Introduction

Nomoclastidae Roewer, 1943, is a family of Neotropical species of harvestmen distributed from Costa Rica to Northern Brazil, it includes the subfamilies Zamorinae Kury, 1997 and Nomoclastinae Roewer, 1943, the first including the monotypic genus Zamora Roewer, 1927 and the second comprising the genera Callcosma Roewer, 1932, Globitarsus Roewer, 1913, Kichua Pinto & Bragagnolo, 2015, Liomma Roewer, 1959, LisareaRoewer, 1943, MeridanatusRoewer, 1943, Micropachylus Roewer, 1913, Napostygnus Roewer, 1929, Nomoclastes Soerensen, 1932, Quindina Roewer, 1914 and Troya Roewer, 1914 (Kury et al., 2024; Kury & Villarreal, 2015; Pinto-da-Rocha & Bragagnolo, 2017; Villarreal & Kury, 2023). The family is defined by the following characters: no sexual dimorphism in its chelicerae and pedipalps (except Liomma); low, elliptical ocularium, with a median depression; ventral plate of the penis sub-rectangular to trapezoidal in shape; no lamellar setae on the ventrodistal margin of the ventral plate; short, stout stylus with a rounded head, mounted on a long, columnar glans; glans without processes except in Napostygnus (Pinzón-Morales & Pinto-da-Rocha, 2020). Nomoclastidae currently includes 12 genera (8 monotypic) and 29 species, including the one herein described (Kury et al., 2024; Pinto-da-Rocha & Bragagnolo, 2017; Pinzón-Morales & Pinto-da-Rocha, 2020; Villarreal & Kury, 2023).

The genus Quindina:Quindina is characterized by a type of body armor similar to that of Callcosma and Kichua, with a pair of long spines on area III and free tergites with a pair of spines each. It differs from other genera of the family by most species having a row of tall tubercles separated from each other on the lateral margins of dorsal scutum, very elongated tubercles on area III of doral scutum, absence of lateral margin elevation, and the presence of an apical prodorsal projection on femur IV (Pinto-da-Rocha & Bragagnolo, 2017; Pinzón-Morales & Pinto-da-Rocha, 2020). It is distributed in Costa Rica, Panama, Colombia and Ecuador. Eleven species had previously been described in this genus, four of which from Colombia: Q. bella Roewer, 1914 (type species) and Q. marginata (Roewer, 1963), from the Andean region; Q. discolorPinzón-M. & Pinto-da-Rocha, 2020 and Q. hermesiPinzón-M. & Pinto-da-Rocha, 2020, from the Caribbean region (Pinto-da-Rocha & Bragagnolo, 2017; Pinzón-Morales & Pinto-da-Rocha, 2020).

Parental care and nest building: Exclusive paternal care evolved independently in fifteen arthropod lineages, which include approximately 1 500 species. This is a very small fraction of the diversity of this group (with more than 1.2 million species), making it probably the rarest form of postzygotic parental care investment (Requena et al., 2014). Egg guarding behavior is observed in most orders of the class Arachnida, but paternal investment in postzygotic care is restricted to members of the order Opiliones. For example, in Gonyleptidae, adult males of Ampheres leucopheus (Mello-Leitão, 1922) protect batches of eggs, attached by a mucous substance to the underside of leaves, and newly hatched nymphs (Hara et al., 2003). Three species of Progonyleptoidellinae Soares & Soares, 1985 show a similar behavior, with eggs being laid on the underside of leaves, covered by a thick mucous layer, while two species of Gonyleptinae Sundevall, 1833 lay eggs in natural cavities, in trunks or on the ground. In all these cases, eggs can be found in different stages of development with males taking care of them (Machado et al., 2004). Maternal care is also present in the order Opiliones and has been reported for at least five families of the suborder Laniatores Thorell, 1876. There are cases of maternal care in which females care for eggs in cavities and leaves or for example, two species of Cranaidae, belonging to the genus Phareicranaus Roewer, 1913, have females that protect groups of eggs in cavities found between the roots of trees. The same type of behavior has been observed in females of the Gonyleptidae subfamily Goniosomatinae (Machado & Warfel, 2006).

The aforementioned cases imply paternal or maternal care of the eggs in surfaces or cavities, but without the construction of complex structures to protect the eggs. Rodriguez and Guerrero (1976) provided the first description of harvestmen nests for Quindina albomarginis (Chamberlin, 1925). The authors indicated that the construction of each nest corresponded to a single male, and that the material used came from the outer bark or outer bark remains of the trunk on which it was built, or from logs and debris from the vicinity. Mora (1990) complemented these observations and indicated that, for the construction, the males use bark removed with the chelicerae and mixed with mud collected from the crevices of the trunk. She also noted that material could be collected from the forest floor, combined with salivary secretions and later applied on the substrate to assemble the nest, with a mud wall then applied to the edges, with an average nest diameter of 33 mm and an average nest wall height of 8.3 mm. Pinto-da-Rocha and Bragagnolo (2017) described four species of Quindina from Panama, three with the same type of nest (Q. moraePinto-da-Rocha & Bragagnolo, 2017, Q. burbayarPinto-da-Rocha & Bragagnolo, 2017 and Q. kunaPinto-da-Rocha & Bragagnolo, 2017), which they called “open mud nest” or “reproductive arena”. Quesada-Hildalgo et al. (2019) and Rojas et al. (2019) also reported this type of nest for Q. limbata (Roewer, 1943) in Costa Rica, calling it “mud-nest” or “cup-like mud nest”.

Our aim in this study is to describe a new species of the genus Quindina, which builds a new type of nest. In addition, we strive to determine the position of the species in the phylogenetic tree of its respective genus and hypothesize about the origins for the two nest types recorded in this group.

Materials and methods

Collection and conservation of specimens: We collected the male specimens in the vereda Mundo Nuevo and females in the vereda El Líbano, in the municipality of La Calera, Cundinamarca, Colombia. The specimens were gathered at the site where the nests were found, by direct observation of the tree branches, the underside of trunks and the interior of an abandoned hut. After locating the nests, we waited until nightfall to collect the male individuals coming back to the nests. Two male individuals were fixed in 70 % ethanol; two others were transported alive, in order to observe their nest building in terrariums. Females were collected under logs and rocks near a road in the vereda El Líbano. The two female specimens collected were fixed in 70 % ethanol. All specimens were deposited in the collection of the “Museo de Ciencias de la Universidad El Bosque (MCUB)”.

Measurement and description of the specimens: We observed the general characters of the body and appendages under a stereoscope, with magnifications of 10-180X. The penis was observed under an optical microscope, with magnifications of 400X and 1 000X. In both cases, photographs were taken with a digital camera attached to one of the eyepieces. In the photographs taken from the stereoscope, a millimeter scale was included. In the microscope photographs, dimensions were calculated using the millimeter scale recorded at 4X magnification. These scales were used to measure the parts with ImageJ software (Rueden et al., 2017).

All measurements are given in millimeters. Illustrations were made with LibreOffice Draw (The Document Foundation, 2022) and Gimp software (Kimball et al., 2022). The delineation of the parts was drawn following the contour of the digital photographs. Descriptions and terminology follow the ones proposed by Pinto-da-Rocha and Bragagnolo (2017) and Pinzón-Morales and Pinto-da-Rocha (2020). Description of genitalia follows the findings of Kury and Villarreal (2015) and Kury (2016). Indications for nomenclature of the structures, measurements and relative positions of structures, follow Acosta et al. (2007).

Distribution records: At the type locality (vereda Mundo Nuevo, municipality of La Calera) a photographic record of the vegetation type was taken using a drone. The collection records of the type specimens was complemented by previous observations made by Héctor Lancheros in the rural area of the Bogotá, Capital District, and other new records made in the municipalities of Choachí (Department of Cundinamarca), Saboyá (Department of Boyacá) and Une (Department of Cundinamarca), by biologists Daniel Arias Cuellar, Christian Franco García and Osvaldo Villarreal, respectively. The distribution map was assembled using QGIS software (QGIS.org, 2021), based on the coordinates recorded with Viking software (Battaglia et al., 2021).

Description of the nests: The fresh weight of twelve nests was recorded in the field, using an electronic scale with 3-digit precision. Length and width dimensions for each one were measured using ImageJ software (Rueden et al. 2017), based on photographs taken on the scale plate, using its diameter as a point of reference. Thickness was recorded using a digital caliper. The dry weight of eight nests was recorded after drying them in a hot air circulation oven at 30 °C. Four nests were used for behavioral observations in the laboratory, where we paid special attention to the care provided by adult males and egg development and hatching. These four nests were also used to analyze the materials that compose them. In each nest, the type of plant material from which it hung was identified. The remains of liverworts present in nests were identified down to genus level.

Observations of parental care and development: The first observations of parental care were made directly in the field, from the beginning of the day until the early evening hours (from 18:00 to 20:30). For observations under controlled conditions in the laboratory, two terrariums were prepared, each containing a section of branch from which hung a nest in a plastic box, with the respective adult male. Observations were made in each terrarium for two weeks. For developmental observations (i.e. changes in egg coloration, hatching and nymph development), two complementary set-ups were made on absorbent paper in Petri dishes, each with only one nest. Observations for each petri dish lasted between 20 days and one month, until hatching ceased. Adults and nymphs were fed using cooked rice with peas and carrots, cooked spaghetti and dead fruit flies, in both terrariums and petri dishes.

Phylogenetic analysis: The matrix used for the phylogenetic analysis was originally published by Kury and Villarreal (2015), including 77 morphological characters and 38 taxa (26 from the clade Laminata). It was later modified by Pinto-da-Rocha and Bragagnolo (2017) who added 13 species in Nomoclastidae, adjusted the coding of characters 11 and 12, and added 16 new characters (78 to 93), two of which originated from the subdivision of the original character 13 (characters 78 and 82). To this, Pinzón-Morales and Pinto-da-Rocha (2020) added two new species of the genus Quindina. With the addition of the new species, the matrix now comprises 93 characters and 54 taxa (Table 1), all belonging to the suborder Laniatores. The taxa comprise 48 species of the superfamily Gonyleptoidea (infraorder Grassatores), of which 42 belong to the clade Laminata, which is divided into 20 species of Nomoclastidae and 22 of Microsetata. Outgroups comprise five species from three other superfamilies of the infraorder Grassatores (Samooidea, Epedanoidea and Assamioidea) and the root is one species from the superfamily Triaenonychoidea of the infraorder Insidiatores (the latter is the first outgroup) (Kury & Villarreal, 2015). The description of the characters can be found in Pinzón-Morales and Pinto-da-Rocha (2020). Species of the most recently added genera to Nomoclastidae (Globitarsus, Liomma, Lisarea, Meridanatus, Micropachylus and Troya) were not included due to lack of reproductive data and several morphological data for matrix. The parsimony analysis was carried out using the software TNT v.1.5 (Goloboff & Catalano, 2016). The heuristic method (“traditional search”) was applied with the tree bisection-reconnection (TBR) option with equal weights. From the trees with maximum parsimony, two cladograms were obtained, one using the strict Nielsen consensus method and another using majority consensus, with 50 % cutoff. Subsequently, the implicit weighting method was applied, which decreases the weight value of homoplastic characters in the analysis based on a concavity constant (k), the value of this constant is negatively related to the decrease in the weight of these characters in the analysis (Goloboff, 1993). To obtain the “optimal” value of the constant k for our data set, the TNT script setk.run was used (Goloboff et al., 2008; Villarreal & García, 2021). Finally, the successive weighting method was applied with the rescaled consistency index rc (Farris, 1989), using the script rewt.run. The latter is a technique that decreases the weight of homoplastic characters after obtaining a cladogram and repeats the process until successively equal trees are obtained (Farris, 1969). Relative Bremer supports were calculated for the trees obtained with the four methods, which provide an approximate measure of the amount of favorable/contradictory evidence for each group (Giribet, 2003; Goloboff & Farris, 2001), and the tree with the highest support values was chosen. Sensitivity plots (Navajo rugs) were used to compare the coincidence of the groups obtained with the different methods and hypotheses.

Hypotheses of the origin of the nests: To propose hypotheses of origin of the nests, the most parsimonious reconstructions method was used (Goloboff, 2022, Chapter 3), with the TNT Characters/Reconstructions option in the optimize menu. For this, the genus Quindina section of the chosen tree was used, with an external group and the character nest presence with three unordered states: 0 = absent, 1 = pendular disc nest, 2 = open mud nest, ? = nest construction unknown. The reconstructions obtained were stored in the display buffer. The frequency with which each nest type occurred in a reference group (each group is one of the main clades that make up the genus Quindina) was tabulated for each possible “condition” (ancestral, derived or in an internal group). The combination of each nest type with the reference group was called “characteristic”, and for each characteristic, the condition with the highest frequency was listed. This is an illustrative method to indicate the possible nodes of origin of each nest type, as unknown data for most species of Quindina does not allow for making a complete reconstruction of the evolution of the nest type and its origin.

All figures (illustrations, map, photographs and cladograms) were diagrammed using LibreOffice Draw graphics editing software (The Document Foundation, 2022).

Results

Quindina pendula Lancheros, Arias-Cuellar & Pinto-da-Rocha sp. nov. zoobank.org:act:69BE1437-BEA0-4006-9528-329D725401DF

Etymology: The specific epithet, pendula, is a Latin term meaning “pendant” and refers to the pendular disc nest, contrasting with the open mud nest attached to the subtract.

Type material: Holotype (MCUB-R-AR-000938) adult male preserved in 70 % ethanol, the nest is preserved dry, Colombia, Cundinamarca, La Calera, vereda Mundo Nuevo (4°39’55.7’’ N & 73°51’ 16.9” W; 2 693 m elevation), 16.II.2019, Daniel Arias Cuellar leg. Paratypes: two males (MCUB-R-AR-000936), from the same locality as the holotype, 19.XII.2020, Daniel Arias Cuellar leg.; two females (MCUB-R-AR-000939), Colombia, Cundinamarca, La Calera, vereda El Libano (4°39’44’’ N & 74°0’31’’ W; 3 006 m elevation), 27.III.2021, Nicolás Briceño Avellaneda leg.

Diagnosis: Ocularium with a group of 3-6 tubercles near each eye, towards the center. Prosoma with dark yellow background on most tubercles both dorsally and ventrally. Area I with a group of uniformly tubercles forming a triangle on each side. Area III with two or three pale yellow patches on each side. Area IV with one pale yellow patch on each side. Dorsal scutum areas and tergites without white spots. Free tergites predominantly black on the sides and yellow in the center.

Comparison: Comparison: Resembles Q. bella, Q. bimaculata, Q. discolor, Q. hermesi, and Q. marginata, by the lack of enlarged tubercles on the lateral margins of the dorsal scutum. It can be distinguished from Q. bella by the presence of only small yellow tubercles on the lateral margin of the dorsal scutum (a group of larger, white tubercles on each side of the lateral margin in Q. bella); from Q. bimaculata by the presence of dark yellow background around most of the tubercles on the dorsal and ventral body, and in free tergites I-III (yellowish white tubercles and a white patch with three tubercles in each side of the lateral margin of the dorsal scutum in Q. bimaculata); from Q. discolor by the dark yellow background of the prosome with black areas (presence of a white patch on the posterior half of the prosoma in Q. discolor); from Q. hermesi by presence of only yellow and black coloration (anterior margin tubercles white and a white spot on each side of the lateral margin near the back of the dorsal scutum in Q. hermesi); and from Q. marginata by the smaller number of tubercles in the ocularium and the absence of these in the posterior part of the prosoma, the presence of only yellow, brown and black coloration and a pair of parallel spines in the free tergite III (presence of 18 tubercles in the center of the ocularium and 23 behind of this, silver tubercles and absence of a pair of parallel spines in the free tergite III in Q. marginata).

Description, Male Holotype (MCUB-R-AR-000938)

Measurements: Carapace length: 1.4; carapace width: 2.2; dorsal scutum length: 3.1; dorsal scutum width: 2.8; femur: 1.6, 3.6, 2.4, 3.3; patella: 0.6, 1, 0.9, 1.1; tibia: 1.1, 2.5, 1.6, 2.1; metatarsus: 1.9, 3.2, 2.5, 3.6. Variation in male measurements (Table 2).

Live color (Fig. 1A): Dorsal shield with brown background, ocularium with black patches on the lateral areas, middle part of the prosoma with a mushroom-shaped set of brown patches. Black patches in 49 % of the prosoma, the other 51 % of the prosoma brown. Areas I and II with dark yellow dots. Spines of area III black with two bright yellow patches on each side. Most of the dorsal and ventral tubercles have dark yellow patches. Lateral areas with small irregular black patches. Lateral edges of the shield are dark yellow. Free tergites black with light yellow lateral patches. Legs light yellow with a pattern of irregular black patches. Pedipalps and chelicerae are dark yellow with black patches. Ventral body dark yellow with heterogeneous black patches. Free sternites with black lines and light yellow patches. UV fluorescence is observed in free tergites I to III (Fig. 2C and Fig. 2D).

Fig. 1
Quindina pendula sp. nov. male holotype (MCUB-R-AR-000938). A-B. Habitus, A, dorsal view; B, lateral view. C. Chelicerae hand with fixed and movable fingers. D-E. Pedipalp, trochanter to tarsus, C, meso-dorsal view; D, ventral view. F. Leg I, trochanter to tarsus in ventral view. Scales: 1 mm.

Fig. 2
Observations of the adult male and nymphs of Quindina pendula sp. nov. in the terrarium. A-B. Adult male tending the nest in the terrarium. C-D. UV observation for evidence of fluorescent areas. E. 2-day-old nymphs on the nest. F. 21-day-old nymph. Photos: A, B, C and D. Daniel Arias; E and F. Héctor Lancheros.

Color in ethanol: Background of the dorsal scutum dark yellow. Ocularium with black patches on the lateral areas. Black patches in 49 % of the prosoma, the other 51 % of the prosoma dark yellow (Fig. 1A). Area I and II with dark yellow dots. Spines of area III black, each with two or three pale yellow patches forward on each side (Fig. 1A). Most of the dorsal and ventral tubercles have dark yellow patches. Lateral areas with small irregular black patches (Fig. 1B). Free tergites with black lateral patches. Yellow pedipalps with black patches. Chelicerae with a dark yellow background with heterogeneous black patches. Legs yellow with black patches. Ventral part is dark yellow with heterogeneous black patches. Free sternites with black lines and dark yellow patches.

Dorsal scutum (Fig. 1A): Type β dorsal scutum, anterior margin with one tubercle on each side. Preocular monticle with some granules. Carapace with one tubercle at each side of the preocular monticle, and one tubercle forward or behind of each ozopore. Ocularium with a small median depression, with two paramedial groups of 3-6 tubercles. Lateral margins with a row of 19-20 tubercles from center of coxa III to almost posterior margin, elevated posteriorly. Grooves of the dorsal scutum are well marked. Area I with 7-8 tubercles forming a triangle on each side. Area II with a row of eleven tubercles. Area III with a pair of sharp, slightly divergent spines, directed backwards; with four tubercles forward of each spine and one or two tubercles at the base or over these. Area IV with four tubercles on each side. Posterior margin with nine tubercles. Free tergites: I and II with a row of twelve tubercles; III with a pair of parallel spines directed backwards, with four tubercles between spines and four at each side. Anal plate with nine small-scattered tubercles.

Ventral: Coxa I with 13 scattered tubercles. Coxa II with 28 tubercles. Coxa III with 21 scattered tubercles. Coxa IV with more than 50 tubercles.

Legs (Fig. 1F): I and III with small tubercles on the front, II with one medium-sized and one large tubercle on the front. Trochanters: I-IV with small scattered setiferous tubercles in the dorsal, ventral and lateral parts. Femora I-IV with rows of various setiferous tubercles; the apex of the femora II-IV presents two spiny lateral tubercles, the retrolateral larger than the prolateral. Tarsal articles: 6(3)/12-13(3)/7/8.

Chelicerae (Fig. 1C): Basal segment with a few small setiferous tubercles on the bulla. Movable finger with four teeth; fixed finger with five teeth.

Pedipalps (Fig. 1D and Fig. 1E): Trochanter with a few small dorsal tubercles, one ecto-lateral and one ventral setiferous tubercles. Femur with small scattered dorsal tubercles. Tibia: mesal iiIi, ectal iIi; Tarsus: mesal IiIi, ectal IiIi.

Penis (Fig. 3): Ventral plate oval in shape, lateral margin with a basal constriction, half of the distal edge curved, three pairs of spatulate macrosetae C; one pair of spatulate macrosetae A directed towards the truncus, almost as long as macrosetae C; one pair of conical macrosetae D; three pairs of macrosetae E on ventral surface of ventral plate. Glans wide and conical, stylus long, slightly wide at apex, with submedian ventral process. Type 4 microsetae distributed in the ventral side of the ventral plate.

Fig. 3
Penis of Quindina pendula sp. nov., holotype (MCUB-R-AR-000938). A. dorsal view, B. ventral view, C. ventrolateral view, D. lateral view. MS = macrosetae, μS 4= type 4 microsetae. Scale: 0.01 mm.

Female Paratype (MCUB-R-AR-000939)

Measurements: Carapace length: 1.2; carapace width: 1.8; dorsal scutum length: 2.8; dorsal scutum width: 2.5; femur: 1.1, 2.1, 1.5, 2.2; patella: 0.4, 0.6, 0.7, 0.7; tibia: 0.8, 1.4, 1.0, 1.4; metatarsus: 1.1, 1.8, 1.5, 2.3. Variation in female measurements (Table 2).

Females differ from males in several morphological aspects and have dark brown spots instead of black spots (Fig. 4). Anterior margin of the scutum with two tubercles on each side. Ocularium with two paramedial groups of tubercles. Lateral margins without a row of tubercles. Area III and free tergite III lacking a pair of spines. Free tergites II and III with sharper and larger tubercles (Fig. 4B).

The number of tubercles is different in areas I-IV of the dorsal scutum, the free tergites and the anal plate. Area I with a row of two or four tubercles on each side. Area II with a row of three tubercles on each side. Area III with four to eight tubercles on each side. Area IV with three to four tubercles on each side. Posterior margin with ten tubercles. Free tergites: I with a row of ten tubercles; II with a row of 13 tubercles; III with eleven tubercles. Anal plate with two rows of four small tubercles. Tarsal counts smaller than males: 5(3)/8(3)/6/6.

Fig. 4
Comparison of male and female of Quindina pendula sp. nov.. A. Dorsal view of the male holotype (MCUB-R-AR-000938), B. Dorsal view of a female paratype (MCUB-R-AR-000939). Scale 1mm. Photos: Héctor Lancheros.

Distribution (Fig. 5): Recorded in the department of Cundinamarca: municipalities of La Calera (Fig. 6), Choachí (Fig. 7A, Fig. 7B, Fig. 7C), Tena; and in the Distrito Capital locality of Chapinero (Bogotá). The elevation range is between 2 040 and 3 150 m and the observed vegetation corresponds to the Andean Forest formation. In addition, there are photographic records of the same type of nests of this species in the municipality of Saboyá (department of Boyacá), at an elevation of 2 619 m, these could belong to the same species, but there is no record, of the individuals from this locality, deposited in scientific collections.

Fig. 5
Sites where Quindina pendula sp. nov. or pendular disc nests have been recorded in the departments of Cundinamarca and Boyacá. The color of the dots indicates the type of record; the blue dot represents the record of the nests without the individuals, so there is no certainty that it is the same species.

Fig. 6
Municipality of La Calera, vereda Mundo Nuevo. Elevation: 2 693 m. A. General view of the vegetation cover in the area. B. Relief of the area. C. Abandoned hut where two nests were observed. D. Nest suspended from liverworts on a tree branch. E. Adult of Quindina pendula sp. nov. protecting the nest. Photos: A, B, C and D. Héctor Lancheros; E. Daniel Arias.

Fig. 7
A-C. Observations in the municipality of Choachí, Forest near La Chorrera waterfall, elevation: 2 620 m. A. High cover of liverworts on tree branches in the forest interior. B. Nest hanging from liverworts on a tree branch. C. Male harvestman found in the nest. D-F. Observations of juvenile individuals in the wild. D-E. Shortly after hatching the nymphs remain in the nest. F. Juvenile individual found on the ground under a log. Photos: A, B and C. Daniel Arias; D, E and F. Héctor Lancheros.

Nest characteristics: We report a new type of nest in Quindina pendula sp. nov. This species builds pendular disc-shaped nests in which the female lays the eggs that are cared for by the male. Most of these pendular disc nests were observed on tree branches in well-preserved Andean Forest areas, generally on branches colonized by foliose liverworts. Apparently, the nest is composed of mud. However, upon close examination, one can observe only plant fragments such as bark segments and bryophytes (Fig. 8), united by an adherent substance that is presumably secreted by the harvestmen. A detailed examination of one of the nests revealed the presence of small arthropods on its surface: a Psocoptera nymph of the family Ectopsocidae and a mite, apparently of the order Mesostigmata (Fig. 8E and Fig. 8F).

Fig. 8
Nest structure of Quindina pendula sp. nov. A. General view of the nest. B-C. Detail showing the elements from which it hangs, a thin branch and a leaf of Tillandsia sp., we observed that the nest is composed of small fragments of plant bark. D. Phyllidium of a liverwort that is part of the nest structure. E-F. Mite, probably of the order Mesostigmata, and Psocoptera nymph of the family Ectopsocidae found in the nest. Photos: Héctor Lancheros.

The nests were found suspended from various structures such as sections of leafy liverworts (Frullania atrata, F. cf. brasiliensis and Plagiochila sp.), thin branches of trees, leaves of bromeliads, among others (Fig. 8B, Fig. 8C, Fig. 9 and Fig. 10). Seven nests were observed in a single tree, four of these on a branch. Two nests were observed (Fig. 10) in an abandoned hut in the forest of the vereda Mundo Nuevo, one hanging from the lower part of the sill beam (Fig. 10B and Fig. 10C) and the other inside near the roof, suspended from a blackberry plant that was growing partially inside the hut (Fig. 10D and Fig. 10E). Nests generally hang from a single point, but occasionally hang from two (Fig. 9C and Fig. 10C); In most observations the nest was at the terminal part of the supporting element, but occasionally it was observed in the middle part, with a section of the element protruding at the bottom (Fig. 9C, Fig. 9D and Fig. 7D).

Fig. 9
Nests of Quindina pendula sp. nov. found in tree branches, suspended from different structures A. tree with high epiphytism of foliose liverworts. B. Nest built on the apex of an epiphyte of the genus Tillandsia. C. Nest built around a liverwort with two support points. D. Nest built around the middle part of a liverwort. E. Nest built on the terminal part of a slender branch. Photos: Héctor Lancheros.

Fig. 10
Nests of Quindina pendula sp. nov. found in an abandoned hut inside the forest of the vereda Mundo Nuevo. A. Hut hidden inside the forest. B. Nest hanging on the lower part of the sill beam. C. Detail of the nest hanging on the sill beam. D. Adult harvestman tending a nest on a blackberry plant. E. Nest suspended from the blackberry plant (Rubus sp.) inside the hut. Photos: Héctor Lancheros.

The observed values of mass and dimensions were as follows: fresh mass 0.12-0.6 g; dry mass 0.08-0.01 g; length 16-20 mm, width 15-21 mm, average diameter 15-21 mm; thickness 3.80-4.35 mm; surface area (sum of the two faces) 353-693 mm². The value of fresh mass is highly variable due to the different degree of hydration, which depends on the location of the nest and the precipitation before data collection. For dry mass n = 8, for the other variables n = 12.

Observations on parental care: Adults were not observed in the nests during the day. In the field, after 6:00 p.m., we observed that males remained constantly on or near the nest (Fig. 6E, Fig. 10D and Fig. 10E). In the lab, we observed that, at night, the male takes care of the nest, even if it is damaged, repairing the nest when it suffered minor damage (Fig. 2A, Fig. 2B, Fig. 2C and Fig 2D). During the day the male remained hidden at the bottom of branches. We also observed that, when the nest of another male is placed inside the box, the male continues to take care of it, even if it is not his own. We have not recorded any females in or near the nests.

Developmental stages from laying to hatching: Initially the eggs were completely covered by nest material, but as they developed and increased in size, they became partially exposed. It appears that deterioration of the nest due to rain also causes them to gradually become exposed. The fact of being partially uncovered does not cause direct changes in the color of the eggs, they only change color gradually as they advance in their development. Preliminary observations show four stages of development from egg laying to hatching: white eggs with no visible differentiated structures; brown eggs with visible appendages; black eggs; emergence of the anterior part of the body (Fig. 11). At least 22 days elapse between laying and the second stage. Then, brown eggs take approximately 12 days to develop into black eggs and 10 to 12 days elapse between the third and fourth stages. In total, at least 45 days elapse from laying to the beginning of hatching (Fig. 11).

Fig. 11
Nest of Quindina pendula sp. nov. with eggs at different stages of development, the circles on the absorbent paper are 1 mm in diameter, the eggs have a diameter of 0.55 mm. A-B. 10 days after collection, with black and white eggs, two hatchings are also observed (orange circles). C-D. 22 days after collection, on one side of the nest there are eggs in the second and third stage of development, brown and black, respectively; also, an egg shell is observed after hatching (green circle). E-F. 22 days after collection, on the other side of the nest there are eggs in the first, third and fourth stage of development: white, black and initiating hatching (red circles), respectively. Photos: Héctor Lancheros.

In the field, the nymphs remain in the nest shortly after hatching (Fig. 7D and Fig. 7E). They can fall to the ground when hit by raindrops or can take shelter again in the holes on the nest that remain after hatching. This was observed once in the first nest recorded in the vereda Mundo Nuevo during the day, when the adult male was absent. A juvenile individual was found on the ground, under a log; this may be the environment in which they remain before reaching maturity and initiate nest construction (Fig. 7F). As in the natural environment, in the terrarium, nymphs remained in the nest for the first few days (Fig. 2E and Fig. 2F).

Phylogenetic analysis: Using the equal weights method, we obtained 30 equally parsimonious trees with 380 steps. From these, we obtained a strict consensus tree of 416 steps and a majority consensus tree of 386 steps. Using the implicit weighting method, with k = 9.2188, obtained with the setk.run script, we obtained a tree of 381 steps. Using successive weighting, with the rc index, we obtained a tree of 381 steps. With all methods, we recovered Nomoclastidae as a monophyletic unit, sister to Microsetata; in the latter, the relative position of the species is similar to that obtained by Kury and Villarreal (2015), with the exception of the species in the Ampycus telifer, Neopachyloides sp., Glysterus metatarsalis and Nesopachylus monoceros clade, with the strict consensus, implicit weighting and successive weighting methods (Appendix 1A, Appendix 1B, Appendix 1C, Appendix 1D). The relative position of the other Gonyleptoidea families, as well as that of the outgroups, was the same as that obtained Kury and Villarreal (2015) using the successive weighting method but slightly different with the implicit weighting method, the latter differs in the position of Sorensenius and Dibunus (Appendix 1C, Appendix 1D).

Concerning the relative position of the genera within Nomoclastidae, the different methods showed Zamora as sister to the other genera. Nevertheless, the methods do not agree on their relative positions. The strict consensus tree with equal weights showed a polytomy between all species of Callcosma and the genera Kichua, Napostygnus, Nomoclastes and Quindina (Appendix 1A). The majority consensus tree with equal weights showed a polytomy between Napostygnus, Nomoclastes and the clade (Kichua (Callcosma Quindina)) (Appendix 1B). The successive and the implicit weighting tree shows a well-defined topology, with Nomoclastes sister to the group ((Kichua (Napostygnus Callcosma)) Quindina) (Appendix 1C, Appendix 1D); in addition, the highest relative Bremer support values were obtained with the successive weighting method (Appendix 1D).

Regardless, the main object of this study is the genus Quindina, the topologies we obtained with the implicit and successive weighting methods show Quindina marginata as sister to the remaining species of the genus (including Q. pendula sp. nov.). Similarly, we recovered Quindina pendula sp. nov. as sister to the remaining 10 species of the genus. Sensitivity plots (Navajo rugs) indicate that the two topologies coincide in seven of the twelve species of the genus Quindina, the relative position varying within the clade formed by Q. hermesi, Q. bella, Q. kuna, Q. limbata and Q. albiocularia. We obtained the highest values of Bremer relative support with the topology recovered by the successive weighting method (values between 53 and 74 for each clade within the genus), so we chose this cladogram for the comparison with the other methods (with the sensitivity plots) and the approach of the possible hypotheses of the origin of the pendular disc-shaped and open mud nests (Fig. 12).

Fig. 12
Cladogram of Nomoclastidae obtained by successive weighting with rc, including the new species. Sensitivity plots (Navajo rugs) indicate, with red squares, agreement with groups obtained by other methods: strict Nelsen consensus (NE)., majority consensus > 50 % (MA) and implicit weighting with k = 9.2188 (IW). Numbers in each branch indicate relative Bremer supports. Next to each species there are drawings indicating the type of nest observed in each; nest descriptions are presented in the boxes on the right. The other species of Gonyleptoidea have been omitted as they are not the subject of discussion in this paper (Appendix 1D).

Hypotheses of the origin of the nests: We have obtained seven hypotheses for the origin of the two types of nests in the genus Quindina using the method of most parsimonious reconstructions with the tree obtained by the successive weighting (Fig. 13). The characteristics observed in the hypotheses are summarized in Table 3. For each characteristic (nest type for each reference group) we give the frequency of each condition (ancestral, derived or in an internal group), the sum being equal to seven. The last column shows the condition that we observed most frequently (in most hypotheses).

Fig. 13
Hypothesized origin of nests in the genus Quindina. A. Independent origin of the pendular disc nest in Q. pendula and the open mud nest in the sister group of this. B. Origin of the pendular disc nest in the common ancestor of the sister group of Q. marginata and modification to the open mud nest in the sister group of Q. pendula. C. Origin of the open mud nest in the common ancestor of the sister group of Q. marginata and modification toward the pendular disc nest in Q. pendula. D. Origin of the pendular disc nest in the common ancestor of Quindina spp. and modification to the open mud nest in the sister group of Q. pendula. E. Origin of the open mud nest in the common ancestor of all Quindina species and modification toward the pendular disc nest in Q. pendula. F. Origin of the pendular disc nest in an ancestor external to the genus Quindina with modification toward the open mud nest in the sister group of Q. pendula. G. Origin of the open mud nest in an ancestor external to the genus Quindina with modification toward the pendular disc nest in Q. pendula.

The obtained results differ on the origin of the nests and the type of nest that first appeared. The lack of data for other nomoclastid genera hindered us to know if these genera share nest construction (Fig 13F, Fig. 13G) or if this is exclusive for Quindina spp. (Fig. 13A, Fig. 13B, Fig. 13C, Fig. 13D, Fig. 13E). The absence of reproductive data for the first derived species of Quindina, Q. marginata, is also an obstacle in understanding how nest construction evolved within the group. At this moment it is impossible to know if the pendular disc nest is exclusive for Quindina pendula sp. nov., or if it evolved first and then was replaced by an open mud nest. The only consensus among all hypotheses is that most species of Quindina share an open mud nest as a strategy to lay eggs and raise nymphs. Thus, this clade is composed of species for which we observed an open mud nest in the field (Q. albomarginis, Q. morae, Q. burbayar, Q. kuna, and Q. limbata) and other species that we hypothesize also build this kind of nest.

Discussion

Previously described nest-building harvestmen are distributed across the territories of Costa Rica (Q. limbata) and Panama (Q. albomarginis, Q. burbayar, Q. kuna and Q. morae) (Pinto-da-Rocha & Bragagnolo, 2017), making Q. pendula sp. nov. the first nest-producing species recorded in Colombia. In addition, this new species represents the first record of a pendular disc nest, built with fragments of bark and other plant remains. The pendular disc nest differs significantly in shape, substrate type and substrate attachment, from the previously reported nest type, which is an open mud nest. This last kind of nest is totally attached to the substrate, surrounded by walls and built with bark fragments and other materials. Both nest types do share some similarities, specifically in the type of materials which compose them (Mora, 1990; Rodríguez & Guerrero, 1976). Nevertheless, the composition of the previously reported nests needs to be re-examined through detailed microscopic observations in order to verify its accuracy.

The presence of small arthropods on the surface of the pendular disc nest, such as psocids and mites, could be related to their use of the nest as a food source. Insects of the order Psocoptera are mainly phytophagous, feeding on algae, fungi, pollen and fallen plant parts. On the other hand, Mesostigmata mites are mainly free-living predators, where some are phoretic on insects and some consume pollen, nectar and fungi (Dhooria, 2016, Chapter 5). The Ectopsocidae (Psocoptera) may roam the nest consuming plant debris or the mycelium of fungi growing on its surface. Meanwhile, the mites may also be in the nest to eat mycelium or, instead, act as a predatory species, consuming small arthropods such as psocoptera or harvestmen nymphs.

The addition of Q. pendula sp. nov. to the phylogenetic analysis result in great differences in the relationships within Quindina when compared with Pinzón-Morales and Pinto-da-Rocha (2020); the implicit and successive weighting methods coincide only in Q. marginata as sister group of the remaining species in the genus and in the clade (Q. albomarginis (Q. morae Q. burbayar)). However, the topology of the different nomoclastid genera is identical between our findings with these two methods and the previous study.

Aside from Quindina, no reproductive studies have tackled other nomoclastid genera. Field observations did not find nests in the area of occurrence for Kichua rheimsaePinto-da-Rocha & Bragagnolo 2017, Callcosma abrapatriciaPinto-da-Rocha & Bragagnolo 2017 or C. gracilima Roewer 1932 (Cabra, Rheims, pers. comm.; Pinto-da-Rocha pers. obs.). However, the lack of observation of nests in these other species is not sufficient evidence of their absence, given that the respective field trips lasted only a few days. It is more likely that all or at least most of the species of the genus Quindina build nests, since this behavior has been observed in half of the known species. Furthermore, the ones that have no records (Q. marginata, Q. discolor, Q. bimaculata, Q. hermesi, Q. bella and Q. albiocularia) have few observations (Pinto-da-Rocha & Bragagnolo, 2017; Pinzón-Morales & Pinto-da-Rocha, 2020). The fact that we did not find the nestless species on the same branch, but on different branches (interspersed with species that do build nests) reinforces the idea that the absence of nest reports for these species is due to a lack of field observation.

We also call attention to the fact that 44 years elapsed between the description of Q. albomarginis (Chamberlin, 1925) and the publication of the first article describing open mud nests (Rodríguez & Guerrero, 1976). In the case of Q. limbata, 71 years separate the species description (Roewer, 1943) from the first published description of its nest (Rojas & Solano, 2014). The other three species with known nests were reported by Pinto-da-Rocha and Bragagnolo (2017). At this point, it is difficult to know if the pendular disc nest is an autapomorphy of Q. pendula sp. nov. or if it arose earlier, since we have no available data for either Q. marginata or the other nomoclastid genera. More knowledge of parental care in Q. marginata would help us clarify the origin of the nests in the genus Quindina, as this is the sister group for the remaining species of the genus. In the event that it builds a pendular disc nest, hypotheses about its origin in the most recent common ancestor of the genus or even in an ancestor external to the genus would be more plausible. On the other hand, if Q. marginata builds open mud nests, it would be more likely that the pendular disc nest is an autapomorphy of Q. pendula sp. nov. Finally, in the case that Q. marginata does not build nests, it would be more probable that the origin of the nests is in one of the internal groups of the genus Quindina.

This study enriches our knowledge on parental care behavior and nest construction in the genus Quindina. Until recently, only one type of nest had been reported for this genus in the scientific literature (Pinto-da-Rocha & Bragagnolo, 2017; Quesada-Hildalgo et al., 2019; Rojas et al., 2019) and now we know that there are variations in the shape and structure of the nests. Likewise, we can observe that our knowledge about the species within this genus, as well as for the whole Nomoclastidae family, has increased steadily in the last ten years (Kury et al., 2024; Kury & Villarreal, 2015; Pinto-da-Rocha & Bragagnolo, 2017; Pinzón-Morales & Pinto-da-Rocha, 2020; Quesada-Hidalgo et al., 2019; Requena et al., 2014; Rojas et al., 2019; Rojas & Solano, 2014; Villarreal & Kury, 2023). Beyond the genus Quindina and order Opiliones as a whole, the results obtained in this study are of great importance in our understanding of egg care and nest construction in arthropods. The construction of hanging nests with fragments of plant material has not been previously reported in other groups of arachnids or in any of the classes within Arthropoda, so our findings are relevant for any further studies about the evolution of parental care in this phylum.

Ethical statement: the authors declare that they all agree with this publication and made significant contributions; that there is no conflict of interest of any kind; and that we followed all pertinent ethical and legal procedures and requirements. All financial sources are fully and clearly stated in the acknowledgments section. A signed document has been filed in the journal archives.

See supplementary material

a53v72n1-suppl1

Acknowledgments

We express our gratitude to all the people who supported the development of this work: Nicolás Briceño, who assisted us with the collection of specimens; Manuel David Cortés who identified the bryophyte species; Osvaldo Villarreal, who provided us with the setk script and recorded the species and nests in the municipality of Tena (department of Cundinamarca); Christian Franco García, who recorded the pendular disc nest in the municipality of Saboyá (department of Boyacá); Daniela Ahumada, who helped us with the extraction and observation of the male genitalia; Conchita Pinzón, who offered initial feedback on the phylogenetic analysis and species description; Paula Nuñez, who aided in the elaboration of the illustrations; Víctor Rodríguez Saavedra, who registered the specimens in the collection of the Museo de Ciencias de la Universidad El Bosque MCUB; Cristina Rheims, who kindly reviewed the language of an early version of the manuscript; David Hernández, who kindly checked and reviewed the language for the final manuscript; and the anonymous reviewers, who provided their valuable feedback. We also give thanks to the Willi Hennig Society for providing us with the TNT v.1.5 software. The participation of Héctor Lancheros in this study was done within the framework of the “Ecological Interactions of Plants in Colombian Ecosystems: Phase I” project, from the Biology Research Group GRIB of Universidad El Bosque. This study was funded by CAPES, CNPq (306722/2018-6), FAPESP (BIOTA, 2013/50297-0), NSF (DOB 1343578), and NASA to Ricardo Pinto-da-Rocha.

References

Publication Dates

History