Web construction of Linothele macrothelifera (Araneae: Dipluridae) William G. Eberhard: Smithsonian Tropical Research Institute and Escuela de Biologı ´a, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica. E-mail: william.eberhard@gmail.com Nicolas A. Hazzi: Programa Acade ´mico de Biologı ´a, Universidad del Valle, Cali, Colombia Abstract. Direct behavioral observations, plus deductions made from studying the lines in recently built webs, showed that Linothele macrothelifera Strand 1908 lays swaths of lines in relatively stereotypic ways that differ during sheet web and tube construction. Sheet construction occurs in brief bursts interspersed with returns to the retreat. The legs are not used to manipulate lines; the spinnerets attach lines to the substrate and are probably used as sense organs. Asymmetrical use of the spinnerets during sheet construction results in an increase in the variety of orientations of lines in the sheet. Keywords: Mygalomorph, sheet web construction behavior Spiders in the family Dipluridae build some of the most elaborate prey capture webs among mygalomorph spiders (Coyle 1986). In the subfamilies Ischnothelinae and Euagri- nae, several species build complex arrays of numerous short tunnels that connect multiple small sheets and that mostly capture ambulatory prey (Coyle 1986, 1988, 1995; Coyle & Ketner 1990). Some species in the subfamily Diplurinae, including species in the genera Linothele, Trechona and Diplura, construct a single large horizontal sheet with a tubular retreat. Some of these sheets are suspended in the air many cm above the ground, and have tangles that extend up to a meter or more above the sheet, while others are built on the surface of the leaf litter or some other substrate (Coyle 1986; Paz 1988; Viera et al. 2007). It appears that other than the brief mention by Paz (1988) of the behavior of L. mega- theloides Paz & Raven 1990, nothing is known regarding the behavior patterns used by diplurids to build their webs. This note reports observations of the building behavior of Linothele macrothelifer Strand 1908, which builds sheet webs on the surface of forest leaf litter. This species, as is typical of non-orb weaving spiders in general, adds lines to its webs on successive nights. Our observations make use of the technique of damaging webs in the field and then observing newly constructed replacement webs, whose more sparse lines facilitate determination of patterns in the spider’s building behavior (e.g., Eberhard 1987; Benjamin & Zschokke 2003; Lopardo & Ramirez 2007). METHODS We made field observations on 1–4 December, 2011, near the end of the rainy season, at the Reserva Forestal de Yotoco (03u519500N, 76u269170W), a 550 ha patch of subtropical wet forest (Florez 1996), between 1300 and 1700m AMSL in the Western Cordillera of the Andes near Buga, Colombia. Sheet web construction behavior of one adult female was recorded using a SONY DCR-TRV50 video camera equipped with +7 close up lenses and infrared illumination. Individual lines emerging from her spinnerets were visible in some frames due to occasionally favorable angles of illumination. We collected portions of webs on small cardboard frames coated with double-sided adhesive tape, taking care to avoid including lines of other webs (e.g., of ochyroceratids) that were often built near the diplurid webs. Photographs of new webs were obtained by destroying sheets (leaving the tunnel mouth intact) in the afternoon and then coating webs with talcum powder the following morning. We include multiple web photographs because webs varied substantially in some respects (e.g., Figs. 1 & 2). Not all spiders whose webs were observed in the field were collected; we judged them to be mature females on the basis of the sizes of the spiders and their tunnels. We made further observations of sheet web and tube construction in captivity by two other adult females by covering the bottom of each of two 30 3 20 cm terraria with moist earth and creating tubular retreat cavities by inserting one finger. Video recordings were made from above with a digital Canon PowerShot A800 camera. Means are followed by 6 1 SD. Because of the small samples, they are meant only to provide general descriptions of magnitudes, rather than to characterize the behavior of this species. RESULTS We followed the webs of 12 individuals of L. macrothelifera over the course of 2–4 days. All consisted of a horizontal sheet extending from the edge of a tunnel, and were attached to the upper surfaces of leaves and twigs in the leaf litter (Figs. 1–4). The sheets were not perfectly flat, but followed the general contours of the objects in the litter. Usually nearly all lines formed a single sheet (Fig. 3), but some webs were somewhat multi-planar (Fig. 4). The individual lines in the sheets were relatively thin, and many were damaged or at least severely disorganized by the relatively moderate rains that fell daily (Fig. 5). In contrast, the walls of the tubular silk retreats were more dense and more protected, and persisted after rains. Thus the spiders largely rebuilt their sheets but not their tubes every evening following a rain. There were often large drops of water trapped in the complex, multilayered web near the mouth of the retreat, as also reported for L. megatheloides (Paz 1988). Sheet construction behavior.—Spiders in the field were out of sight in their silk tunnels during the day, and came to the mouth of the tunnel about 18:00–18:30 to rest motionless, facing outward. Web construction in the field occurred in 2013. The Journal of Arachnology 41:70–75 70