All spiders are known to produce silk for the whole life and for realization of different vital functions (Foelix, 1996; Craig, 1997; Craig et al., 2000; Craig, 2003; Krafft and Cookson, 2012). Correspondingly, spiders possess several kinds of abdominal silk glands and up to six types of silk of different morphology, amino acid content, and protein structure (Foelix, 1996; Craig et al., 2000; Blamires et al., 2017). Representatives of other arthropod groups belonging to different evolutionary lineages, such as insects, crustaceans, pseudoscorpions, and mites, which produce silk, reveal only one type of silk glands synthesizing silk for a quite local and specific functional task (e.g., Sutherland et al., 2010; Annamalai and Jayaprakash, 2012; Kronenberger et al., 2012; Kakui and Hiruta, 2014; Büsse et al., 2015; Osborn Popp et al., 2016). Consequently, both glands and silks in arthropods are expected to be non-homologues because they are thought to have developed many times in the course of evolution of different groups (Sehnal and Akai, 1990; Craig et al., 2000; Young and Merritt, 2003; Sutherland et al., 2010; Ashton et al., 2011; Büsse et al., 2015). At the same time, based on the molecular data, it was hypothesized that “the spider orb web evolved only once” in the evolution (Blackledge et al., 2009).

Silks in arthropods and, in particular, in spiders are mostly built of fibrillary proteins (Kovoor and Zylberberg, 1980, 1982; Stubbs et al., 1992; Gosline et al., 1999; Sponner et al., 2005; Yonemura et al., 2006; Sponner et al., 2007; Blackledge et al., 2009; Blackledge, 2013; Arakawa et al., 2022) but, owing to their different origin, show quite different composition and structural properties (Vollrath et al., 1996; Gosline et al., 1999; Craig et al., 2000; Vollrath, 2000; Vollrath and Knight, 2001; Bakker et al., 2006; Schniepp et al., 2023). On the fine microscopic level, it was shown that generally each individual silk fiber might consist of an envelope, variously expressed, and of a core inward substance containing various fibrils differently organized (Kovoor and Zylberberg, 1980; Stubbs et al., 1992; Li et al., 1994; Vollrath et al., 1996; Sponner et al., 2007; Riekel et al., 2019, 2020; Blamires et al., 2023; Perera et al., 2023). Silk threads reveal a heterogeneous protein composition (Stubbs et al., 1992; Wang et al., 2022) and the presence of lipids in the envelope (core-skin structure) (Sponner et al., 2007). The expression of these two components of the spider silk fibers and their composition varies significantly in different spider species (Gould et al., 1999; Knight and Vollrath, 2002; Hajer et al., 2009). There is no doubt that silk in its different forms and composition, providing important functions, significantly increases the vital resources of arthropods (Alberti and Ehrnsberger, 1977; Manson and Gerson, 1996; Hajer et al., 2009; Clotuche et al., 2011; Kronenberger et al., 2012; Kakui and Hiruta, 2014; Hatano and Nagashima, 2015; Blamires et al., 2017).

Silk of the water spider Argyroneta aquatica (Clerck) was investigated previously by SEM methods with respect to diving bell and egg cocoon (Bakker et al., 2006). The authors recognized four types of silk threads in this species–threads of the diving bell, anchor threads, walking threads, and threads of the egg cocoon. Only threads of the first type show a diameter bigger than 1 µm; threads of other types seem to be much delicate and thinner (Bakker et al., 2006). The structure and function of the “attachment discs” in this species, providing the attachment of silk threads to substrates under water, were studied in detail recently (Schaber et al., 2023). The attachment discs, with special regard to piriform silk, was also studied in representatives of four aerial spider families using TEM methods (Wirth et al., 2019).

Surprisingly, the phenomenon of silk production was discovered in water mites (Acariformes, Hydrachnidia) (Shatrov et al., 2014, 2016; Shatrov et al., 2023) as related to the activity of their dermal glands (Shatrov, 2013; Shatrov and Soldatenko, 2016, 2022)—an acquisition and synapomorphy of freshwater mites after their ancestors have come down to fresh waters ca. 235 MYA (Dabert et al., 2016). The silk of Limnochares aquatica (L.) (Limnocharidae) and Limnesia maculata (O.F. Müller) (Limnesiidae) and some other water mite species was examined using both SEM and TEM methods. It was shown that silk in these two species is represented by only one type of threads quite similar in their organization (Shatrov et al., 2014, 2016; Shatrov et al., 2023). These threads are extremely long, unbranched, uniform tubes of around 1 to 2 µm in diameter containing a core substance (residual substance) or without it. In the latter case, silk threads are flattened on the substrate, as observed by SEM, resembling an elastic ribbon. With this character, silk threads of water mites are similar to ribbon-like threads of sicarid spider Loxosceles laeta (Nicolet) (Knight and Vollrath, 2002; Schniepp et al., 2023). In both cases, these ribbon-like threads are able to stick together either by van der Waal’s forces (Knight and Vollrath, 2002) or by the adhesiveness of the thread surface itself which bears fine fibrils of the thread’s wall rising vertically (Shatrov et al., 2023). This feature seems to give a possibility to web function as an elastic net (Knight and Vollrath, 2002).

In terrestrial mites, the production of silk is known in representatives of several families (Tetranychidae, Eriophyidae, Camerobiidae, Cunaxidae, and Bdellidae) (Wallace and Mahon, 1972; Bolland, 1983; Gerson, 1985; Alberti and Coons, 1999; Shin et al., 2006). Silk in these mites provides many important vital functions, such as in prey capturing, communication, migration, nest buildings, and others (Manson and Gerson, 1996; Clotuche et al., 2011; Kanazawa et al., 2011; Le Goff et al., 2011; Fernandez et al., 2012; Yano, 2012). Apparently, the same role is played by silk in other arthropods. Unfortunately, among many sophisticated methods applied for the study of the silk structure and composition in spiders and other arthropods, TEM methods were used only sporadically (Kovoor and Zylberberg, 1980; Stubbs et al., 1992; Li et al., 1994; Vollrath et al., 1996; Sponner et al., 2007; Wirth et al., 2019; Perera et al., 2023).

The main purpose of this study is to examine the ultrastructural organization of silk in the water spider Argyroneta aquatica (Clerck, 1757) (Araneae, Cybeidae) in comparison with that of some acariform mites, both water mites Limnochares aquatica (L., 1758) (Limnocharidae) and Limnesia maculata (O.F. Müller, 1776) (Limnesiidae), studied previously (Shatrov et al., 2016, Shatrov et al., 2023) and terrestrial aerial tetranychid two-spotted spider mite Tetranychus urticae (C.L. Koch, 1836) (Tetranychidae) studied previously only by SEM (Gerson, 1985).

The water spider A. aquatica has a Palearctic distribution and the only species from the order Araneae living underwater (Masumoto et al., 1998; Bakker et al., 2006). The spiders maintain an air layer around the abdomen and thorax by hydrophobic setae (Schaber et al., 2023). This air volume is used for the construction of the diving bell and spinning of dragline and piriform silk with the help of the attachment discs (Bakker et al., 2006; Schaber et al., 2023). The spiders construct the egg cocoon (egg sac) inside the diving bell from silk, which was investigated previously by SEM methods (Bakker et al., 2006). The authors divided this silk into four categories, and the diving bell, in turn, was found to consist of three different types of threads. Threads of these types correspond well in size to threads found in our study, in particular, to types a, b, c, and d used apparently as a capturing net. Probably, threads of the same origin in the spiders may be used for different purposes—building of the egg cocoon and formation of the capturing net. In the recent work, functioning of the attachment discs in A. aquatica was considered in tight association with the secretion of both dragline and piriform silks and the piriform adhesive cement (Schaber et al., 2023). In light of these data, it is clearly seen that the a- and, probably, e-types in our study closely correspond to dragline silk; the b-, c-, and d-types are directly associated with different types of piriform silk threads, whereas the amorphous substance is nothing more than the piriform adhesive cement (Schaber et al., 2023). Nearly the same mechanism of attachment of silk threads to various substrates, with the help of the glue-coated piriform nano-silk fibers due to the action of disk-like structures, was also shown for several species of terrestrial spiders (Wirth et al., 2019).

Despite the fact that, in nature, A. aquatica “spends its entire life cycle underwater” (Schaber et al., 2023), it rests for some time within the diving bell (Masumoto et al., 1998). Our observations show that the spider spends some time within the restricted air capacity above the water, the probable analogue of the diving bell. This resting outside the water most likely is just because of the fact that the container was maintained without any additional structures where the spider could fix its diving bell to. Importantly, that the spider produced web only periodically as well as constructed diving bells (Masumoto et al., 1998) as an alternate activity with periods of long rest.

Although it seems nothing is known about the formation and structure of a capturing net in this spider, it is apparent, however, that the spider constructs the web from the same threads used to build the egg cocoon. Nevertheless, the assortment of threads used for the formation of the egg sac (Bakker et al., 2006) appears to be wider than that used to produce the capturing net. In this regard, it is nearly possible to assume that (i) the spider uses different kinds of threads for every case and (ii) the formation of a kind of a capturing net in our observations is only the imitation of producing of the type of the egg cocoon in the absence of any favoring materials like wooden sticks, etc. Although we tried to exclude any hampering factors, like plants, substrate particles, etc., from our work with silk, in samples of silk associations examined in TEM, many additional particles of unknown origin were observed. However, the amorphous substance sometimes indicated in SEM most likely is a piriform adhesive glue secreted together with the piriform silk.

Morphologically, silk threads of the water spider A. aquatica are infinitely long, uniformly wide, and unbranched. However, threads of the a-type show a wide spectrum of their diameter and internal organization. In contrast with the “walking threads” (Bakker et al., 2006), which “are branched off into several fibers as soon as the substrate is reached” (p. 140), we did not observe branching of any threads even if they were detached from the walls of the container. In contrast to a- and e-type threads revealing a variously expressed internal structure, threads of the other types show a quite simple organization, probably possessing fine longitudinal fibers inside. Importantly, the a-type threads or their portions, which are free from the internal core substance, were not seen flattened on the substrate in SEM in contrast with the similar threads of the water mites with a significantly larger diameter (Shatrov et al., 2014; Shatrov et al., 2016; Shatrov et al., 2023).

The comparison with the studied silk threads of terrestrial spiders (e.g. Stubbs et al., 1992; Li et al., 1994; Vollrath et al., 1996; Gosline et al., 1999; Gould et al., 1999; Sponner et al., 2007; Hajer et al., 2009) does not reveal an exact morphological similarity between the studied silk threads of water and terrestrial spiders. The core substance in terrestrial spiders may be built of “pleated fibril-like structures, which are arranged in two concentric cylinders” in a coaxial manner as in Nephila clavipes (L.) (Li et al., 1994; Vollrath et al., 1996) or of a bundle of parallel nanofibrils as in Theridiosoma gemmosum (C.L. Koch) (Hajer et al., 2009; see also Blamires et al., 2017). In the latter case, the egg cocoon is constructed from several thread types of quite different diameter values from 600–800 nm up to 4–6 µm. Alternatively, in the cobweb scaffolding threads of Latrodectus hesperus (Chamberlin et Ivie) (black widow), thicker fibers (300 nm) oriented parallel to the thread axis and thinner ones (10–100 nm) oriented across the thread axis were identified in the highly ordered regions of the silk threads (Gould et al., 1999). In Argiope aurantia Lucas, a core substance of threads of the egg case consists of numerous electron-light fibrils immersed into the electron-dense matrix (Stubbs et al., 1992).

The envelope morphology of the spider silk threads disclosed previously in a number of works (e.g. Stubbs et al., 1992; Sponner et al., 2007) attracted lesser attention than that of the core substance and is completely different by its organization to that revealed in A. aquatica—for example, the wall of the thread fibers may be nearly imperceptible, like in Argiope (Stubbs et al., 1992), or may consist of a number of thin layers of differently expressed spidroins, as occurs in Nephila clavipes dragline silk (Sponner et al., 2007). The adaptations to different ecological conditions and feeding diets probably greatly influence on the spider silk and its morphology and properties.

The recent studies of the spider silk ultrastructure based on different methodological approaches revealed a quite complex nanostructure of the spider silk fibers (e.g. Schniepp et al., 2023; Blamires et al., 2017; Riekel et al., 2019; Wirth et al., 2019; Riekel et al., 2020; Wang et al., 2022; Blamires et al., 2023; Perera et al., 2023). In particular, it was shown that the spiders’ dragline fiber consists of a protein core of tightly packed nanofibrils within a protein matrix (Blamires et al., 2023). The proposed dragline silk threads in the present study show a quite different organization in comparison with the above-mentioned ones. Undoubtedly, nanofibers of the external fibrillary wall may provide both adhesive properties and extension of fibers, whereas the internal core may contain greatly coiled protein fibers, which may stretch and thus reveal the high tensile strength of these threads (Foelix, 1996; Blamires et al., 2017; Blamires et al., 2022). At the same time, it is hardly possible to suspect that exfoliation of these threads could give multiple fine fibers as it is shown in some spiders (Perera et al., 2023) or even a more complicated internal structure (Riekel et al., 2019, 2020). In any case, it may be disclosed by methods that are more sophisticated like AFM tomography, X-ray diffraction, gas chromatography, and others (e.g. Schniepp et al., 2023; Riekel et al., 2019, 2020)—for instance, the b-, c-, and d-type threads of A. aquatica are supposedly composed of longitudinally oriented nanofibers like in Trichonephila clavipes (L.) (Perera et al., 2023), although in the latter species these threads could be of a much greater diameter (up to 4 µm). Importantly, the surface of the ribbon-like silk threads in Loxosceles spider reveals fine papillae supposedly contributing to their adhesive properties (Schniepp et al., 2023) that was disclosed by AFM tomography and SEM. Conversely, in A. aquatica, the surface of different silk threads may show only fine striation that may be provoked by some contraction during the sample preparation.

Silk of the water-inhabiting caddisflies (Trichoptera) investigated thoroughly by different approaches (Engster, 1976; Yonemura et al., 2006; Ashton et al., 2011; Hatano and Nagashima, 2015) differs substantially from that of the water spider A. aquatica despite inhabiting similar aquatic environments. First, “trichopteran larval silk is extruded from a pair of silk glands that fold back and forth several times along the ventral side of the alimentary canal and terminate at the spinneret on the ventral side of the mouthparts and alimentary orifice” (Ashton et al., 2011). To form the final thread, the initial silk fibers are fused into a paired elliptical fiber embedded in a layer of glue proteins with adhesive properties (Yonemura et al., 2006). Although these silk fibers are principally proteinaceous, the peripheral layer also shows the presence of both neutral and acidic polysaccharides (Engster, 1976). These composite silk threads of caddisflies are used for two purposes—constructing both retreats and cocoons for larvae and capturing nets to filter food particles (Hatano and Nagashima, 2015). The composite silk of the water spider differs by morphological criteria from that of Trichoptera but quite similar to the latter by their functions.

Another large water-inhabiting arachnid group, water mites, reveals a partial similarity in the silk threads’ organization. The a-type threads of A. aquatica and silk threads of water mites (Shatrov et al., 2014, 2016; Shatrov et al., 2023) are nearly identical by their morphology but are slightly different in size. Importantly, the fibrillary walls in both cases may point to the high adhesive properties of the threads that help to form, owing to their sticking together, a rather strong capturing net. The core substance, characteristic of this type of silk, until fibers are found in it, may play a stabilizing and solidifying role in the thread composition. The latter may indicate a high tensile strength of this kind of threads, which is highly characteristic for the spider silk (Foelix, 1996; Blamires et al., 2017).

Threads of other types, except for the e-type with the simplest organization, in the light of available data, are hard to be directly compared with any known morphology of silk threads in spiders. Nevertheless, due to the possible presence of fine fibers inside, these threads are also capable of stretching. Therefore, owing to their abundance, they possibly play an important role in net construction. The fifth, the largest, and the fewest e-type thread of A. aquatica silk is rarely observed and probably play a general constructive role in the thread association closely resembling thick dragline threads. These threads are comparable with the Nephila dragline silk, where the micro-fibril walls are crossing under certain angles and may move along themselves as well (Vollrath et al., 1996). This type of threads, in contrast to others, is obviously capable of stretching.

Silk of the terrestrial tetranychid, the so-called spider mite T. urticae, demonstrates great simplicity in the organization of its threads (e.g. Gerson, 1985; Shin et al., 2006). This type of silk is highly comparable with the b- and c-types threads of A. aquatica by the dimensional characteristics and the possible fine-fibrillary contents. Despite such simplest morphology, these mites produce this silk in great abundance and use it in quite different ways (Clotuche et al., 2011). Apparently, contrary to the assumption that some insects “produce only one type of silk, and, in general, they use it for one purpose during a limited period in their lives” (Craig et al., 2000), spider mites use this kind of silk, uniform and simply organized, as wide as possible for the realization of different living-related purposes.

The water spider A. aquatica produces a compound silk consisting of several types of threads, which are comparable on one side with those of water mites and on the other side with those of aerial tetranychid mites. This finding shows that, in different evolutionary lineages of arachnids, a small number of possible variations limit the spectrum of available ways to produce silk proteins packed in threads of various morphology. On the other hand, the same water environment, in the cases of water mites and spiders, may provoke a similarity of the silk threads in both formation and morphology. At the same time, water spider produces silk resembling that of aerial mite that point to a wide range of its adaptations to different environmental conditions. However, the sources of these silks are different—if in water spider and water mites the glands producing silk are abdominal (Shatrov et al., 2019; Shatrov et al., 2023), in tetranychid mites these glands are prosomal (e.g., Gerson, 1985), i.e., of a quite different origin.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

The manuscript presents research on animals that do not require ethical approval for their study.

AS: Conceptualization, Data curation, Investigation, Methodology, Project administration, Software, Supervision, Writing – original draft, Writing – review & editing. ES: Data curation, Investigation, Methodology, Software, Visualization, Writing – original draft.