Center (GH940510_R00_21_USA_Bathy_pallid.tif, GH940510_R00_25_USA_Bathy_pallid.tif). (C) Bathyphantes pallidus, male. USA, Maryland, Patuxent Wildlife Research Center (GH930630_R02_07_USA_Bathy_pallid.tif). (D, E) Diplothyron dianae Silva-Moreira & Hormiga, 2022, female. Costa Rica, Parque Internacional La Amistad, Cerro Pittier (GH950613_R09_06_CRI_diplothy.tif, GH950615_R09_08_CRI_diplothy.tif). (F, G) Diplothyron dianae, female. Costa Rica, Cerro Pittier, Parque Internacional La Amistad (GH950615_R09_28_CRI_diplothy.tif, GH950615_R00_32_CRI_Diplothyron.tif).

Figure 55.

Neriene webs (6). (A, B) Neriene helsdingeni (Locket, 1968), female. Cameroon, Mount Cameroon, Etome (GH920116_R02_20_CAM_Nereine_helsd.TIF, GH920116_R02_21_CAM_Nereine_helsd.TIF). (C–E) Neriene helsdingeni, female. Cameroon, Mount Cameroon, Mann's Spring (GH920123_R03_28_CAM_Nereine_helsd.TIF, GH920123_R03_1_CAM_Nereine_helsd.TIF, GH920123_R03_33_CAM_Nereine_helsd.TIF). (F, G) Neriene helsdingeni, female. Cameroon, Mount Cameroon, Mann's Spring (GH920125_R04_22_CAM_Nereine_helsd.TIF, GH920125_R04_23_CAM_Nereine_helsd.TIF, GH920125_R04_22_CAM_Nereine_helsd.TIF). (H) Neriene helsdingeni. Cameroon, Mount Cameroon, Mann's Spring (GH920125_R04_25_CAM_Nereine_helsd.TIF).

Figure 56.

Neriene webs (7). (A–E) Neriene litigiosa (photos by William G. Eberhard). USA, Washington, Whatcom Co. (A) D#54. (B) D#47. (C, E) D#51. (D) D#53.

Figure 64.

Pimoa webs. (A, B) Pimoa cthulhu Hormiga, 1994, female. USA, California, Mendocino Woodlands State Park (DSC_5047_ED.NEF, DSC_5057_ED.NEF). (C) Pimoa cthulhu, female. USA, California, Mendocino Woodlands State Park (DSC03917_ED.jpg). (D) Pimoa cthulhu, female. USA, California, Mendocino Woodlands State Park (DSC03992_ED.jpg). (E) Pimoa breviata Chamberlin & Ivie, 1943, female. USA, Oregon, Azalea State Park (GH900716_R00_29_USA_Pimoa_breviata_ED.TIF). (F) Pimoa breviata, female. USA, Oregon, Humbug Mountain State Park (DSC03994_ED.JPG).

Figure 75.

Putaoa webs. (A, B) Putaoa seediq Hormiga & Dimitrov, 2017, subadult female. Taiwan, Huisun Forestry Station, Xiaochu Mtn. (A) Sheet web, from above. (B) Same web as in A, from below, showing retreat entrance funnel (DSC_1012_ED.NEF, DSC_1014_ED.NEF). (C, D) Putaoa seediq, juvenile. Taiwan, Huisun Forestry Station, Xiaochu Mtn. Lampshade web, slightly from below. (D) Same as in C, from the side (DSC_1030_ED.NEF, DSC_1031_ED.NEF). (E, F) Putaoa seediq, juvenile. Taiwan, Huisun Forestry Station, Xiaochu Mtn. Sheet web, from above. (F) Same web as in E, from below, showing retreat entrance funnel (DSC_1022_ED.NEF, DSC_1025_ED.NEF).

In dome + tube webs, the upper tip of a clear dome extended to form a short tube; some tubes led to the substrate, as in Laminacauda ansoni (Fig. 36D), and other tubes of this same species ended in the upper tangle (Fig. 36A).

Flat. In flat sheets, the entire sheet was in more or less the same plane. Many flat sheets were approximately horizontal, as in Tenuiphantes flavipes (Blackwall, 1854) (Fig. 77A), Walckenaeria(?) sp. (Figs. 77D, E), and Tapinopa bilineata Banks, 1893 (Fig. 76D), but some flat sheets were substantially inclined, as in Mecynidis sp. (Fig. 47B). Sheets that sagged slightly, apparently because of their coat of powder, were classified as flat. Some flat sheets had lines attached to them that pulled the sheet slightly upward, producing a bumpy flat sheet, as in Neriene variabilis (Figs. 53D–F) and Neriene sp. GH02 (Figs. 53A, B). The edges of some flat sheets curled slightly upward, as in Laminacauda tuberosa (Fig. 39H, left).

Figure 77.

Tenuiphantes and Walckenaeria? webs. (A) Tenuiphantes flavipes (Blackwall, 1854), female. Denmark, Hestehaven (GH940901_R00_11_DEN_Tenuiph_flav_ED.TIF). (B) Tenuiphantes flavipes, female. Denmark, Hestehaven (GH940901_R00_33_DEN_Tenuiph_flav_ED.TIF). (C) Tenuiphantes flavipes, female. Denmark, Hestehaven (GH940901_R00_12_DEN_Tenuiph_flav_ED.TIF). (D, E) Walckenaeria? sp., female. Costa Rica, Parque Internacional La Amistad Cerro Pittier (GH950613_R00_36_CRI_Walcken_ED.TIF, GH950613_R00_37_CRI_Walcken_ED.TIF).

Tent-shaped. In a ridge tent sheet, an uppermost central ridge, supported by a single long line from one edge of the web to the other, slanted upward at one end toward a nearby large leaf and sloped downward on both sides of the central ridge, as in Laminacauda sp. “chdes” (Fig. 38F).

Sandwich. Sandwich sheets, as in Tapinopa bilineata (Fig. 76D) and Laminacauda magna (Fig. 41D), consisted of a pair of more or less flat parallel sheets, one above the other. They were attached near the base of a tree and their edges were loosely joined, thus enclosing the spider, which walked on the underside of the upper sheet.

Figure 41.

Laminacauda webs (6). (A, B) Laminacauda magna Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0018.NEF, DSC_0013.NEF). (C, D) Laminacauda magna (no sex info). Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0103.NEF, DSC_0105.NEF). (E) Laminacauda magna, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_0093.NEF).

Saddle-shaped. Saddle-shaped sheets resembled dome-shaped sheets but with one side sloping upward, as in Neriene helsdingeni (Fig. 55D).

Tube at the Edge. Some sheets were continuous with the wall of a tube at the edge located at one edge of the sheet, as in Putaoa seediq (Figs. 75C–F) and (possibly) in Pimoa cthulhu (Fig. 64C). In P. seediq the tube served as the spider's retreat. Funnel-like silk tubes leading into a retreat were also common in Orsonwelles species (e.g., O. polites Hormiga, 2002; Figs. 62B, C, E). Tubes were not always easy to distinguish from runways (below).

Figure 62.

Orsonwelles webs (4). (A) Orsonwelles polites Hormiga, 2002, juvenile. Hawaii, Oahu, Honouliuli Forest Reserve. Note discarded old sheet at the bottom (Orsonwelles polites_J_ow27plates.tiff). (B, C) Orsonwelles polites, female. Hawaii, Oahu, Mount Kaala Natural Area Reserve. Detail of funnel (B) (Orsonwelles polites_F_ow29plates.tiff, Orsonwelles polites_F_ow28plates.tiff). (D) Orsonwelles polites, female. Hawaii, Oahu, Wainae Kai (Orsonwelles polites_F_ow26plates.tiff). (E) Orsonwelles polites, female. Hawaii, Oahu, Palikea (Orsonwelles polites_F_ow30plates.tiff).

Complex Sheet Form. A few complex sheets had forms that did not fit easily into the preceding categories, as in Laperousea sp. GH01 (Fig. 45E) and Australian genus 3 sp. GH01 (Figs. 18A, B).

Figure 18.

Australian genus 3 webs. (A, B) Australian genus 3 GH01, female. Australia, New South Wales, Dorrigo National Park (DSC_2817.NEF, DSC_2819.NEF). (C) Australian genus 3 GH01, female. Australia, Queensland, Lamington National Park, Bina Burra (GH020418_R06_08_AUS_ED.tif). (D, E) Australian genus 3 GH02, female. Australia, Queensland, Danbulla State Forest (GH020421_R09_03_AUS.tif, GH020421_R09_05_AUS.tif). (F) Australian genus 3 GH02, female. Australia, Queensland, Danbulla State Forest (GH020421_R09_11_AUS.tif).

Figure 45.

Laperousea webs. (A, B) Laperousea blattifera (Urquhart, 1887), male. Australia, Queensland, Lamington National Park (GH020417_R03_27_AUS_Laperousea.TIF, GH020417_R03_30_AUS_Laperousea.TIF). (C, D) Laperousea sp. (GH02)(GH1677), female. Australia, Victoria (DSC_1292.NEF, DSC_1294.NEF). (E, F) Laperousea sp. (GH01), female. Australia, Tasmania, Cradle Mountain National Park (DSC_0257.NEF, DSC_0256.NEF).

Naked and Nearly Naked Sheets

Many naked sheets and nearly naked sheets had few or no lines either above or below. Naked sheets were never cup- or dome-shaped. The association between curved sheets and lines above and below it was probably because the sheet could only be pulled into a curve by additional lines out of the sheet's plane.

Naked sheets varied in several ways. Some were near the substrate, as in Himalaphantes sp. (Figs. 28D, E), Tenuiphantes flavipes (Figs. 77A–C), Agyneta sp. Maryland (Figs. 12G, H), Sphecozone ardens Millidge, 1985 (Fig. 63B), and Australolinyphia remota Wunderlich, 1976 (Fig. 19G). (About 25 other webs of this species also consisted of nearly naked sheets.) Others were elevated, as in Agyneta spp. Colombia and Hawai‘i (Figs. 13D, E), Laminacauda malkini (Fig. 37A), Dubiaranea sp. Ecuador (Figs. 24D, E), and Grammonota sp. (Fig. 28A). Some were sparsely meshed, as in Laminacauda malkini (Fig. 37A), whereas others were densely meshed, as in Agyneta sp. CR01 (Fig. 11B) and Grammonota sp. (Fig. 28A). A few nearly naked sheets near the substrate had patches of slime, as in Laminacauda magna (Figs. 42A, B). (We inferred that the slime was from the spider and not some other agent, such as a snail, from the pattern and distribution of the slime.) A more extreme case was Tapinopa longidens, whose naked sheet was covered with slime (Fig. 76A). Nielsen (1932) and Bristowe (1941, 1958) also described a slime-covered sheet in this species; web photos of this species (Shinkai, 1979; Shinkai and Takano, 1984) also show tiny naked sheets near the ground that were not obviously shiny, but it is possible that a shine would only be apparent at certain angles of illumination, so a slimy surface cannot be confidently ruled out). The stickiness of the glistening silk in one L. magna sheet was no different from that of the adjacent sheet, as assessed by light contact with the blunt metal end of a mechanical pencil. The majority of the many L. magna webs examined lacked slime, so slime is an uncommon feature in this species.

Figure 13.

Agyneta webs (3). (A, B) Agyneta sp. females. South Africa, Sodwana Bay National Park (GH010407_R08_02_SAF_Meioneta.tif, GH010407_R08_04_SAF_Meioneta.tif). (C) Agyneta sp., male (same species as above). South Africa, Sodwana Bay National Park (GH010404_R06_13_SAF_Meioneta.tif). (D) Agyneta sp., male. Colombia, Valle del Cauca, Farallones de Cali (GH980213_R03_04_COL_Meioneta.tif). (E) Agyneta sp. HI01, female. Hawaii. Oahu, Mt. Ka‘ala (GH950813_R04_01_HAW_Meioneta.tif). (F) Agyneta sp. (undet.), female. USA, North Carolina (GH910622_R00_17_USA_Meioneta.tif).

Figure 19.

Australian genus 4 and Australolinyphia webs. (A, B) Australian genus 4 GH01. Australia, Tasmania, Weldborough Pass (DSC_0339.NEF, DSC_0343.NEF). (C) Australian genus 4 GH01. Australia, Tasmania, Weldborough Pass (DSC_0348.NEF). (D, E) Australolinyphia remota Wunderlich, 1976, subadult male. Australia, Queensland, Lamington National Park (GH020416_R03_11_AUS_Austral_remot.tif, GH020416_R03_12_AUS_Austral_remot.tif). (F, G) Australolinyphia remota, female. Australia, Queensland, Lamington National Park (GH020416_R03_15_AUS_Austral_remot.tif, GH020416_R03_16_AUS_Austral_remot.tif).

Figure 24.

Dubiaranea webs (4). (A) Diplothyron nubilosus Moreira & Hormiga, 2022, female. Panama, Parque Internacional La Amistad (DSC_4011.NEF). (B) Dubiaranea hospita (Keyserling, 1886), female. Trinidad and Tobago, Trinidad, Brasso Seco (DSC_0554.NEF). (C) Dubiaranea hospita (Keyserling, 1886), female. Trinidad and Tobago, Trinidad, Brasso Seco (DSC_0544. NEF) . (D, E) Dubiaranea sp. DE1, female. Ecuador, Napo Prov., Sierra Azul (GH960615_R04_20_ECU_Dubiar_sp.tif, GH960615_R04_23_ECU_Dubiar_sp.tif).

Figure 28.

Grammonota and Himalaphantes webs. (A) Grammonota sp., juvenile. Costa Rica, La Selva (GH930403_R06_15_CRI_.tif). (B, C) Himalaphantes sp., female. Taiwan, Nantou County, Beidong Mtn. (DSC_0991.NEF, DSC_0994). (D, E) Himalaphantes sp., female. Taiwan, Nantou County, Beidong Mtn. (DSC_0999.NEF, DSC_1005.NEF).

Figure 42.

Laminacauda webs (7). (A–E) Laminacauda magna Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island. (B) Undusted, detail showing glistening silk (DSC_0692.NEF, DSC_0694.NEF, DSC_0690.NEF, DSC_0697.NEF, DSC_0696.NEF).

Figure 63.

Ostearius and Sphecozone webs. (A) Ostearius melanopygius (O. Pickard-Cambridge, 1880), female with egg sac. Hawaii, Oahu, Palikea (GH990409_R04_05_USA_Ostear_melano.tif). (B) Sphecozone bicolor (Nicolet, 1849), female. Chile, Parque Nacional Huerquehue (DSC_2621_ED.NEF). (C, D) Sphecozone bicolor, female. Chile, Parque Nacional Huerquehue (DSC_2600_ED.NEF, DSC_2595_ED.NEF). (E, F) Sphecozone bicolor, juvenile. Chile, Parque Nacional Puyehue (GH001230_R02_27_CHI_Sphecozone_ED.TIF, GH001230_R02_30_CHI_Sphecozone_ED.TIF).

Patterns of Lines in Sheets

Density. The “mesh” of a sheet refers to the density of lines in the sheet. Densities of lines in the sheet ranged from sparse, as in Laminacauda ansoni (Figs. 36D, E), Australian genus 2 sp. GH01 (Figs. 14C–F) and Dubiaranea lugubris Millidge, 1991 (Figs. 22C, D), to medium dense, as in Dubiaranea insulana Millidge, 1991 (Figs. 23C–E) and Dubiaranea fulgens (Millidge, 1985) (Fig. 22A), to dense, as in Floronia bucculenta (Fig. 26D), Agyneta sp. CR01 (Fig. 11B), and (at least in the central area) Dubiaranea sp. Ecuador (Fig. 25C) and Dubiaranea sp. Costa Rica (Fig. 22F). Very dense sheets were those in which individual lines could not be easily distinguished, as in Pocobletus versicolor(Figs. 73C, E).

Figure 14.

Australian genera 1 and 2 webs. (A, B) Australian genus 1 GH06, female. Australia, Tasmania, Mt. Field National Park (DSC_1585.NEF, DSC_1587.NEF). (C, D) Australian genus 2 GH01, female. Australia, Queensland, Lamington National Park (GH020416_R03_07_AUS_.tif, GH020416_R03_09_AUS_.tif). (E, F) Australian genus 2 GH01, female. Australia, Queensland, Lamington National Park (GH920719_R01_24_AUS_.tif, GH920719_R01_26_AUS_.tif).

Figure 22.

Dubiaranea webs (2). (A, B) Dubiaranea fulgens Millidge, 1985, female. Chile, Parque Nacional Puyehue (GH010101_R05_19_CHI_Dubiar_fulg.tif, GH010101_R05_22_CHI_Dubiar_fulg.tif). (C, D) Dubiaranea lugubris Millidge, 1991, female. Colombia, P.N. Purace, Laguna San Rafael (GH980215_R04_01_COL_Dubiar_lug.tif, GH980215_R04_11_COL_Dubiar_lug.tif). (E, F) New genus MPME sp. 1, female. Costa Rica, Cerro de la Muerte (GH930221_R02_17_CRI_Dubiar_sp.tif, GH930221_R02_10_CRI_Dubiar_sp.tif).

Figure 23.

Dubiaranea webs (3). (A, B) Dubiaranea caledonica (Millidge, 1985), female. Chile, Monumento Nacional Contulmo (DSC_1997.NEF, DSC_2000.NEF). (C–E) Dubiaranea insulana Millidge, 1991, female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island (DSC_416.NEF, DSC_418.NEF, DSC_428.NEF). (F, G) Dubiaranea sp. DE2, female. Ecuador, Napo Prov., Sierra Azul (GH960611_R01_15_ECU_Dubiar_sp_ED.tif, GH960611_R01_11_ECU_Dubiar_sp_ED.tif).

Figure 25.

Dubiaranea webs (5). (A–C) Dubiaranea sp. DE1, female. Ecuador, Napo Prov., Sierra Azul (GH960610_R01_04_ECU_Dubiar_sp.tif, GH960610_R01_08_ECU_Dubiar_sp.tif, GH960610_R01_09_ECU_Dubiar_sp.tif). (D, E) Diplothyron nubilosus Moreira & Hormiga, 2022, female. Costa Rica, Cerro de la Muerte (GH930222_R02_23_CRI_Dubiar_spec.tif, GH930222_R02_27_CRI_Dubiar_spec.tif).

Figure 26.

Diplothyron and Floronia bucculenta webs. (A) Diplothyron nubilosus Moreira & Hormiga, 2022, male. Costa Rica, Cerro de la Muerte (GH930222_R02_34_CRI_Dubiar_sp.tif). (B, C) Diplothyron nubilosus Moreira & Hormiga, 2022, male and female. Panama, Parque Internacional La Amistad (DSC_4009.NEF, DSC_4010.NEF) (note the spider on the sheet in panel C). (D, E) Floronia bucculenta (Clerk, 1757), female. Sweden, Tullbotorp (GH940813_R00_06_SWE_Floron_bucc.tif, GH940813_R00_09_SWE_Floron_bucc.tif).

The density of lines in the sheet often varied in different parts of the sheet and was lower near the edge than it was in the rest of the sheet in Dubiaranea sp. Costa Rica (Fig. 22F), Dubiaranea sp. Ecuador (Figs. 23F, G), Labulla thoracica (Fig. 31G), and Laminacauda magna (Fig. 42A). Greater density at the edge of the sheet that abutted broadly with a sheltering object, which occurred in Laminacauda magna (Fig. 41C, top right), Laminacauda sp. (Fig. 44E), Putaoa seediq (Fig. 75A), Orsonwelles macbeth Hormiga, 2002 (Fig. 60G, upper left), and O. polites (Fig. 62B), may have been associated with the spider resting at or just beyond the edge of the sheet, but further observations will be needed to test this idea. Differences in density could result from lines having accumulated near the retreat in older sheets. Thus, the sheet of Pityohyphantes costatus (Hentz, 1850) shown in Figures 65A–C, which abutted broadly with the substrate where the spider rested but had no local increase in density, may have been recently built (this web was photographed at about 9 p.m.).

Figure 60.

Orsonwelles webs (2). (A) Orsonwelles graphicus (Simon, 1900), female. Hawaii, Hawaii, Kahaualea Natural Area Reserve (Orsonwelles graphicus_F_ow40plates.tiff). (B) Orsonwelles graphicus, female. Hawaii, Hawaii, Pu‘u Maka‘ala Natural Area Reserve (Orsonwelles graphicus_F_ow41plates.tiff). (C, D) Orsonwelles graphicus, female. Hawaii, Hawaii, Pu‘u Maka‘ala Natural Area Reserve (Orsonwelles graphicus_F_ow42plates.tiff, Orsonwelles graphicus_F_ow43plates.tiff). (E) Orsonwelles macbeth Hormiga, 2002, female. Hawaii, Molokai, Kamakou Preserve (Orsonwelles macbeth_F_ow23plates.tiff). (F) Orsonwelles macbeth, female. Hawaii, Molokai, Kamakou Preserve (Orsonwelles macbeth_F_ow24plates.tiff). (G) Orsonwelles macbeth, female. Hawaii, Molokai, Kamakou Preserve. Runway into retreat is in the upper part of the web (Orsonwelles macbeth_F_ow25plates.tiff).

Figure 65.

Pityohyphantes webs. (A–C) Pityohyphantes costatus (Hentz, 1850), juvenile. USA, Virginia, Jefferson National Forest, Mtn Lake Biological Station (DSC_2798.NEF, DSC_2802.NEF, DSC_2796.NEF). (D, E) Pityohyphantes costatus, subadult male. USA, Virginia, Jefferson National Forest, Mtn Lake Biological Station (DSC_2807.NEF, DSC_2814.NEF). (F–H) Pityohyphantes costatus, female. USA, Maryland, Patuxent Wildlife Research Center (GH940510_R00_02_USA_Pityohy_cost.TIF, GH940510_R00_09_USA_Pityohy_cost.TIF, GH940510_R00_04_USA_Pityohy_cost.TIF).

Lines Radiating from a Corner. Many lines radiated from one corner of the sheets (where the spiders were usually found) in Agyneta spp. Colombia, Costa Rica, Maryland, Hawaii, and South Africa (Figs. 11, 12A, B, 13B–D); this pattern did not occur in sheets of other Agyneta species from Hawaii (Fig. 13E) and Maryland (Figs. 12G, H). This radiating pattern has not been seen outside the genus Agyneta.

Frame Lines. Often sheets had long frame lines at their edges, as in Grammonota sp. (Fig. 28A), Floronia bucculenta (Figs. 26D, E), and Bathyphantes pallidus (Fig. 20A). Other sheets had no frame lines, and the edge of the sheet was anchored to the substrate by many separate lines, as in Novafrontina sp. Brazil (Fig. 57C) and Australolinyphia remota (Fig. 19E). In some cases, as in Walckenaeria(?) sp. (Figs. 77D, E), the same sheet lacked frame lines at one edge but had clear frame lines at another. In Sphecozone ardens (Fig. 63B) the frame lines were all relatively small. Another variant, seen in Laminacauda magna (Fig. 42A) and Australian genus 1 spp. GH03 and GH05 (Figs. 17B, E, F), had long marginal pseudo-frame lines at the extreme edge; the mesh of the sheet was so sparse at the edge, and these lines supported very few sheet lines.

Figure 17.

Australian genus 1 webs (3). (A, B) Australian genus 1 GH03, female. Australia, Victoria, Dandenong National Park (DSC_1495.NEF, DSC_1497.NEF). (C) Australian genus 1 GH03, female. Australia, Victoria, Dandenong National Park (DSC_1485.NEF). (D) Australian genus 1 GH05, female. Australia, Tasmania, Weldborough Pass Scenic Reserve (DSC_0362.NEF). (E, F) Australian genus 1 GH05, female. Australia, Tasmania, Mt. Field National Park (DSC_1592.NEF, DSC_1596.NEF).

In some sheets with frame lines, many lines in the sheet made V-shaped intersections with the frame, as in Agyneta sp. Guyana (Fig. 12H), Diplothyron sp. Costa Rica (Fig. 20G), and Dubiaranea sp. Ecuador (Fig. 24E). V-shaped attachments presumably resulted when, during sheet construction, the spider attached the line it was laying to the frame and then turned back sharply (Benjamin and Zschokke, 2004). Nearly identical patterns occur at the edges of some pholcid and theridiid webs (Eberhard, 1992, 2020), and direct behavioral observations of one pholcid confirmed that the “V” patterns were also produced when the spider turned back after reaching the edge of the sheet (Eberhard 1992). In other webs, sheet lines made few or no V-shaped intersections with the frame, as in Bathyphantes pallidus (Figs. 20B, C), Labulla thoracica (Fig. 31B), and Dubiaranea sp. Costa Rica (Fig. 22F). Some sheets had a mix of multiple V-shaped intersections and few V-shaped intersections on different frame lines, as in Pityohyphantes costatus (Fig. 65A, right and left, respectively), Australian genus 1 sp. GH03 (Figs. 17A–C), and Labulla thoracica (Fig. 31G, left and lower edges).

Some sheets had secondary frame lines, relatively straight lines near the edge of the sheet that were attached at both ends to a frame line and that had multiple sheet lines attached to them, as in the near edge of the Orsonwelles polites web in Figure 62D. The edge of the dense central portion of the sheet was thus scalloped and was attached to a few, perhaps stronger frame lines that formed a second, more peripheral edge.

Parallel Lines. Some sheets had sets of two or more nearby parallel lines that ran nearly parallel to each other for short distances, sometimes straight and sometimes in curves, as in Diplothyron sp. (Fig. 20G) and Agyneta sp. Maryland (Fig. 12H). Perhaps the most striking parallel lines were the pairs of sticky lines that wandered across the sheet of Dubiaranea lugubris (Figs. 22C, D). We do not know whether these pairs were laid simultaneously or sequentially.

Skeleton Web Lines. Some sheets included possible skeleton web lines that were longer and straighter than other lines in the same sheet, as in Labulla thoracica (Fig. 31G); some skeleton web lines were approximately parallel to a nearby frame line, as in Dubiaranea sp. Ecuador (Fig. 25C). In contrast, no possible skeleton web lines were visible in other sheets, as in Australian genus 1 spp. GH03 and GH05 (Figs. 17B, F).

Repairs. Apparent repairs of holes in the sheet occurred in several species, including Floronia bucculenta (Fig. 26E), and Orsonwelles falstaffius Hormiga, 2002 (Figs. 59D, E). One web of Tenuiphantes flavipes (Figs. 77A, C) had three zones with different densities, perhaps corresponding to successive repairs.

Figure 59.

Orsonwelles webs (1). (A) Orsonwelles ambersonorum Hormiga, 2002, female. Hawaii, Oahu, Mount Tantalus (Orsonwelles ambersonorum_F_ow31plates_E.tiff). (B) Orsonwelles calx Hormiga, 2002, subadult female. Hawaii, Kauai, Laau Ridge (Orsonwelles calx_F_ow32plates_E.tiff). (C) Orsonwelles falstaffius Hormiga, 2002, female. Hawaii, East Maui, Waikamoi Preserve (Orsonwelles falstaffius_F_ow35plates_E.tiff). (D, E) Orsonwelles falstaffius Hormiga, 2002, juvenile, detail of repair work in main platform. Hawaii, East Maui, Waikamoi Preserve (Orsonwelles falstaffius_J_ow38plates_E.tiff, Orsonwelles falstaffius_J_ow37plates_E.tiff). (F) Orsonwelles falstaffius Hormiga, 2002, subadult female. Hawaii, East Maui, Waikamoi Preserve (Orsonwelles falstaffius_sF_ow36plates_E.tiff). (G) Orsonwelles falstaffius Hormiga, 2002 (no sex information). Hawaii, West Maui, Puu Kukui (Orsonwelles falstaffius_X_ow39plates_E.tiff).

Dimples. A dimple was where a “tensor line” (Suter, 1984) that attached more or less perpendicularly to the sheet pulled the sheet out of the plane of nearby portions of the sheet. Most tensor lines were attached to tangle lines, but the lines pulling the sheet into dimples ran directly to the substrate in Neriene montana (Nielsen, 1932, vol. 1, fig. 59).

Downward-directed dimples pulled the sheet downward in areas where the sheet curved upward, as in cup-shaped sheets in Frontinella spp. (Figs. 27A–D), Dubiaranea Trinidad and Tobago (Figs. 24B, C), and Acroterius sp. GH02 (Fig. 78). Upward-directed dimples pulled the sheet upward in some dome-shaped sheets, as in Diplothyron dianae (Fig. 20D), Laetesia raveni (Fig. 34B), and Laetesia sp. GH02 (Fig. 34E). These dimples were not as pronounced as many downward dimples and were sometimes nearly imperceptible, as in Neriene litigiosa (Fig. 53G). The inverse relation between the directions of dimples and the local curvature of the sheet is logical if dimples are sites where the sheet is being pulled toward a tangle (Suter, 1984) (see Discussion). There were, however, exceptions. Some dimples occurred in relatively flat sheets, as in Neriene oxycera Tu & Li, 2006 (Fig. 54C), and Laperousea sp. GH1677 (Fig. 45C); as just noted, some strongly curved sheets lacked large dimples, as in Neriene litigiosa (Fig. 53G).

Figure 54.

Neriene webs (5). (A) Neriene oxycera Tu & Li, 2006, male and female. Thailand, Chiang Mai, Doi Chiang Dao (GH031002_R01_14_THA_.TIF). (B, C) Neriene oxycera, female. Thailand, Chiang Mai, Doi Chiang Dao (GH031002_R01_22_THA_.TIF, GH031002_R01_24_THA_.TIF) (D, E) Neriene oxycera, female. Thailand, Chiang Mai, Doi Chiang Dao (GH031002_R01_28_THA_.TIF, GH031002_R01_31_THA_.TIF).

Some dimples were large, as in Laperousea sp. GH01 (Fig. 45F), Neriene clathrata (Fig. 51B), Neriene oxycera (Figs. 54B, C), and Linyphia triangularis (Fig. 46C). Others were of more moderate size, as in Australian genus 4 (Figs. 19A–C) and Laetesia sp. GH02 (Fig. 34E), and some were small and barely perceptible, as in Neriene litigiosa (Fig. 53G) and Dubiaranea distincta (Nicolet, 1849) (Fig. 21E). Other less pronounced local irregularities in the slope were abundant in some sheets but were not counted as dimples. Concentrations of tangle lines, curves in the sheet surface, and lack of depth of field probably obscured dimples in some web photos.

Figure 21.

Dubiaranea webs (1). (A–E) Dubiaranea distincta (Nicolet, 1849) (= Dubiaranea aysenensis (Tullgren, 1902)), Chile, Parque Nacional Puyehue. (A, B) Female (GH001231_R04_28_CHI_Dubiar_aysens.tif, GH001231_R04_25_CHI_Dubiar_aysens.tif). (C) Male and subadult female (GH001229_R01_17_CHI_Dubiar_aysens.tif). (D, E) Female (GH010102_R06_17_CHI_Dubiar_aysens.tif, GH010102_R06_19_CHI_Dubiar_aysens.tif).

Figure 46.

Linyphia and Lepthyphantes webs. (A, B) Linyphia triangularis (Clerck, 1757), female. Sweden, Tullbotorp (GH940815_R00_17_SWE_Linyph_triangul.TIF, GH940815_R00_20_SWE_Linyph_triangul.TIF). (C) Linyphia triangularis, female. Sweden, Tullbotorp (GH940815_R00_04_SWE_Linyph_triangul.TIF). (D, E) Linyphia triangularis, female. Hestehaven Forest, Denmark (GH940901_R00_20_DEN_Linyph_triangul.tif, GH940901_R00_21_DEN_Linyph_triangul.tif). (F) Lepthyphantes turbatrix (O. Pickard-Cambridge, 1877), male. USA, Virginia, Jefferson National Forest (DSC_2820.NEF).

Runways. A runway (termed a “funnel” by Hormiga, 2002) was a relatively densely meshed extension of the edge of a sheet to a sheltering object like a tree trunk where the spider waited. Runways were present at the edges of the sheets of Orsonwelles polites (Fig. 62) and O. malus Hormiga, 2002 (Fig. 61D, lower left), and Australian genus 1 sp. GH01 (in which the short runway was expanded slightly at its tip) (Figs. 15C–E). Some runways were difficult to distinguish from tubes (above).

Perforated Sheets. Nearly all sheets were continuous, without any objects such as a twig passing through them or through their tangles. In a few, an object such as a twig penetrated or passed through the sheet. Microlinyphia dana (Fig. 49I) and Juanfernandezia melanocephala (Figs. 29A–C) had a small open space in the sheet around the object. In contrast, the sheet of Tapinopa bilineata (Figs. 76E, F) had no openings. Given that these objects all appeared to be rigid and probably immobile, the open spaces may have resulted from the spider's construction behavior, rather than from damage to the sheet produced by movements of the object. The holes around penetrating objects in the sheets of two other linyphiids, Neriene coosa and Neriene litigiosa, may have been caused by movements of the flexible plant structures that frequently penetrated sheets (Eberhard, 2022; W. Eberhard, unpublished).

Figure 49.

Neomaso and Notholepthyphantes webs. (A–C) Neomaso pollicatus (Tullgren, 1901), female. Chile, Zapallar (DSC_2280.NEF, DSC_2298.NEF, DSC_2309.NEF). (D) Neomaso patagonicus (Tullgren, 1901), female. Chile, Puyehue National Park (GH001229_R01_31_CHI_Neomaso.tif). (E) Neomaso sp., subadult female. Chile, Puyehue National Park (GH001229_R01_31_CHI_Neomaso.tif). (F, G) Notholepthyphantes australis (Tullgren, 1901), female. Chile, Puyehue National Park (GH001231_R04_21_CHI_Notholep_austr.TIF, GH001231_R04_19_CHI_Notholep_austr.TIF). (H, I) Notholepthyphantes australis, female. Chile, Puyehue National Park (GH010102_R06_12_CHI_Notholep_austr.TIF, GH010102_R06_13_CHI_Notholep_austr.TIF).

Central Hub-like Areas. Another study described hub-like areas in the central, uppermost portions of the domes in Diplothyron simplicatus (F. O. Pickard-Cambridge, 1902) and Neriene coosa, where the spider rested (Eberhard, 2022). The areas usually had one or more holes, and the lines seemed to be less dense and shorter (Eberhard, 2022), giving the impression that the spider may have removed and then replaced lines. It was not possible, however, to follow individual lines reliably, even in sheets built in captivity, so this impression could not be quantified. Similar apparently modified areas occurred in Florinda coccinea (Hentz, 1850) sheets built in captivity (W. Eberhard, unpublished).

Similar hub-like modifications were seldom clear in this study. One possible example is Australian genus 1 sp. GH05 (Fig. 17D); less certain cases occurred in the sheets of Dubiaranea insulana (Fig. 23D, at the peak of the largest dimple), Laetesia sp. GHG02 (Figs. 35A, E, in the middle near the top of the dome), and close to the very center of the sheet of both Australolinyphia remota (Fig. 19G) and Australian genus 2 sp. GH01 (Fig. 14E). One possible explanation for this difference is that most of the closeup photos of sheets in this study were not of the central area but rather the edges, where tangle lines did not obscure the sheet. Other possible observational difficulties resulted from apparent raindrop damage, as in Laminacauda propinqua (Fig. 39C), repairs of damage to sheets (above), and dense tangle lines that obscured the sheet structure. Nonetheless, it was clear that the central area was not modified in some clear photos of apparently intact, naked sheets in Australian genus 1 sp. GH03 (Figs. 17A, B), Laminacauda sp. “chdes” (Fig. 38F), and (probably) Dubiaranea sp. Ecuador (Fig. 25C). In general, it is uncertain whether the species in this study had similar hubs.

Figure 35.

Laetesia webs (4). (A, B) Laetesia GH02, female. Australia, Queensland, Atherton Tablelands (DSC_8056.NEF, DSC_8051.NEF). (C, D) Laetesia GH02, male, female. Australia, Queensland, Atherton Tablelands (DSC_8040.NEF, DSC_8045.NEF). (E) Laetesia GH02, female. Australia, Queensland, Atherton Tablelands (DSC_8049.NEF).

Another possible reason for a lack of a modified central area would be if the spider rests at the edge of the web rather than in the center. Thus, Diplothyron dianae (Fig. 20B), Labulla thoracica (Figs. 31B, F), and Himalaphantes sp. (Fig. 28E) all apparently lacked a modified central area, and their sheets were especially dense at one protected edge where it abutted against the substrate (Figs. 20A, upper edge; 31B and 28E, upper right; 31F, middle right). Further observations will be needed to determine the frequency of hub-like areas and test the hypothesis that they are associated with the spider's resting site.

Sticky Lines. We can only comment briefly on the presence of sticky lines because they could not be distinguished from nonsticky lines when coated with powder. Many lines in the sheet of Dubiaranea lugubris bore droplets so large that they were easily distinguished with the naked eye (Figs. 22C, D). Smaller droplets that are difficult or impossible to distinguish with the naked eye are known in both the sheets and the tangles of several other linyphiid species (summary in Eberhard, 2021) and may well have been present in some (or even all) of the species illustrated here. An estimated 50% of lines in the sheet and at least some lines in the upper tangle of a web of a Frontinella pyramitela web built in captivity bore small droplets, and the majority of lines in the sheet and at least some of those in the upper tangle in a Florinda coccinea web built in captivity and in the sheet of Neriene litigiosa also bore droplets (W. Eberhard, unpublished). The Brazilian species Pocobletus riberoi (Lemos & Brescovit, 2013) builds a sheet web with an upper and lower tangle; lines in at least the upper tangle were sticky, even after it was lightly dusted with cornstarch (as assessed by light contact with the blunt metal end of a mechanical pencil in three different webs).

The triad of spigots on the posterior lateral spinnerets, which is known in araneoid spiders to produce lines coated with droplets of sticky liquid, is present in almost all linyphiids that have been examined (e.g., Hormiga, 1994b, 2000), including the genera whose webs are illustrated here, and droplets on lines have been observed directly in others (summarized in Eberhard, 2021). The presence of triads in many of the species of genera whose webs are illustrated here, including Agyneta, Diplothyron, Frontinella, Laminacauda, Lepthyphantes, Linyphia, Microlinyphia, Neriene, Orsonwelles, Ostearius, and Walckenaeria, implies that many of these webs had sticky lines. Tiny droplets of liquid also occur in the webs of some theridiids, such as Anelosimus spp. and Tidarren sisyphoides (Walckenaer, 1841) (Madrigal-Brenes and Barrantes, 2009).

Tangles

A “tangle” was defined as a three-dimensional network with no perceptible organization. Because of the large intraspecific variation in the outlines of the tangles above and below sheets (e.g., Figs. 58, 78, 79A, B; also Eberhard, 2022), we did not include the forms of the outlines of tangles (triangular, etc.) in our analyses. The tangle above the main sheet was termed the upper tangle; one below the main sheet was the lower tangle. (These tangles have also been called the upper and lower scaffolding; Arnedo et al., 2009). It is worth noting that linyphiids and pimoids always moved under their sheets (except for brief forays to attack prey in the upper tangle), which means that the lines above the sheet (and also most if not all of those below the sheet) were presumably the result of building behavior and not incidental drag lines left behind as the spider moved about its web.

Figure 58.

Novafrontina webs (2). (A, B) Novafrontina uncata (F. O. Pickard-Cambridge, 1902), female. Costa Rica, Estación Biológica La Selva (GH930406_R08_34_CRI_Novafro_uncata.tif, GH930406_R08_35_CRI_Novafro_uncata.tif). (C, D) Novafrontina uncata, female. Costa Rica, Sirena, Parque Nacional Corcovado (GH930303_R00_01_CRI_Novafro_uncata.tif, GH930303_R00_15_CRI_Novafro_uncata.tif). (E, F) Novafrontina uncata, subadult female. Costa Rica, Estación Biológica La Selva (GH930406_R08_26_CRI_Novafro_uncata.tif, GH930406_R08_31_CRI_Novafro_uncata.tif). (G, H) Novafrontina uncata, female. Costa Rica, Sirena, Parque Nacional Corcovado (GH930303_R00_29_CRI_Novafro_uncata.tif, GH930303_R00_23_CRI_Novafro_uncata.tif).

Upper Tangles. Upper tangles took many forms. Tall upper tangles were more than a sheet diameter tall, as in Acroterius spp. (Figs. 78, 79B), Pocobletus sp. GH02 (Fig. 74C), Diplothyron sp. (Fig. 20D), and Novafrontina uncata (F. O. Pickard-Cambridge, 1902) (Fig. 58A). Tall upper tangles such as those of Pocobletus sp. Dominican Republic were consistently dense (Fig. 68B) or relatively dense (Fig. 68C). At the other extreme were low upper tangles, which were usually sparse, as in Laetesia raveni (Fig. 33B), but sometimes dense, as in Novafrontina uncata (Fig. 58H). Some sparse upper tangles were restricted to locations just above the site near the edge of the web where the spider apparently rested, as in Agyneta spp. (Figs. 11C, D, 12A, B) and Dubiaranea sp. Ecuador (Fig. 25B). (This spider rested off the web during the day but in the center at night.) Some other apparent retreats lacked tangles, however, as in Agyneta sp. (Figs. 11A, B, E–H). One such tangle contained two white egg sacs (Fig. 11D). (In contrast with Theridiidae, linyphiid egg sacs are seldom associated with webs; e.g., Nielsen, 1932.) Upper tangles never included sheet-like structures like those in some lower tangles (below).

Figure 68.

Pocobletus webs (3). (A) Pocobletus sp. GH33, male. Dominican Republic, Sierra de Bahoruco, Reserva Cachote (GH050408_R01_36_DR_Pocobletus_ED.jpg). (B) Pocobletus sp. GH33, female. Dominican Republic, Sierra de Bahoruco, Reserva Cachote (GH050409_R02_36_DR_Pocobletus_ED.jpg). (C, D) Pocobletus sp. GH33, female. Dominican Republic, Sierra de Bahoruco, Reserva Cachote (GH050409_R03_01_DR_Pocobletus_ED.jpg, GH050409_R03_07_DR_Pocobletus_ED.jpg). (E, F) Pocobletus sp. GH01, female. Panama, Parque Nacional Altos de Campana (DSC_2693_ED.NEF, DSC_2695_ED.NEF).

In general, lines in the upper tangle did not have discernible patterns, but a few upper tangles were largely composed of long straight, approximately vertical lines, as in Acroterius sp. (Fig. 78) and Dubiaranea sp. Ecuador (Fig. 23G).

Lower Tangles. Lower tangles were usually less dense than upper tangles, although Australian genus 3 sp. GH02 (Fig. 18F) and Laminacauda sp. “fCh” (Fig. 43E) were exceptions. Some lower tangles included sheets (see “double sheets” above), but the distinction between lower sheets and tangles was sometimes difficult; some sheets were so sparse (e.g., Figs. 55G, H of Neriene helsdingeni) that it was not easy to distinguish them from small tangles. The sizes, locations, and forms of lower sheets varied. The dense lower tangle of one web of Australian genus 3 sp. GH02 lacked a second sheet but had a curved upper margin that was parallel to and just below the principal sheet (Fig. 18F). Some cup-shaped sheets had a centered lower tangle that was limited to the space directly below the lowest portion of the cup (where spiders generally rested), as in Laminacauda sp. “fCh” (Figs. 43D, E). When the sheet of Agyneta semipallida (Chamberlin & Ivie, 1944) (Figs. 11E, F) was carefully peeled away, a sparse tangle was revealed below. In some photos it was not possible to check for lines directly below the sheet, so similar inconspicuous tangles may have occurred but been missed in some webs (e.g., other Agyneta species).

Combinations of Sheet and Tangle Traits

There were many combinations of the different sheet and tangle traits. To cite just a single example, the web of Pocobletus sp. Dominican Republic (Figs. 68C, D) was an elevated, moderately cup-shaped, dense unobstructed sheet with long frame lines; a wide, tall, dense upper tangle; and a less dense, narrow lower tangle that was centered under the lower portion of the cup. Many different combinations are noted in the discussions below. Further work will be needed to determine whether, as seems likely, some taxa are characterized by particular combinations of these web traits.

Cyatholipidae

As noted in the introduction, Cyatholipidae appears to be the family most closely related to linyphioids. The webs of one species in each of five cyatholipid genera are illustrated in Figures 3–7. These webs share several traits seen in linyphioid webs. They all have elevated, nearly horizontal, flat or nearly flat, unobstructed, and densely meshed sheets that lack dimples. Apparent repairs of holes were visible in the sheets of Forstera sp. (Figs. 3A, B, D–F) and Wanzia sp. (Figs. 7E–G). The density of lines was reduced near the edge of the sheet in Forstera sp. (Figs. 3D, F, 6D). Frame lines with “V” attachments and scattered cases of parallel lines were clear in areas of one web of Forstera sp. (Fig. 3) where density was lower.

One difference with nearly all linyphioid webs was the following combination of traits: Forstera sp., Matilda sp., Ulwembua sp., and one web of Wanzia were nearly (Fig. 5A) or entirely naked (Figs. 3A, B, D, 4A, B, 6A–D, 7A) and were also (except for Matilda sp.) substantially elevated above the ground. Teemenaarus sp., however, had a tangle above and below the sheet, with an open space just below the sheet (Fig. 4D), as in linyphioids. Tekelloides sp. differs from all known linyphioids in that the spider rests on lines in the sparse tangle below its dense sheet rather than on the sheet itself (Forster and Forster, 2005). One web of Wanzia fako Griswold, 1998, had a nearly planar tangle below one side of the sheet (Figs. 7B, C) (perhaps below the spider's resting site?) but none above. Although we do not have quantitative data, in many cases the sheet seemed to be larger relative to the spider's size than is typical for linyphioids (e.g., in Fig. 4A the spider is minute relative to the size of the sheet). Cyatholipids generally rested near the central portion of the sheet.

Physoglenidae

As noted in the introduction, physoglenids may be closely related to the group linyphioids + Cyatholipidae. The webs of one species in each of seven genera are illustrated in Figures 8–10. Physoglenid webs shared several traits with those of cyatholipids and linyphioids. All webs included an elevated, approximately horizontal sheet that was sparse to moderately dense. In all cases in which it was possible to judge, the sheet was at least weakly domed. The spiders rested at the peaks of these domed sheets (Figs. 8A, D). The sheet of Pahora sp. had a tangle above and below (Fig. 8E).

There were also several differences. All physoglenid sheets were less dense than all cyatholipid sheets. They also differed from cyatholipids in lacking clear “V” attachments to the frame lines. The webs of Chileotaxus sans Platnick, 1990, differed from all linyphioids and cyatholipids in forming a sharply curved dome that extended in places into a large nearly horizontal sheet with a very low, sparse upper tangle (Figs. 8A, B). The top of one such dome was just under a plant leaf (Fig. 8B); these web traits are very similar to those of the pholcid Modisimus bribri Huber, 1998 (Eberhard, 2020). The sheets of Mangua medialis Forster, 1990, and Pahora sp. were also domed but had taller tangles above the sheet (Figs. 8D, E), as did the less sharply domed sheet of Paratupua sp. (Fig. 9E). In Physoglenes puyehue Platnick, 1990 (Figs. 9A, B), and Paratupua sp. (Fig. 9D) the upper tangles were sparse and limited to only the higher portions of the weakly domed sheets.

Diversity of Web Forms

Nearly all of the species in this study built sheets, and nearly all of these sheets were continuous and more or less horizontal. Thus, to a first approximation, this survey confirms the view that linyphioid webs are relatively uniform. The sheets and the lines accompanying them varied, however, in many respects, including the following: presence and relative size and density of tangles above or below the sheet; placement, especially close to the substrate, whose contours define the contour of the web (defined as “substrate webs” in Hormiga [2007] and Arnedo et al. [2009]); densities and several geometric patterns of lines in the sheet; coatings of “slime” or sticky droplets on lines in the sheet; the presence of a runway (that is sometimes tube- or funnel-like) from a more sheltered site to the sheet; the degree of definition of the plane itself; the presence of downward- or upward-projecting dimples in the sheet; the presence of a cylindrical retreat at one edge of the sheet; curves in both the main sheet and lower sheets (cup-shaped, dome-shaped, flat, etc.); perforation of the sheet by objects like twigs and leaves; and the presence of additional sheets and their relative sizes and densities.

Even though current knowledge of linyphiid webs is still only fragmentary, it is clear that the label “sheet web” has hidden a great deal of diversity. This diversity in linyphiid webs resembles, although at a smaller scale, the same pattern of diversity seen in spider webs in general, and in sheet webs in particular (Eberhard, 2020). Previous statements to the effect that linyphiid webs lack diversity are thus only partially correct. Linyphiid webs do indeed usually include sheets, but they have many variations on this basic design. The term “sheet web” has thus been something of a linguistic trap that has hidden many different designs under a single name. For further discussions of problems associated with the term “sheet,” see Peters and Kovoor (1991), Blackledge et al. (2009, supplementary materials), Eberhard and Hazzi (2017), and Eberhard (2020).

It is worth noting, however, that this variation may have limitations: no known webs with a dome-shaped principal sheet have a second sheet; flat sheets near the ground nearly always have little or no upper tangle; in cases of intraspecific variation in the presence of a sheet, the lower sheet, but never the principal sheet, is sometimes reduced to only a tangle; in only a single species (Tekelloides sp., a cyatholipid rather than a linyphioid) does the spider rest in a tangle below a sheet rather than on the sheet itself; to the best of our knowledge, the same species never builds both cup- and dome-shaped sheets.

Intrageneric Similarities and Differences

If one assumes that the species that are presently grouped in genera are each other's closest relatives (i.e., the genera are monophyletic), data are now sufficient to begin searching for patterns in web evolution by checking for similarities and differences within genera. Although phylogenetic analyses have tested the monophyly of some of the genera treated here, such as Orsonwelles (Hormiga et al., 2003), Laminacauda (Arnedo and Hormiga, 2021), and Diplothyron (Silva-Moreira and Hormiga, 2022), most linyphiid genera have not received a phylogenetic treatment. We discuss below the genera in which the webs of more than a single species have been documented, in decreasing order of the apparent intrageneric diversity, and of the sample size by genus. We include information from other publications with web photographs and mention the sample size for each species and the fraction of species in the genus whose webs are known to emphasize the limits of current knowledge. (Webs are known in only two or three species in 14 of the 21 genera.) Both species and genera can be expected to display greater ranges of diversity when larger samples become available (see the section on intraspecific variation below).

Linyphiidae

High Intrageneric Diversity. Laminacauda (Webs of 11–12 species). The erigonine genus Laminacauda comprises 40 species, mainly from South America, but new species remain to be described (G. Hormiga, unpublished). Some of the undescribed species are included in our sample, which is largely composed of species endemic to the south Pacific archipelago of Juan Fernandez. The 11–12 species of Laminacauda (Figs. 36–44) show the greatest intrageneric diversity yet documented in a linyphiid genus. The diversity of species and webs in Juan Fernandez is most likely the result of an adaptive radiation in depauperate environments (Arnedo and Hormiga, 2021). Most sheets were medium to sparsely meshed and highly elevated with moderately sparse tangles, but with exceptions (below), and they showed many other differences that were so striking that nearly every species (and even some different conspecific individuals) exhibited distinctive variations in design, to the extent that some of the species in Juan Fernandez could be identified by just examining their webs. The upper end of the domed sheet extended into a tube in some L. ansoni (Figs. 36A, D, E); in one of these webs (but not the others) the tube reached the substrate (the trunk of a tree), and often these silk tubes led to retreats in circular openings on tree trunks made by xylophagous insects. No sheet was discernible, however, in another L. ansoni web (Fig. 36C), and the sheet of still another individual of the same species appeared to be roughly cup-shaped (Fig. 36B).

The sheets of L. malkini were flat, relatively small, and sparse, with very sparse upper tangles. One sheet formed a tube at one edge (Fig. 37C), whereas others lacked even a hint of a tube (Figs. 37E, F). The sheets of L. rubens (Figs. 38A–E) were unique in being vertical and very close (on the order of the length of a spider or less) to the trunk of a tree. The upper edge of some sheets was curved to form a tube in which the spider rested (Figs. 38B, C), but others lacked a clear tube (Fig. 38A). A tangle “above” the sheet varied greatly in its extension (compare Fig. 38A with Fig. 38D). In addition to the data provided by the webs photographed, the webs of many other individuals of L. ansoni and L. rubens were examined, because both species are relatively common in the native forests of Robinson Crusoe Island. The sheet of Laminacauda “chdes” (also from Robinson Crusoe Island) had a different, ridge tent–like form just above the leaf litter; an uppermost central ridge slanted upward toward a large leaf with no discernible tangle lines (Fig. 38F). The web of L. tuberosa was, in contrast, unremarkable, with a dense weakly cupped sheet that sloped slightly upward at some edges, with a very sparse upper tangle and moderate lower tangle (Fig. 39H). The webs of L. propinqua (a species endemic to Alejandro Selkirk Island) were weakly domed sheets near the leaf litter whose mesh varied from dense (Fig. 39C) to only moderate (Figs. 39D, F). Upper tangles were absent or only sparse above. One sheet had plant stems projecting through it (Fig. 39B).

All six webs of Laminacauda magna (another Robinson Crusoe endemic) had a naked or nearly naked, medium-meshed, bumpy horizontal sheet near the ground (Figs. 40–42). Two of these sheets had small glistening patches (Figs. 40B, 42A, B) that were not perceptibly sticky, but similar patches were absent in other webs of this species. A horizontal view of one web revealed that it was a “sandwich,” with a second, less densely meshed horizontal sheet immediately below the principal sheet (Fig. 41D). It was difficult to distinguish the presence of the lower sheet in a vertical view of this same web from above (Fig. 41C), leading us to question whether some or all of the other webs of this species (all seen from above) may have also had second lower sheets. This possibility of a second sheet seemed unlikely, however, in some especially clear photos (Figs. 40A, 42A).

One web in each of two other Robinson Crusoe species, Laminacauda sp. “fPC” and Laminacauda sp. “fCh” (Fig. 43), resembled those in some other genera outside the subfamily Erigoninae, with highly elevated, dense, moderately cup-shaped sheets with multiple dimples; a very sparse upper tangle; and a moderately dense lower tangle. In Laminacauda sp. “fPC” the lower tangle was as wide as the sheet, whereas in Laminacauda “fCh” the lower tangle was limited to the central portion of the sheet under the lowest portion of the cup.

In some senses, the most distinct web was that of the continental species L. parvipalpis, whose five webs were all near the ground. Three webs (Figs. 44B, C, E) appeared to be more or less horizontal sheets, whose sparse mesh became gradually more dense near one edge, giving the impression that there was a retreat under the adjoining leaf litter (there is an upper tangle in Fig. 44B, little if any above Fig. 44C, and none in Fig. 44E). This sheet design differed from all other known linyphiid webs. One other web (Fig. 44D) was a loose tangle in which we could not discern any organization (perhaps because it was in the early stages of web construction); the other web was even more difficult to describe (Fig. 44A).

One web of a Laminacauda newtoni Millidge, 1985, from continental Chile (Miller, 2007: fig. 1A) also suggested a sparsely meshed, perhaps naked sheet near the substrate whose more sheltered side abutted on the substrate and was somewhat more densely meshed, but details could not be resolved confidently. The web of an additional Laminacauda sp., also from continental Chile (Miller, 2007: figs. 1B, C), was near the substrate, but the photo was not clear enough to resolve details other than long lines in what appeared to be an upper tangle. The simple, somewhat irregular webs of these three continental species are representative of the webs of other continental species and of other South American erigonines that live in the leaf litter or among rocks and possibly represent the plesiomorphic web architecture of Laminacauda.

In sum, the webs of Laminacauda species differed in many respects, including web site (near leaf litter, highly elevated), sheet mesh (sparse, dense, asymmetrically dense near one edge, shiny patches), sheet form (cup, dome, flat, curled to form lateral tube, tube-shaped extension of top of dome, sandwich), and sheet orientation (horizontal, vertical).

Acroterius (Webs of Two Species)

The recently erected genus Acroterius includes a dozen described species from China, but an unknown number of yet to be named species exist, including the two species from Taiwan and northern Thailand whose webs are illustrated here. The webs of these two species differ dramatically. The web of Acroterius sp. Thailand (Figs. 65F, G) had a weakly domed, only moderately densely meshed sheet, a moderate tangle above, and perhaps little or no tangle below. The sheet was pulled upward in numerous small dimples (Fig. 65G). In contrast, nine webs of Acroterius sp. GH01 (from Taiwan) (Figs. 78, 79) had a dense sheet whose form ranged from sharply cupped to nearly flat (Fig. 79F). All of the sheets had multiple downward-directed dimples. Seven of these webs had a second, less dense, and smaller diameter cup-shaped sheet immediately below the larger sheet, although in some cases it was strongly reduced (Fig. 78); the second sheet was apparently lacking, however, in the other two webs, which had only a sparse tangle below the sheet (Figs. 79E, F). All webs had an upper tangle but its height varied. In one case it extended more than two sheet diameters above (Fig. 78); in other webs it was smaller (Figs. 79A–C). The density of the upper tangle ranged from dense (Figs. 78, 79B) to sparse (Fig. 79E). None of the Taiwan webs was dome-shaped.

Moderate Intrageneric Diversity. Agyneta (Webs of 12 Species). Agyneta is a large genus with about two hundred described species and a world distribution; many undescribed species are presumed to exist. Twelve species of Agyneta showed a mix of consistency and variation in their webs. All webs were naked or nearly naked planar sheets near the ground. The sheets of perhaps eight species had the otherwise unique pattern of many lines radiating from one corner of the sheet, which appeared to have a small tangle retreat (Figs. 11–13). In three of these species a small tangle was immediately above the retreat (Figs. 11C, D, 13A–C). Three other species, however, had similar naked planar sheets near the ground but lacked the converging lines (Figs. 12G, H, 13E, F), and one (Fig. 13D) was more elevated. A 12th species differed strikingly (Figs. 12D–F). It had a unique trough-like sheet built close to the upper surface of a curled leaf in the ground litter, with an extensive upper tangle that spanned the leaf's curl and a much smaller lower sheet. Pocobletus (Webs of 10?) Species, All but One Undescribed)

The American genus Pocobletus groups 13 described species, mostly form the Neotropical region, but many others remain to be described (G. Hormiga, unpublished). Six webs of Pocobletus sp. GH32 from Brazil (Figs. 66 A–F, 67A, B) had highly elevated, flat or very weakly cupped dense sheets with few or no dimples. They were unusual in having lower tangles that were more extensive than the upper tangles; one web lacked an upper tangle altogether (Fig. 66C). Four of the lower tangles were more or less limited to the central, lowest area of the sheet, presumably where the spider rested. Five webs of Pocobletus sp. GH33 from the Dominican Republic were similar in having highly elevated dense sheets and moderately dense lower tangles limited to the central portion of the sheet. They differed from the Brazilian species in having clearly cup-shaped sheets with downward-projecting dimples (especially in the central area in Fig. 68D), and in having moderately dense to dense upper tangles. Two webs of Pocobletus sp. GH02 from Costa Rica (Figs. 74C–F) were similar in having moderately to steeply sided cup-shaped sheets, with a few downward-directed dimples in the central portion. The lower tangles were denser in the central portion of the sheet but extended to the sheet's edges. In contrast, the lower tangle in one web of Pocobletus sp. GH05 from Colombia was limited to the central portion of the sheet, and there was little if any upper tangle. This web was similar to those just mentioned in having a highly elevated, dense, steeply sided, cup-shaped sheet (Fig. 72E).

Figure 66.

Pocobletus webs (1). (A, B) Pocobletus sp. GH32, subadult male. Brazil, Rio Jufari (DSC_9240.NEF, DSC_9237.NEF). (C, D) Pocobletus sp. GH32, female. Brazil, Rio Jufari (DSC_9278.NEF, DSC_9280.NEF). (E, F) Pocobletus sp. GH32, subadult male. Brazil, Rio Jufari (DSC_9287.NEF, DSC_9288.NEF).

Figure 67.

Pocobletus webs (2). (A, B) Pocobletus sp. GH32, male. Brazil, Rio Jufari (DSC_9294.NEF, DSC_9295.NEF). (C, D) Pocobletus sp. GH33, female. Dominican Republic, Sierra de Bahoruco, Reserva Cachote (GH050407_R01_03_DR_Pocobletus_ED.jpg, GH050407_R01_08_DR_Pocobletus_ED.jpg). (E, F) Pocobletus sp. GH33, subadult male. Dominican Republic, Sierra de Bahoruco, Reserva Cachote (GH050408_R02_06_DR_Pocobletus_ED.jpg, GH050408_R02_12_DR_Pocobletus_ED.jpg).

Figure 72.

Pocobletus webs (7). (A, B) Pocobletus sp. GH03, female. Ecuador, Napo, Sierra Azul (GH960612_R01_27_ECU_Pocoblet_sp3ED.tif, GH960612_R01_32_ECU_Pocoblet_sp3ED.tif). (C, D) Pocobletus sp. exeGH03, female. Colombia, near Cali (GH980211_R00_19_COL_Exechop_sp3_ED.tif, GH980211_R00_21_COL_Exechop_sp3_ED.tif). (E) Pocobletus sp. exeGH05, female. Colombia, Cundinamarca (GH980201_R00_09_COL_Exechop_sp5_ED.tif). (F, G) Pocobletus versicolor (Millidge, 1991), juvenile. Costa Rica, Estación Biológica La Selva (GH930403_R06_01_CRI_Exechop_versic_ED.tif, GH930403_R06_07_CRI_Exechop_versic_ED.tif).

Four of the five Pocobletus sp. GH01 from Panama (Figs. 68E, F, 69B–D, F) resembled the Brazilian species in having highly elevated, dense, flat (Fig. 69D) or weakly cup-shaped sheets that may have had fewer dimples than the Dominican Republic species (Fig. 68F) and dense, moderately tall upper tangles. They differed in having especially tall, dense, and wide lower tangles that were less strictly limited to only the central area of the sheet; in three, the lower tangle was as dense or denser than the upper tangle (Figs. 68E, F, 69B, D). The other web (Fig. 69A) was too battered to be useful.

Figure 69.

Pocobletus webs (4). (A) Pocobletus sp. GH01 (female with egg sac, no voucher). Panama, Parque Nacional Altos de Campana (DSC_2866_ED.NEF). (B) Pocobletus sp. GH01, female. Panama, Parque Nacional Altos de Campana (DSC_2848_ED.NEF). (C, D) Pocobletus sp. GH31, juvenile. Panama, Parque Internacional La Amistad (DSC_4015_ED.NEF, DSC_4023_ED.NEF). (E) Pocobletus sp. GH10, female. Ecuador, Parque Nacional Yasuni (GH960621_R07_25_ECU_Pocoblet_sp10_ED.TIF). (F) Pocobletus sp. GH01, female. Panama, Parque Nacional Altos de Campana (DSC_2835_ED.NEF).

Four of the five webs of Pocobletus sp. GH11 from Guyana (Figs. 70A–F, 71A) also had highly elevated, dense, flat or weakly cup-shaped sheets with dense upper tangles, but they differed from all of the previous Pocobletus webs in that the lower tangle was a clear sheet (Figs. 70A–F, 71A). Three of four webs of Pocobletus sp. GH03 (Figs. 71F, G, 72A–D) also had sheet-like lower tangles that were just under the central portion of a highly elevated, dense, cup-shaped sheet. The slopes of the sides of the cup tended to be steeper than those of sp. GH11, varying from steep (Fig. 72E) to moderate (Fig. 72D) to weakly sloped (Fig. 72G). One web had numerous downward-directed dimples in the central two-thirds of the sheet (Fig. 72D). Four webs of still another species, P. versicolor, had clearly planar lower tangles (Figs. 72F, G, 73A–F). One of these webs was unique in having very long anchors that ran in only four different directions (Fig. 73E); the sheet of the lower tangle of this web was also unusual in being very close to the underside of the major sheet (Fig. 73F).

Figure 70.

Pocobletus webs (5). (A, B) Pocobletus sp. GH11, female. Guyana, Gunn's landing strip (GH990716_R07_26_GUY_Pocoblet_sp11_ED.TIF, GH990716_R07_31_GUY_Pocoblet_sp11_ED.TIF). (C) Pocobletus sp. GH11, female. Guyana, Gunn's landing strip (GH990706_R03_22_GUY_Pocoblet_sp11_ED.TIF). (D, E) Pocobletus sp. GH10, female. Ecuador, Parque Nacional Yasuni (GH960620_R07_13_ECU_Pocoblet_sp10_ED.TIF, GH960620_R07_16_ECU_Pocoblet_sp10_ED.TIF). (F) Pocobletus sp. GH11, female. Guyana, Gunn's landing strip (GH990706_R03_20_GUY_Pocoblet_sp11_ED.TIF).

Figure 71.

Pocobletus webs (6). (A) Pocobletus sp. GH11, female. Guyana, Gunn's landing strip (GH990717_R08_03_GUY_Pocoblet_sp11_ED.tif). (B) Pocobletus sp. GH28, female. Argentina, Iguazu (MR990713_X01_09_ARG_Pocoblet_sp28_ED.jpg). (C) Pocobletus sp. GH28, female. Argentina, Iguazu (MR990716_X00_01_ARG_Pocoblet_sp28_ED.jpg). (D) Pocobletus sp. GH28, sex? Argentina, Iguazu (MR990716_X00_03_ARG_Pocoblet_sp28_ED.jpg). (E) Pocobletus sp. GH28, female. Argentina, Iguazu (MR990716_X16_36_ARG_Pocoblet_sp28_ED.jpg). (F) Pocobletus sp. GH03 subadult female. Ecuador, Napo, Sierra Azul (GH960614_R00_25_ECU_Pocoblet_sp3_ED.tif). (G) Pocobletus sp. GH03, male. Ecuador, Napo, Sierra Azul (GH960614_R02_18_ECU_Pocoblet_sp3_ED.tif) (photos B–E by Martín J. Ramírez).

Three webs of Pocobletus sp. GH28 from Argentina (Figs. 71B–E), with dense, nearly flat sheets, differed from the others in being near the ground; three were next to tree trunks. They all had short upper tangles and a lower tangle that was sharply limited to the center of the sheet.

One web of Pocobletus sp. GH01 from Costa Rica (Figs. 74A, B) differed from the other Pocobletus species in being weakly domed rather than cupped; both the upper and lower tangles were limited to the central portion of the sheet. Although we do not have photographs, we observed multiple webs of Pocobletus riberoi at a single locality in the Atlantic forest in southern Brazil. Their webs, built on fallen trees and roots close to the forest floor, had upper and lower tangles and a horizontal, flat main sheet with downward-pointing dimples (pulled down by lines attached to the lower tangle). The upper and lower tangles were sometimes cone-shaped (with the bases facing the sheet). At least some of the silk lines of the upper tangle were sticky.

In sum, the webs of Pocobletus generally had cup-shaped sheets with tangles above and below, but most species had at least one detail that set them apart from the webs of other congeners. Traits that varied included elevation above the substrate; the height, width, and density of the lower tangle; the height and density of the upper tangle; the form of the sheet; and the presence and location of dimples in the sheet.

Australian Genus 1 (Webs of Five Species)

This undescribed genus is from eastern Australia and Tasmania that putatively groups many as yet undescribed species. Seven webs of species GH01, three webs of GH03, two webs of GH05, and one web of GH06 were all elevated, naked, or nearly naked, approximately horizontal sheets with long frame or pseudo-frame lines that were located alongside vertical tree trunks; the sheets had low- to medium-density meshes that were more sparse near some edges. The two webs of GH02, in contrast, were just above the forest floor and had a sparse upper tangle above the central portion of the sheet. At least two of the GH01 sheets had a small, densely meshed “runway” extension on the sheltered side (Figs. 15C–E) that expanded slightly near the trunk, apparently forming a shelter for the spider that rested there during daytime. Similar extensions may have also been present in two other GH01 webs (Figs. 14B, 16A), but the photos were not clear on this point. Appropriate closeup photos were available to check for runways in two other species: the sheet of GH05 had a runway (Fig. 17E), but that of GH03 lacked a runway (Figs. 17A, C).

Figure 15.

Australian genus 1 webs (1). (A–C) Australian genus 1 GH01, female. Australia, New South Wales, Dorrigo National Park (DSC_2776.NEF, DSC_2779.NEF, DSC_2780.NEF). (D–F) Australian genus 1 GH01, female. Australia, New South Wales, Dorrigo National Park (DSC_2794.NEF, DSC_2796.NEF, DSC_2801.NEF). (G, H) Australian genus 1 GH01, female. Australia, Queensland, Tamborine National Park (GH020417_R04_31_AUS_Laetesia_.tif, GH020417_R04_36_AUS_Laetesia_.tif).

Figure 16.

Australian genus 1 webs (2). (A, B) Australian genus 1 GH01, male and female. Australia (GH920718_R01_05_AUS_Laestesia.tif, GH920718_R01_10_AUS_Laestesia.tif). (C) Australian genus 1 GH01, female. Australia, Queensland, Lamington National Park (GH020414_R01_08_AUS_Laetesia_.tif). (D, E) Australian genus 1 GH02, female. Australia, Tasmania, Franklin-Gordon Wild Rivers National Park (DSC_0506.NEF, DSC_0511.NEF). (F, G) Australian genus 1 GH02, female. Australia, Tasmania, Cradle Mountain National Park (DSC_0268.NEF, DSC_0270.NEF).

In sum, the webs of different species in this genus were similar in being more or less naked, relatively sparsely meshed sheets. The sheets differed, however, in the presence of short runways at the edge.

Microlinyphia (Webs of Two Species)

This genus includes 12 described species from the Holarctic region and Africa. Two webs of the Malagasy Microlinyphia simoni van Helsdingen, 1970, were large (note the spider in Fig. 48E), very weakly domed sheets with tall, moderately dense tangles above and moderately dense tangles below (Figs. 48A–E). One web of Microlinyphia dana (Chamberlin & Ivie, 1943) from North America was more clearly domed, and the upper tangle was much shorter and concentrated in the central portion of the sheet (Fig. 48F).

Figure 48.

Microlinyphia webs. (A) Microlinyphia simoni van Helsdingen, 1970, female. Madagascar, Station Forestale Analamazaotra, Mitsinjo (DSC_5990.TIF). (B, C) Microlinyphia simoni, female. Madagascar, Station Forestale Analamazaotra, Mitsinjo (DSC_5921.TIF, DSC_5925.TIF). (D, E) Microlinyphia simoni, female. Madagascar, Ranomafana (DSC_5449.TIF, DSC_5454.TIF). (F) Microlinyphia dana (Chamberlin & Ivie, 1943), female. USA, Washington, Olympic National Park, Elwha River (GH900802_R00_31_USA_Microliny_dana.tif).

Lepthyphantes (Webs of Two Species)

Lepthyphantes is a large genus with more than 160 species described, mainly from the Holarctic region and Africa. Many of the species currently placed in this genus may not be congeneric (Tanasevitch [2022] lists only seven species in Lepthyphantes sensu stricto). One somewhat battered web of a male of the North American Lepthyphantes turbatrix (O. Pickard-Cambridge, 1877) was apparently naked, sparsely meshed, slightly domed, and attached at one edge by short anchors to a tree trunk (Fig. 46F). The web of L. minutus (in Denmark) was built against a tree trunk, with a gently sloping sheet (possibly bent centrally like a ridge tent) that abutted on one side against the trunk and a tall dense tangle above and little or no tangle below; the spider apparently rested off the web under flakes of bark or lichens (Nielsen, 1932: fig. 409).

Low Intrageneric Diversity. Orsonwelles (Webs of Eight Species). The webs of eight species of Orsonwelles (out of a total of 13 described species), a genus endemic to the Hawaiian Islands in which spiders rest at or beyond the edge of the sheet during the day, share a distinctive trait not known in any other linyphiid—a densely meshed “runway” with many parallel lines leading from an edge of the web to the spider's hiding place on the substrate (Fig. 62: O. polites; Fig. 61D: O. malus, lower left). Both the spiders and some of their webs are among the largest in the family Linyphiidae. Their webs were highly elevated, flat or slightly curved sheets built near large objects (except for Orsonwelles ambersonorum Hormiga, 2002) (Fig. 59A), with an upper tangle that varied in height and density and an absent or skimpy lower tangle that was often reduced to long dimple lines, as in O. falstaffius (Figs. 59D–G), Orsonwelles ambersonorum graphicus (Simon, 1900) (Figs. 60A–D), O. macbeth (Figs. 60E–G), Orsonwelles calx Hormiga, 2002 (Fig. 59B), and Orsonwelles othello Hormiga, 2002 (Figs. 61A, B). One slightly domed web of O. malus completely lacked a lower tangle (Fig. 61D), and another with a weak cup had a tangle below the lower portion of the cup (Fig. 61E). The lower tangles of the flat sheets of O. polites varied from absent (Fig. 62D) to sparse (Fig. 62A).

Figure 61.

Orsonwelles webs (3). (A, B) Orsonwelles othello Hormiga, 2002, female. Hawaii, Molokai, Kamakou Preserve (Orsonwelles othello_F_ow34plates.tiff, Orsonwelles othello_F_ow33plates.tiff). (C) Orsonwelles malus Hormiga, 2002, female. Hawaii, Kauai, Waialae cabin area (Orsonwelles malus_F_ow20plates.tif). (D) Orsonwelles malus, female. Hawaii, Kauai, Kokee State Park. Runway into retreat is in the bottom left corner (Orsonwelles malus_F_ow22plates.tiff). (E) Orsonwelles malus, juvenile. Hawaii, Kauai, Kokee State Park (Orsonwelles malus_J_ow21plates.tiff).

Juanfernandezia (Webs of Two Species)

The genus Juanfernandezia is endemic to the Juan Fernandez islands and includes two species, one of them undescribed (Arnedo and Hormiga, 2021). Three webs of J. melanocephala (Selkirk Island) (Figs. 29A–E) and five of Juanfernandezia sp. (Robinson Crusoe Island) (Figs. 30A–E) were very similar. All were densely meshed flat sheets close to but slightly above the forest floor, lacking tangles or long frame lines with only a few attachments to the substrate. Several sheets of both species had apparent repairs of holes (e.g., Figs. 29C–E; an especially large repair in Fig. 30B). One J. melanocephala sheet had a stem protruding through it (Fig. 29C).

Figure 30.

Juanfernandezia webs (2). (A–E) Juanfernandezia n. sp., female. Chile, Juan Fernandez Archipelago, Robinson Crusoe Island. (A) DSC_0004.NEF. (B) DSC_0423.NEF. (C) DSC_0430.NEF. (D) DSC_0011.NEF. (E) DSC_0331.NEF.

Pimoidae

Moderate Intrageneric Diversity. Pimoa (Webs of Two Species). The genus Pimoa groups about 80 described species from eastern North America, southern Europe, and Asia. The photographed webs of both of the North American Pimoa species were elevated. The three webs of P. cthulhu (Figs. 64A–D), from California, differed from all known linyphiid webs. They were built against vertical walls in large cavities (e.g., at the base of a tree trunk) and had an extensive planar vertical sheet parallel to this surface that seemed, at least in one web (Figs. 64A, B), to be a continuation of the top of a vaguely saddle-shaped sheet. In at least one (Fig. 64C) and perhaps another (Figs. 64A, B) was a tubular retreat near the top. Two webs of the American Pimoa breviata Chamberlin & Ivie, 1943 (Figs. 64E, F), varied substantially. One was a sparsely meshed, possibly weakly dome-shaped sheet with little or no tangle above or below and was also built against a vertical tree trunk. The spider rested near the trunk with no perceptible retreat in one case (Fig. 64F); the second web (Fig. 64E), also against a tree trunk and with little or no tangle, had a more extensive sheet with a denser mesh and a retreat under the tree bark; other webs of P. breviata in the same area also had retreats. Hormiga (1994a: figs. 5, 6) depicts an additional web of P. breviata, a horizontal sheet with a relatively open mesh. European species of Pimoa usually inhabit the outermost section of caves where they can be found on cave walls or among large rocks on the ground, hanging on horizontal sheet webs (Isaia et al. 2011; Mammola et al., 2016).

Summary of Intrageneric Variation: Rampant Divergence

In sum, the data from the present study and from previous publications present a clear pattern: there is frequent substantial variation among the webs of congeneric (i.e., closely related) species. This pattern is clear even though present coverage is only fragmentary (two-thirds of the 21 genera we reviewed above were represented by only two to three species), and incomplete coverage undoubtedly results in a bias to underestimate the diversity of web forms.

Several additional factors also indicated that current knowledge seriously underestimates the diversity of linyphiid webs. In the first place, the taxonomic coverage in the family is only fragmentary. Counting the >60 species represented here, combined with previously published web photos from just under 30 other species, web forms have been documented in just under 2% of the known linyphioid species (and this is an overestimate because it ignores the many species not yet described). At the level of genera, the present paper adds information on 26 genera; combined with the published web photos for approximately 12 other genera; this represents a total of only about 6% of the 625 genera in Linyphiidae and the two genera in Pimoidae. This lack of taxonomic coverage, in combination with the high diversity of web forms even within a single genus and the consistently overly typological nature of current knowledge of linyphiid web designs, suggests that a great deal of diversity is yet to be discovered.

It seems likely, nevertheless, that there are phylogenetic signals in the intrageneric differences. For example, the genus Neriene, for which we have web information on the greatest number of species (17 of 60 described species), there seems to be one group with elevated, deep, cup-like sheets; another group with elevated, pronounced, dome-shaped sheets; and one species with a flat web near the ground. It remains to be determined, of course, whether these represent natural groups of species.

There are several suggestions for possible convergences in different genera. The vertical trunk sheets in the erigonine Laminacauda rubens were similar to those of the distantly related “micronetine” Obscuriphantes obscurus (Blackwall, 1841) which built dense vertical sheets against tree trunks (Nielsen, 1932: fig. 410). The O. obscurus webs differed, however: the upper edge curved toward the trunk where it was broadly attached; long, straight lines anchored the sheet to branches above and below; and the spider rested facing downward on the trunk and holding the sheet with its hind tarsi (the spider dropped readily when disturbed) (Nielsen, 1932). Vertical sheets are also built against tree trunks by Drapetisca socialis but are relatively much smaller and probably function mostly as trip lines to alert the spider to the presence of prey (Schütt, 1995). The asymmetrical, vertical sheets formed by the upward extension of one upper wall of a cup in Neriene GH03 (Figs. 52A, B) resembled the downward extension of the lower wall of a dome in the distantly related Putaoa seediq (Figs. 75C–F). The slime on sheets of Tapinopa spp. resembled that on Laminacauda magna webs. (However, although slime seems to be a typical feature in the former species, it is only occasionally found in the latter.) The pattern of intrageneric diversity we have found in linyphioids resembles that in the webs of another large araneoid family, Theridiidae (Eberhard et al., 2008).

Our sample of Pimoidae is too small to justify strong conclusions, but there are hints of the same combination of intrageneric diversity and convergence in Pimoa; P. cthulhu also had a downward extension of the lower wall of a dome (Fig. 64C). No webs are known in Nanoa enana (Hormiga et al., 2005), the single species of the only other pimoid genus.

The fragmentary nature of the data on webs, along with the lack of a well-supported phylogeny of the family Linyphiidae (but see Fig. 2 and Arnedo et al., 2009; Wang et al., 2015; Silva-Moreira et al., in prep.), have discouraged us from attempting to trace web evolution within Linyphiidae on a phylogenetic tree. A well-supported outline of their phylogeny will be crucial in classifying linyphioid webs and grouping them in biologically realistic ways and to trace the evolutionary changes in web architecture.

Intraspecific Variation

Web traits showed substantial intraspecific variation (with coefficients of variation ranging up to >90%) in two linyphiids, Diplothyron simplicatus and Neriene coosa (Eberhard, 2022). The variables included the symmetry and curvature of the dome-shaped sheet, the sheet's orientation with respect to horizontal, and relative sizes of the tangles compared with the diameter of the sheet. Suter (1984) also found substantial variation in the curvatures of the cups of Frontinella pyramitela. Alderweireldt (1994) linked substantial variation in the size of Bathyphantes gracilis (Blackwall, 1841) webs to the sizes of available spaces in which to build in cultivated fields. Despite the relatively small samples in the present study, several examples were of significant intraspecific variation.

Two Pityohyphantes costatus webs were nearly naked, open-meshed sheets that sloped up to a sheltering object on one side where the spider rested (Figs. 65A–C); in contrast, a third web had a weakly dome-shaped sheet with no sheltering object and a medium-dense upper tangle (Figs. 65F, G). One of these webs had numerous V-shaped lines at one edge of the sheet (Figs. 65A, C), but another web lacked them (Fig. 65F). In one web of Mecynidis sp. the lower tangle formed a clear sheet (Figs. 47A–C), but in another the principal sheet was just above a large leaf, with only a few lines and no sign of a sheet in the lower tangle (Figs. 47D, E). In Novafrontina uncata the relative height of the upper tangle varied from about 0.5 to three sheet diameters (Figs. 58A, H). The form of the sheet of Putaoa seediq varied from sharply dome-shaped or conical (Figs. 75C, D) to quite flat (Fig. 75A). Laminacauda magna may have varied in the presence or absence of a second “sandwich” sheet (Fig. 41D). Even a single web sometimes provided variations, as in the Walckenaeria sp. web with long frames with many “V” attachments of sheet lines along with one elevated edge that lacked clear frame lines and had long anchor lines instead (Fig. 77E). Undoubtedly much further intraspecific variation remains to be documented.

In sum, it is probable that linyphiid webs usually show substantial intraspecific variation in a number of traits, which implies that the small samples of webs for the species in this study (as well as the web descriptions throughout the published literature) likely underestimate of the diversity of webs built by any single species. This observation, in turn, implies that the high degree of intrageneric diversity described above is probably a substantial underestimation.

Some of this intraspecific variation probably represents adjustments to the environment rather than imprecision. For instance, the N. uncata web with an especially low upper tangle (compare Figs. 58G, H with 58A–E) was built at a site where there were no potential attachment sites higher above the sheet. Thus, some differences in web design probably resulted from variation in web site choices by the spiders. Nothing is known concerning how linyphioid spiders explore and choose web sites. Benjamin and Zschokke (2004: 122) mentioned that Linyphia hortensis, which built naked sheets, performed a “short exploratory stage with the spider moving in a space where the web was later built,” but gave no further details. Repairs and additions of lines to webs on subsequent nights, which have been documented in Drapetisca socialis (Schütt, 1995), Linyphia triangularis (Benjamin et al., 2002), and Diplothyron simplicatus and Neriene coosa (Eberhard, 2022), probably add further intraspecific variation. The almost complete absence of webs in this study that had stems and leaves projecting through them (the only clear exceptions are Neriene coosa [Eberhard, 2022] and Linyphia hortensis [Nielsen, 1932]) implies that exploratory behavior by linyphioids before web construction must inform the spider regarding the sizes of open spaces in which to build.

The Evolutionary Derivations and Convergences Involving Linyphioid Webs

Recent phylogenomic analyses (Kulkarni et al., 2021: fig. 4) suggest that Linyphiidae and Pimoidae form a clade with Cyatholipidae (but see Kallal et al., 2021). Few studies have documented cyatholipid webs, but all available data and our own observations indicate that they build horizontal sheet webs. Griswold (1987: 500) mentions that cyatholipids build “small horizontal sheet webs”, and Davies (1978: 293) described the web of Teemenaarus as a “filmy sheet web against the bark of forest trees.” Our own observations of cyatholipids webs in the genera Alaranea, Forstera, Matilda, Teemenaarus, Tekella, Tekelloides, and Wanzia agree (Figs. 3–7) that they all built densely meshed, naked or nearly naked horizontal sheets in forests, usually above the ground on vegetation and against tree trunks. These genera have triads on their posterior lateral spinnerets, implying that at least some lines in their webs are sticky. They also have reduced numbers of aciniform spigots (Griswold, 2001), implying that they do not attack prey by wrapping them. The major difference of the webs of these small to very small spiders compared with linyphioids is their combination of a lack of tangles and elevated web sites. They share with linyphiids the relatively long and thin legs (Griswold, 2001; Jocqué and Dippenaar-Schoeman, 2007) but have only short medium paturons.

The preliminary suggestion is thus that the most recent common ancestor of linyphiids, pimoids, and cyatholipids built sheet webs. These three families may be closely related to physoglenids, which also build elevated horizontal sheets, with more or less sparse tangles (Figs. 8–10); at least some also have triads (G. Hormiga, unpublished) (also implying sticky silk in their webs) and relatively long, thin legs (Forster et al. 1990), and at least some have reduced aciniform spigots (data are incomplete), implying that they do not attack prey by wrapping them.

Farther back in time, linyphioids, cyatholipids, and physoglenids are thought to be derived from an orb-weaving ancestor (Kulkarni et al., 2021). The typical predatory strategy of orb weavers differs in basic ways from that of these sheet weavers: orb weavers employ relatively low numbers of strong lines that bear relatively large amounts of adhesive and that are placed in precise locations with respect to one another; in contrast, the linyphiids (and probably also their allies) rely on much larger numbers of relatively fine lines that bear much smaller droplets of adhesive and that are placed in much less consistent patterns with respect to each other.

Additionally, linyphiids differ from many orb weavers in making extremely rapid attacks on prey. Some orb weavers, such as the tetragnathid Leucauge mariana (Taczanowski, 1881) (Briceño and Eberhard, 2011) and the araneid Trichonephila clavipes (Linnaeus, 1767) (Robinson and Robinson, 1973; Eberhard, 2020), also attack prey rapidly, but many other orb-weaving spiders have relatively shorter legs and are much slower (summary in Eberhard, 2020). The typical linyphiid body plan noted above, with relatively long, thin legs, operating in conjunction with their nearly always unobstructed sheets that have an open space below that facilitates rapid movement and robust chelicerae and reduced aciniform spigots associated with rapid biting attacks, suggests specialization for speedy attacks on prey. The linyphiids' legs contrast with the short heavy legs of some slowly attacking orb weavers, such as the araneids Micrathena and Gasteracantha and theridiosomatids (Eberhard, 2020), and the thick, powerful legs of many webless spiders (Wolff et al., 2022).

One consequence of rapid linyphiid attacks is that the prolonged retention of prey in their webs is probably less crucial for prey capture than it is in orbs: probably only a short retention time (perhaps generally <1 second) is necessary for linyphiids to capture many of their prey. Nielsen (1932) observed that the webs of Neriene montana retained flies for only very short periods, and Turnbull (1960) mentions prey that were not attacked readily escaped from webs of this same species. The small droplets of adhesive on linyphiid web lines may be sufficient to achieve momentary retention (Eberhard, 2021). In general, linyphioids may represent an additional example of the pattern noted by Kaston (1964) for web-building species to evolve to rely less on their webs and more on active attack behavior.

We must reemphasize the current lack of data on the presence and distribution of adhesive material in the web, which undoubtedly affect prey retention times. The retention capabilities of the different linyphioid webs probably depend on the presence of sticky lines, the sizes and spacing of the droplets of adhesive, and the densities of sticky (and nonsticky) lines in the sheet and in the upper tangle just above the sheet. Preliminary data from other studies indicate that the fraction of lines in the sheet that bear droplets differs between species; for example, webs of Microlinyphia pusilla had more sticky globules than those of Linyphia triangularis (Benjamin et al., 2002). Neriene radiata lacks triads (there are no aggregate gland spigots accompanying the flagelliform spigot) (Schütt, 1995), but the congeneric N. coosa had abundant droplets on nearly all of the lines in the sheet (Eberhard, 2021). The numbers of aggregate gland spigots in Linyphia hortensis also apparently varied (“in some specimens of L. hortensis one of the spigots of the aggregate gland may still exist”; Schütt, 1995: 559). In at least several species of the erigonine genus Callitrichia the triad on the posterior lateral spinneret is absent in the adults of both sexes (Lin et al., 2022), but nothing is known of their webs. Further observations of linyphioid lines and the droplets on them under the compound microscope may yield important new insights.

The “elevated sheet with tangles above and below” design that is common in linyphioids has evolved convergently in several other families of spiders, including the theridiids Nihonhimea tesselata (Keyserling, 1884), Cryptachaea sp., Tidarren spp., and Chrosiothes sp. (Eberhard et al., 2008; Madrigal-Brenes and Barrantes, 2009); the araneid Cyrtophora and allied genera (Blanke, 1972; Eberhard, 2020); and the diguetid Diguetia spp. (Eberhard, 1967, 2020; Bentzien, 1973). These webs share the linyphioid trend for a lower tangle less dense than the upper tangle and an open space just below the sheet. Additionally, distant retreats with signal lines that lead to the prey capture webs are widespread in groups with relatively sparse sheets such as the theridiid Latrodectus, as well as in many orb weavers (Eberhard, 2020), and are generally lacking in linyphioids. The runways of Orsonwelles spp. and the lines that converge at one corner in Agyneta spp. that were documented here and that also occur in Pityohyphantes costatus (= Linyphia phrygiana) (Emerton, 1902) seem to be as close as linyphiids come to signal lines running to distant retreats; perhaps this lack is related to linyphiid reliance on especially rapid attacks on prey.

Another striking difference between linyphioids and these other families is that linyphioids almost never incorporate objects such as leaves or other detritus into their webs to use them as hiding places (Eberhard et al., 2008; Eberhard, 2020). One web of Laetesia raveni apparently had the top of the dome just under leaves (Fig. 33B) that may have provided protection against predators, but the top of the dome was more typical in other webs of the same species (Fig. 32) and of two other Laetesia species (GH01 and GH02) (Figs. 33C–F, 34D, E, 35A–E), far from any protecting plant structure. The sheet webs of the African genus Mecynidis are typically built under a leaf (Scharff, 1990: fig. 138; Fig. 47A). Webs in one other family, Pholcidae, resembles linyphioids in also building elevated dome-shaped sheets and not using detritus as hiding places. (Their long legs would make hiding this way more difficult.) Pholcids may differ, however, in having only relatively reduced upper tangles, in lacking lower tangles entirely, and in often having larger droplets of adhesive (at least among the few species whose webs are known) (Deeleman-Reinhold, 1986; Eberhard, 1992, 2020; Japyassú and Macagnan, 2004).

Possible Functions

Different portions of linyphiid webs have numerous possible nonexclusive hypothetical functions. For instance, sheets can intercept prey, stop prey, retain prey, protect the spider from prey, transmit vibrations that allow the spider to locate prey, maintain a favorable microclimate for the spider (Toft 1980), and serve as vehicles for pheromones (Nielsen, 1932; Toft, 1980; Schulz and Toft, 1993; Gaskett, 2007). Upper tangles can intercept, stop, and retain prey; provide anchors for lines that tense the sheet and allow it to curve; and protect the spider from predators. The lower tangles can also protect spiders from predators and provide anchors for tensing the sheet and allowing it to curve. Given that some possible prey appear to sense webs visually and then avoid them (Turnbull, 1960), lower visibility, at least of the sheet and upper tangle, may improve prey capture. Experimental tests of the function of different linyphioid web design details are nearly nonexistent, so we cannot address many of these ideas. Nevertheless, a few aspects of the forms and distributions of web components provide tests of hypotheses regarding functions. Functional considerations can help illuminate significance of the diversity of web forms that we have documented.

Space Below the Sheet for the Spider to Run. This study confirms and extends the observations of Nielsen (1932) that cup-shaped and flat sheets consistently had open spaces just below them. No linyphiid web is known in which there are dense tangle lines immediately below the sheet. As argued by Nielsen, the function of this space is probably to allow the spider to move rapidly under its sheet to attack prey and escape from predators. As far as is known, all linyphioid spiders always move on the undersurfaces of their sheets rather than on top of them, as in Linyphia triangularis (Figs. 46A, B), Neriene albolimbata (Figs. 50C, D), and Pocobletus sp. Panama (Figs. 69B, D) (see also Bristowe, 1930; Nielsen, 1932; Suter, 1984; Eberhard, 2022, on Diplothyron simplicatus). We have never observed a linyphiid consistently running on the upper surface of a sheet.

In contrast, no similar open spaces were seen between the upper surface of the sheet and the upper tangle. It appeared that the density of the upper tangle tended to be greatest near the upper surface of the sheet and to decrease gradually farther up. Examples include Frontinella pyramitela (Figs. 27C, D), Pocobletus sp. GH33 (Figs. 68B–D), and Pocobletus sp. GH28 (Fig. 70D). Possible exceptions were Novafrontina uncata (Fig. 59A), Neriene digna (Fig. 50G), Acroterius GH02 (Fig. 78), and Neriene variabilis (Fig. 53F). Greater tangle density just above the upper surface of the sheet may represent an adaptation for retaining prey that have fallen onto the sheet's surface in the cyrtophorine araneid Cyrtophora citricola (Forsskål, 1775), another group that builds a horizontal sheet with an upper tangle that includes many relatively lax lines that may hinder prey movement across the sheet (Eberhard, 2020).

Defense Against Predators. Blackledge et al. (2003) argued that tangle webs have a defensive function. That study, however, focused only on sphecid wasps as predators and was based on the questionable assumption that faunal surveys employing collecting techniques such as beating, sweeping, pitfall trapping, and canopy fogging provide accurate reflections of the relative numbers of prey that are available to these wasps. Additionally, most linyphiids are too small to serve as food for many sphecid wasps (Bristowe, 1941). The study also lumped the many linyphiids that do not build a tangle or a sheet below the main sheet with those that do, so its relevance here is uncertain.

Some data from the present study suggest a defensive function for the lower tangle. That lower tangles were nearly always smaller than the main cup-shaped sheet and were often located just under the bottom of the cup where the spider probably rested in Laminacauda sp. (Figs. 43D, E), Pocobletus sp. GH32 (Fig. 66C), Pocobletus sp. exeGH05 (Fig. 72E) supports the hypothesis that the lower tangle has a defensive function. In our experience, spiders with domed or cupped sheets tend to rest under the peak or under the bottom of the cup (further evidence comes from explicit statements to this effect by Emerton (1902) for Neriene radiata (= Linyphia marginata); by Suter (1984) for Frontinella pyramitela; and by Eberhard (2022) for Diplothyron simplicatus and Neriene coosa, as well as from photos of Shinkai (1979) and Shinkai and Takano (1984) for Turinyphia yunohamensis (Bösenberg & Strand, 1906) and Neriene emphana and from our photos of Neriene albolimbata (Figs. 50C, D), Linyphia triangularis (Figs. 46A, B), and Pocobletus sp. Panama (Fig. 69B). The central positions of these lower tangles thus suggest that their most probable function is to form a physical barrier against enemies that attack the spider from below. In addition to being a physical barrier, the lower tangle may also serve as an early warning device to elicit defensive behavior. Florinda coccinea, which had a moderately sparse lower tangle that was not concentrated below the center of the sheet, dropped quickly to the substrate below when startled (W. Eberhard, unpublished; G. Hormiga, unpublished), and Microlinyphia mandibulata (Emerton, 1882), which apparently lacked a tangle below, also dropped when startled (Emerton, 1902). Behavioral evidence in other families also indicates that a tangle below the spider's resting site (the hub of a horizonal orb) functions to defend uloborid spiders from predators (Lubin et al., 1982; Lubin, 1986; summary in Eberhard, 2020).

Nevertheless, the defensive function hypothesis seems unlikely to explain webs in which the lower tangle consisted of only a few lax lines, some of which were far from the spider's presumed resting site, as in Pimoa breviata (Fig. 64E) and Laetesia raveni (Fig 34B). (Unless, as seems unlikely, these species do not suffer significant predation.)

Upper tangles may also at least occasionally function in protection from predators. In accord with this idea, the upper tangle of Agyneta sp. CR01 (Fig. 11D) was more or less limited to just above the site where the spider rested near the edge of the sheet. Both the lower and (especially) the upper tangle lines of Neriene coosa webs were generally more concentrated just above (as well as just below) the spider's resting place; peripheral portions of the large sheets of this species often had no tangle lines above them (Eberhard, 2022).

Prey Capture I: Sensing Prey. The webs of Erigone dentigera (Emerton, 1902) and (probably) an unidentified Japanese erigonine (Shinkai, 1979) and of the erigonine Mermessus tridentatus (Fig. 47F) were attached at many points to the upper tips of moss plants and other objects near the surface of the ground and lacked clear frame lines. This design may represent an adaptation to enable the spider both to sense and to gain rapid access to pedestrian prey like collembolans walking on the substrate at the edge of the web. This interpretation is supported because some erigonines with webs captured pedestrian prey, and adults of Hypselistes florens (O. Pickard-Cambridge, 1896) and multiple species of Erigone and Oedothorax sometimes captured prey in the field without building any web (Wheeler, 1973; Alderweireldt, 1994).

Prey Capture II: Intercepting and Stopping Prey. Tangles above the sheets of linyphioids and other spiders that build sheets, including theridiids, araneids, diguetids, diplurids, agelenids, and lycosids (Eberhard, 2020) are generally considered to be devices for intercepting and stopping prey in the air above the sheet, knocking them down onto the sheet, and then retaining them on the sheet, allowing the spider to attack (Foelix, 2011). For most groups these hypotheses are not supported by empirical studies, but experimental elimination of tangles above the sheet showed that the tangles both increased prey capture and reduced predation in the araneid Cyrtophora moluccensis (Doleschall, 1857) (Blamires et al., 2013). The knock-down hypothesis can explain why upper tangles in our study consistently failed to extend beyond the edges of the sheet, although minor exceptions occurred in Novafrontina sp. (Fig. 58A) and Diplothyron simplicatus Eberhard (2022). Tangles were nearly as wide as the central sheet, including dome-shaped sheets, as in Frontinella pyramitela (Figs. 27C, D); slightly cup-shaped sheets, as in Pocobletus sp. exeGH03 (Fig. 72D); and flat or nearly flat sheets, as in Pocobletus sp. GH28 (Fig. 71D), Pocobletus sp. GH11 (Figs. 70A, B), Mecynidis sp. (Figs. 47A–C), and Tapinopa vara (Fig. 76B).

In many (although not all) of the webs with substantial upper tangles, the upper tangle diameter decreased the farther above the sheet, and more peripheral portions of the sheet often had fewer knock-down lines above them, as in Pocobletus sp. GH28 (Fig. 71B), Dubiaranea sp. DE2 (Fig. 23G), Frontinella pyramitela (Fig. 27C), and Neriene helsdingeni (Fig. 55D). A few exceptions included Pocobletus sp. GH33 (Fig. 67C). If prey tend to fall directly downward when they strike the upper tangle (observations are lacking), then the combination of this pattern in the upper tangle and the lower densities of lines near the edges of sheets (Eberhard, 2022) implies that peripheral portions of the web may be less able to stop and retain prey. The functional significance of looser mesh near the web's periphery is not clear and deserves further study.

Prey Capture III: Retaining Prey until the Spider Arrives. The widespread existence of sticky lines in sheets indicates that sheets function in retaining prey. In general, the multiple lines attaching the upper tangle to the sheet produced only tiny or imperceptible dimples in the sheet, implying that these lines were under relatively low tensions compared with those on lines in the sheet. Lax lines just above the sheet, as in Linyphia triangularis (Fig. 46E) and Dubiaranea distincta (Figs. 21C, D), may be specialized for entangling prey, and thus increasing the time they are retained after falling onto the sheet (similar, apparently especially lax lines just above the sheet occur in the araneid Cyrtophora citricola (Eberhard 2020) and the theridiid Nihonhimea tesselata (Jörger and Eberhard, 2006).

Collembola as Prey. There are several indications that collembolans (wingless insects that can hop but not fly, and which thus generally have aerial trajectories that are largely up and down) may be important prey for those linyphioid species that build webs on or very close to the substrate. Bristowe (1941: 292) reported that collembolans were the principal prey of some linyphiids in the UK. In the field, Bathyphantes eumenis, which builds webs in irregularities in damp rocks, apparently specialized on collembolans (87% of 165 prey in the field belonged to eight species of collembola) (Rybak, 2007), and collembola were the main prey of Erigone arctica (White, 1852) (a species with small webs on the substrate; Nielsen, 1932) (Van Wingerden, 1978). Finally, the majority (60%) of the potential prey captured by sticky traps placed in habitats where the webs of at least 11 linyphiid species (including adults and juveniles) were built were collembolans (mostly isotomids and entomobryids, both of which can jump) (Harwood et al., 2001).

In contrast, the linyphiid Linyphia triangularis, which builds elevated webs, captured quite different prey. In webs in bushes in a “wet meadow” 90% of the prey from the webs of (apparently adult females) were Aphidoidea, Cicadellidae (= “Cicadina”?) and Nematocera (Nentwig, 1983). In webs in shrubby vegetation of the field layer, 91.6% of 581 prey found in webs of different stages of this species were flying insects, and many were hymenopterans; 7.2% were jumping species (mainly leafhoppers and froghoppers), and only 1.2% were pedestrian (Turnbull, 1960). Weakly flying species (nematocerous flies, parasitic wasps, aphids) were most common, especially early in the season in the webs of juveniles. Microlinyphia mandibulata fed largely on leafhoppers (which both jump and fly) and small dipterans in alfalfa fields (Wheeler, 1973), while both Bathyphantes gracilis and Neriene clathrata in cultivated fields fed mostly (91% and 66%, respectively) on aphids and collembolans (Alderweireldt, 1994).

Water Droplets as Traps? Some sheet-building species, such as Agyneta sp. CR01 (Fig. 11E), Agyneta sp. (Fig. 13A), and Tapinopa longidens (Fig. 76A), consistently build webs extremely close to damp substrates such as wet moss or earth. The cooler air at such a site holds less water than warm air, and water condenses from oversaturated air on anything that can provide a nucleus and then breaks into droplets because of Raleigh instability. Spider webs, such as those of some erigonines that are built in the very humid layer close to a damp substrate, are often covered with tiny water droplets, as in Erigone dentigera (Emerton, 1902) and E. atra (Nielsen, 1932). Such droplets may function in prey capture in the webs of a hahniid spider: prey that struggled to free themselves on droplet-covered sheets sometimes accumulated a larger drop of water on their bodies that appeared to impede their escape (Eberhard, 2019). It remains to be determined whether by building in such habitats linyphiids provide their webs with water droplets that facilitate prey capture.

Forms of the Sheet—Cups, Domes, Dimples, etc. The functional significance of the domed, cup-shaped, and flat forms of sheets is unclear. One might suppose that the upward-slanting curves of cup-shaped sheets function to make it more difficult for prey to walk off the sheet, but this idea runs counter to the more widespread trend toward domed sheets, a design that would by this reasoning promote prey escape. Domed sheets are widespread in pholcids, and also occur in some theridiids (Madrigal-Brenes and Barrantes, 2009; Eberhard, 2020). Domed sheets were always accompanied by a tangle above the sheet, and cup-shaped sheets by a tangle below, but these correlations are probably because of construction constraints: the lines in a sheet are not stiff, so additional lines out of the plane of the sheet are needed to hold the sheet in a curved configuration. Nevertheless, the relatively small tangles associated with strongly domed pholcid sheets show that relatively small tangles can be sufficient to sustain sharply domed sheets (see Eberhard, 2020, on Modisimus). The runways from apparent resting places near the substrate to the sheet, as in Himalaphantes sp. (Figs. 28B, D, E), presumably function to allow the spider to rest in a less exposed position and reduce the danger from predators but nevertheless allow rapid access to the sheet for prey capture.

The sites of dimples in sheets suggest their possible functions. Dimples tended to occur at places where the sheet was curved, with the dimple projecting in the opposite direction from the sheet's curve. For instance, relatively uniformly spaced downward dimples occurred in the uniformly curved cup of one web of Acroterius sp. GH02 (Fig. 78) but were more concentrated near the centers of other, less sharply cup-shaped sheets of the same species (Figs. 79D, F). The downward-directed dimples in the weakly cup-shaped sheet of Pocobletus sp. GH33, in which the curve was mostly limited to the central area, were also limited to the central portion of the sheet (Figs. 67C–F). The dimples were dispersed over nearly the entire sheet in the more uniform, moderately curved cup-shaped sheets of Frontinella pyramitela (Figs. 27C, D) (as also reported by Suter, 1984) and of Frontinella sp. (Figs. 27A, B). Similarly, upward dimples occurred near the tops of domes. The small upward dimples in Diplothyron dianae (Fig. 20D) and Grammonota sp. (Fig. 28A) were at places where an otherwise flat or slightly domed sheet curved downward.

Similar downward-directed dimples occurred where sombrero-shaped sheets curved upward in several other families, including Diguetidae, Araneidae, and Theridiidae (Eberhard et al., 2008; Eberhard, 2020).

The “tensor” lines perpendicular to the sheet that are associated with dimples probably alter the form of the sheet by pulling it downward. In accord with this idea, cutting some tensor lines in Frontinella pyramitela webs caused sheets to become flatter (Suter 1984). These lines may also function to tense the sheet itself (Suter 1984). This duality of function is suggested by the weak dimples in the strongly curved (but perhaps not strongly tensed) sheet of Australian genus 3 GH02 (Fig. 18F). Whatever their function, it seems likely that the tensor lines below cup-shaped sheets must constitute physical obstacles that reduce the speed with which spiders can attack prey; their function must be important enough to compensate for this hinderance to the spider's attacks. It appears, in contrast, that the striking absence of strong upward-directed dimples with tensor lines in dome-shaped sheets, even sharply dome-shaped sheets such as Neriene litigiosa (Fig. 56), is associated with many different lines that attach the sheet to the tangle at different sites, instead of just a few tensor lines. This difference may be associated with lack of selection on lines just above the sheet to facilitate rapid attacks.

Mechanical Stability. Peters and Kovoor (1991) stated that the only function of tangles is to support the sheet. The numerous sheet webs that lacked or nearly lacked upper or lower tangles argue against this idea. Nevertheless, a sheet obviously needs some lines to provide mechanical support, especially to resist stresses such as wind and falling detritus. This function cannot be examined for lack of data. An incidental observation was that many linyphiid webs often survive rain storms more or less intact. The sheets of some Diplothyron simplicatus webs, which have extensive tangles above and below, had repaired sectors of the sheet (Eberhard, 2022). Even species with naked or nearly naked sheets sometimes had repairs of damage to their sheets, as in Floronia bucculenta (Figs. 26D, E) and Tenuiphantes flavipes (Figs. 77A–C), suggesting that these webs may have relatively long functional lives in nature.

The Mystery of Naked Sheets. The frequent absence of upper tangles in some linyphioids (11 species in perhaps 10 different genera had “naked” or “nearly naked” sheets), and the sparse lines above the sheets of others such as the undescribed species Australian genus 3 (Fig. 18E) (which nevertheless had an extensive tangle below), as well as the naked sheets typical of Cyatholipidae (Figs. 3–7), constitutes a puzzle: it is not clear how prey are captured by horizontal sheet webs that lack tangles. One unavoidable conclusion seems to be that many prey in the field must be moving through the air in vertical rather than horizontal directions. Perhaps, for example, the sheets of webs that are located just above but very close to the substrate (and in which tangles were usually reduced or absent) function to capture jumping collembolans.

It is striking that almost all the webs in this study that were found near the ground lacked a tall or thick upper tangle (the exception is one web of Novafrontina uncata on the forest floor; Figs. 58C, D). The existence of a few, sparse lines above many of these nearly naked sheets showed that attachment points were nevertheless available, eliminating the possibility that the spiders could not build lines above their sheets. The implication is that these webs were capturing prey that were moving vertically in the air very near the ground. Hahniids also build approximately horizontal sheets very near the ground without tangles above them (Opell and Beatty, 1976; Eberhard, 2019)

This trend in linyphiid webs near the substrate emphasizes the even more general mystery of how prey are captured by the many sheet webs that lack upper tangles, both those built near the substrate and far above it. The low numbers of attachments of these sheets to the substrate suggest that they capture few pedestrian prey; they would seem to be limited to capturing jumping or falling prey (see also Harwood et al., 2001). At least for naked and nearly naked sheets near the ground, perhaps collembolans are the major prey (above). Another possibility, mentioned in connection with symphytognathoid webs near the ground, that they are designed to capture spores or pollen (Blackledge et al., 2011), is contradicted by the lack of any tendency in the spiders in any of these groups to build their sheets just below fungal fruiting structures, as occurs in the specialized spore-feeding mycetophilid fly larvae (Eberhard, 1970).

The small, flat, horizontal, naked sheets of “juvenile” webs of Bathyphantes eumenis are built in irregularities in damp rocks. Juvenile webs accounted for most prey captures that occurred in captivity (Rybak, 2007), and in the field this species is apparently a specialist on collembolans (87% of 165 prey in the field belonged to eight species of collembola; the different web types were not distinguished in the field) (Rybak, 2007). It appears that these webs have low numbers of attachments to the substrate (only drawings are available), so the alternative possibility that these webs function as sensors for pedestrian prey, as proposed above for erigonine webs, is unlikely.

Capture of collembolan prey does not explain, however, elevated sheets that also lack tangles above them, such as Australian genus 3 GH02 (Fig. 18E). Similar designs also occur in other spider families, including cyatholipids (Figs. 3–7), oxyopids (Coville and Coville, 1980; Griswold, 1983), psechrids (Robinson and Lubin, 1979; Eberhard, 1987), and synotaxids (Eberhard, 2020). The upper tangles of many pholcids are also very low and seem poorly designed for intercepting prey; some (e.g., Modisimus bribri) are built near the undersides of leaves (Eberhard, 2020). Studies of prey captured in the field could help solve some of these mysteries.

In a striking contrast, the symphytognathoid orb weaver families Anapidae and Symphytognathidae that build orbs in leaf litter close to the ground have evolved to do just the opposite—to add a tangle with both sticky and nonsticky lines above their horizontal webs (Lopardo et al., 2011; Eberhard, 2020).

Summary. The various probable functions of different design details of linyphioid webs described above offer a starting point for understanding the significance of the diversity of web forms documented in this study. Different web designs presumably represent specializations to improve particular functional traits. Although there have been almost no experimental studies of function or of how prey capture occurs that have enough detail to begin to reveal the functions of different design features of linyphiid webs, our overview of web diversity is sufficient to begin a tentative discussion of rapid runs to attack prey, defense against predators, prey interception, prey retention times, collembola in webs near the ground, dimples in sheets and their associated tensor lines, and mechanical stability. Many mysteries remain. Perhaps the greatest is the functional significance of building sheets that are either domed, cup-shaped, or flat—a mystery not only in linyphioids, but also in other groups that build aerial sheets, such as pholcids, theridiids, araneids, and diguetids (Eberhard, 2020).

The Linyphioid Syndrome

As noted in the Introduction, linyphioids comprise a large number of species but are relatively uniform rather than diverse in many somatic morphological traits. This study, in combination with previous publications, has documented a similar pattern of relative uniformity in building a horizontal, unobstructed sheet of some kind, with an open space just under the sheet. We have also documented a substantial amount of diversity in the details of this basic web design that includes the sizes, densities, and locations of tangles above and below the sheet; cupped compared with domed compared with flat sheets; sheets with and without dimples; and sheets with and without retreats at one edge. In general, the appropriate phrase for linyphioid webs is “moderate diversity.” Thus, the linyphiid success story of many species that are often very abundant is perhaps characterized not by evolutionary inventiveness, but by relative conservatism in a suite of traits.

We propose that a subset of this suite of conserved linyphiid traits, including the designs of their webs and the spiders' morphology and attack behavior, all contribute to increasing the speed with which linyphiids attack their prey. Linyphioid webs appear to have limited retention capabilities, with lines that bear only tiny droplets of adhesive. (The droplets may self-assemble at or near the point of impact and increase adhesion to prey, at least slightly [Eberhard, 2021].) This limited retention capability in webs is combined with very rapid attacks on prey. The horizontal sheet provides a direct path to prey on any portion of the sheet; the smooth surface of the sheet, the almost universal lack of any objects projecting through it, and the open space just below it except the reduced numbers of lines that connect the sheet to the tangle below where the sheet curves upward, all facilitate rapid dashes. The typical spider somatic morphology, in particular the moderately long, moderately thin legs (in contrast, e.g., with many “stubby-legged” body designs in related groups such as orb-weaving araneoids), also facilitate rapid running. Also contributing to the rapid attacks are immediate biting attacks on prey (rather than wrapping it) used by linyphiids to restrain the prey as soon as the spider arrives (Eberhard, 1967). Biting and then holding on to the struggling prey is facilitated by the robust chelicerae, and perhaps also by the teeth on both margins of the paturon (the basal cheliceral segment) (Draney and Buckle, 2017). Additionally, these biting attacks may explain the reduction in the aciniform gland spigots that provide wrapping silk; the widespread nature of this reduction in linyphiids implies that biting attacks occur in many species whose behavior has yet to be observed. Other web traits also contribute to rapid attacks: the sheet is thin, with most or all lines in nearly the same plane, so the spider can bite the prey through the sheet and grasp it without having to pause to make a hole. Additionally, the sheet is between the spider and the prey and thus protects the spider from damage from the prey's struggles when the spider contacts the prey while biting and holding it. Still another aspect of typical linyphiid behavior that may also facilitate rapid attacks occurs before the arrival of prey: the spider commonly positions itself near the center of the sheet (Bristowe, 1941; Benjamin and Zschokke, 2002) (although with exceptions that hide in retreats off the center during the day: apparently, early diverging linyphioids; and at least some species in Pimoa, Putaoa, Stemonyphantes, and Labulla), thus minimizing the distance that it needs to run to reach prey on the sheet, increasing attack speed. This “fast-attack” strategy is also feasible without a web, and may have thus predisposed some members of the large subfamily Erigoninae to lose webs, although the full extent of the absence of foraging webs in erigonines is unknown.

This suite of linyphiid traits (“the linyphiid syndrome”) contrasts with spiders that have thick or only moderately long legs, like theraphosids, ctenids, and lycosids. These species are built to be able to overcome prey by raw physical force, with their legs forcefully pressing the struggling prey against the spider's chelicerae (e.g., Rovner, 1978, 1980). Some other groups, like theridiosomatids and Hyptiotes (Uloboridae), also have relatively thick legs, to develop the force necessary to tense their webs (ditto the araneids Gasteracantha and Micrathena, which do not run rapidly to the prey when it strikes the web, but rather jerk the web). At the other end of the spectrum are spiders like pholcids with thin and very long legs that also run very rapidly but bite only briefly an extremity of their prey and rely on silk (wrapping + web) to immobilize prey (Eberhard, 2020). Relatively long legs are also associated with rapid running in the salticid Megaloastia mainae Żabka, 1995 (Soley et al., 2016). The linyphiid syndrome is near the pholcid end of this spectrum.

These tentative correlations between behavior and morphology seem to contrast sharply with the findings by Wolff et al. (2022: 1) that few morphological traits of spiders differ between ecological guilds: “Only few traits differed significantly between ecological guilds, . . . it was not possible to unequivocally associate a set of morphometric traits with the relative ecological mode.” That study, however, lumped together all spiders with aerial webs and thus did not account for the evidence of important behavioral differences (e.g., attack behavior in orb webs compared with sheet webs) that might correlate with morphological differences. Analyses of morphology that avoid the temptations of extreme simplification may be more rewarding.

The great British naturalist William S. Bristowe (1939: viii) prefaced his The Comity of Spiders book with a challenging question: Why do nearly half of the species and more than half of the population of British spiders belong to the family Linyphiidae? More than 80 years later we lack a clear answer to that question. The question itself would benefit from taking a more explicit comparative angle. Bristowe's (1941: 503) closing chapter provides what he considered the most satisfactory answer to that question: climate (e.g., tolerance to low temperatures), an abundance of food supply (primarily collembolans and nematocerous flies), an ability to live in diverse habitats, and a partial immunity from predators (many linyphiids are distasteful to spiders and centipedes) would explain the abundance and diversity of linyphiids in the temperate regions. Bristowe's hypotheses remain to be rigorously tested. Whether or not the linyphiid syndrome is responsible, at least in part, for the evolutionary diversification and ecological abundance of linyphiids is unknown. The fact that the families Pimoidae, Cyatholipidae, and Physoglenidae all appear to share several aspects of the linyphiid syndrome but nevertheless have relatively few species that are not as widespread or common suggests that this suite of traits is not the sole explanation for the evolutionary and ecological success of linyphiids.

Sticky Lines and Other Limitations of This Study

Contrary to previous conventional wisdom (e.g., Kaston, 1948; Bristowe, 1958; Comstock, 1967), the webs of many linyphiids contain substantial numbers of small sticky lines. They are difficult to perceive with the naked eye (Millidge, 1988; Peters and Kovoor, 1991; Benjamin and Zschokke, 2002; Eberhard, 2021). In this study most webs were coated with powder to photograph them, so it was generally not possible to distinguish sticky from nonsticky lines. One exception was Dubiaranea lugubris, which had large, closely spaced droplets on lines in the sheet (Figs. 22C, D). Our only other related observation was that the patches of “slime” on Laminacauda magna sheets (Fig. 40) were not sticky. The stickiness of the shiny, apparently slimy material covering of the naked sheets of Tapinopa longidens (Fig. 76A) was not tested. As already noted, fragmentary data indicate that there is probably substantial variation in the presence, the relative numbers, and the locations of sticky lines in linyphiid webs (Kullmann, 1971; Millidge, 1988; Peters and Kovoor, 1991; Schütt, 1995; Benjamin and Zschokke, 2002; Eberhard, 2021). These variables thus constitute important added dimensions of linyphioid web diversity not covered in this study.

All of the photos in this study were made in the field, so the histories of webs were not known; some were probably freshly built, whereas others may have included lines added over a longer period of time. We may have overlooked webs like those of Bathyphantes eumenis (Rybak, 2007) that were so sparse that we did not distinguish them from accumulations of drag lines.

Further Underexplored Aspects of Linyphioid Webs

Tensions on different lines are a largely unexplored aspect of web structure. It is possible, however, to intuit tensions indirectly from our photos. Suter (1984) showed that the “tensor” lines that produced dimples in the sheet of the bowl and doily spider Frontinella pyramitela increased the tensions on nearby lines in the sheet. The functional significance of this pattern of tensions is uncertain. It could simply be a guide to allow the spider to find a central resting site from which to monitor its web (Suter, 1984). In this species, the tension was greater in the “central area” of the cup-shaped sheet where the spider rested; the lines surrounding this area were less tense, whereas lines farther away, near attachments to the substrate, were more tense. We suppose that the dimples that we documented here also produced local increases in tensions on sheet lines. Similar dimples occur in the sheets of several other spiders that have a more or less horizontal sheet suspended between tangles of lines, including diguetids, several theridiids, and the araneid Cyrtophora citricola (Eberhard, 2020).

Another aspect of linyphiid webs needing further exploration is the possibility that sheets sometimes function to accumulate pollen, fungal spores, or water droplets. Convincing demonstrations of pollen feeding are seen in Frontinella pyramitela and Tennesseellum formicum (Emerton, 1882): spiders used specialized behavior to feed on corn (Zea mays Linnaeus, 1753) pollen in their webs that did not simply involve incidental ingestion during web destruction (Peterson et al., 2010: 210). They fed relatively quickly (within about 30 minutes) after pollen was applied, apparently without breaking web lines “by drawing the pollen grains into their mouths with alternating cheliceral movements, chewing with the endites, and periodic expulsion and retrieval of a droplet of salivary liquid from the oral cavity.” On the other hand, F. pyramitela failed to derive nutrition from the abundant pollen of pine trees (Carrel et al., 2000).