Attachment structures
Attachment mechanisms are omnipresent in animals, including insects [38, 39, 40, 41, 42, 43, 44, 45, 46], molluscs [47, 48] reptiles [49, 50], amphibians [51, 52], mammals [53, 54] and fishes [26, 55, 56].
These mechanisms are diverse and can include interlocking structures, such as hooks, locks, clamps or spacers [40], wet and dry adhesion [57, 58], and/or suction cups [24, 26]. Depending on the attachment system, physical effects as friction, mechanical interlocking, muscular force, viscous forces, chemical bonding, capillary effects, van der Waals forces, and electrostatic forces are involved and can lead to permanent, transitory and temporary attachment time to the substrate [40, 59, 60].
With regard to the aquatic environment, two main attachment strategies, bioadhesive secretion or suction attachment, seem to be present as adaptation to the specific physical conditions [see reviews 60,61]. Glue-like bioadhesive secretions include complex mixtures of proteins, lipids and sugars and can be found in echnioderms, mussels, or barnacles. Suction attachment involves muscular contraction to generate pressure differences and can be found in cephalopods, some insect taxa, and fish. In some species, both mechanisms can be found, as in lottiid limpets or fish. Depending on the taxa, attachment is achieved by multiple points of interaction, as in Echniodermata or Cephalopoda, or by one single attachment point, as in limpets or fish [see review 61].
In fish, the attachment structure, i.e., the suction disc, represents a chamber of subatmospheric pressure, to create adhesion by suction to various substrates [7, 11, 18, 23, 25, 26, 55, 56, 62], which can even enable climbing vertical surfaces outside the water column [8, 20, 63, 64, 65].
Performance of the fish sucker depends on many different factors of the attachment structure itself, such as muscle contraction, kinematics, material properties, size, and shape [18, 24, 25, 31, 56, 66, 67, 68, 69, 70, 71]. Additionally, the fish attachment ability is affected by the intensity of the water stream [72, 73] and the substrate curvature and its surface properties [11, 22, 25, 26, 31, 72, 74, 75, 76, 77, 78, 79]. During attachment, the animals maintain their grip by friction with structures at micro- and nanoscale, i.e. papillae or microvilli [11,14,15,16,25,28,55,56,80,81; see reviews 29,30,32]. Additionally, the mucus between and on these structures provides strong contribution to the attachment strength [82].
The whole body of the fish can be modified as ventral sucker (Balitoridae), or the fins have convergently evolved to attachment structures (Gobiesocidae, Cyclopteridae, Liparidae, Echeneidae, Cyprinidae) [18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 79].
In Gyrinocheilidae, some Gobiesocidae, and suckermouth catfishes of Africa (Mochokidae) and South America (Loricariidae, Astroblepidae), the mouth structures are transformed to an oral sucker, allowing animals to adhere to surfaces while simultaneously foraging and performing respiration [6, 7, 11, 12, 13, 20, 24, 74, 83]. This specialized suckermouth is, similar to attachment structures of other fish, highly textured with papillae bearing small keratinized outgrowths of single epithelia cells, i.e., unculi [10, 13, 14, 15, 16, 84, 85, 86].
In general, attachment structures in fish have received attention in the last decades, which even enabled the development of adhesive materials or gripping and adhesive devices [56,69,71,78,82,87,88,89,90,91,92; see reviews 29,30,31,32,93].
However, in contrast to remoras, gobies, hillstream loaches, darters, or clingfish, which were experimentally studied with regard to attachment performance and mechanisms [18, 25, 26, 55, 56, 62, 73, 76, 77, 78, 79, 80, 94], to date only few experiments [11, 72] addresses the real attachment performance of the loricariid suckermouth fish.
We here aim at presenting the structural diversity of papillae and unculi types in Loricariidae. As here only 69 species were studied, we expect that more diversity can be potentially discovered, when more species are included. As comparative experimental data on species with different mouthpart morphologies is lacking, we can only hypothesize about their attachment capabilities. As however, attachment structures were well investigated in various animal phyla and basic principles of adhesion and interlocking are known, we can infer some functionality based on the morphology analysis together with material property estimations and propose hypotheses about the interaction of papillae and unculi with different kinds of substrate (Fig. 4). This, however, needs to be proven experimentally in the future, especially with regard to the strength of current and substrate properties. In addition to the micro- and nanostructures, the mucus covering the mouth apparatus is on expect to contribute to the contact formation and adhesion as well and therefore should be investigated deeply in the future, in the course of controlled experiments on model species with different unculi and papilla types.
The loricariid oral disc is composed of upper and lower jaw and is surrounded by a softer outer rim, which was found to make tight contact with the substrates during attachment [6]. This seems similar to the soft tissues surrounding the suckers of remoras, which conform to the local roughness and curvature of the substrate [95], or to the outer papillae, setae or microvilli (which are of smaller diameter and densely packed) of the clingfish and loaches [25, 26, 55, 56, 80, 81].
The fleshy lips of Loricariidae are highly variable with regard to morphology, size and the content of collagen, which was previously found to relate to the substrate and the flow [96]. The collagen probably reinforces the oral suction cups and reduces slipping, failure or buckling in streams with high flow velocities [96], while being manipulated and bolstered by the jaws and maxillary barbels [12]. The lips are covered ventrally by uniculiferous papillae [9, 10, 12, 13], which probably increase wet friction and hydrodynamic adhesion to reinforce the seal of the oral disc [9, 11]. The geometry and arrangement of papillae in other fish taxa were previously found to support the resistance to shear forces and to arrests cracks at the interface between suction cups and substrate, which would compromise the subambient pressure in the mouth chamber [25, 26, 55, 80, 81]. This mechanism is partially similar to segmented adhesive pads of insects [41, 46, 97].
The unculi, which can be found on top of the papillae, are potentially involved in feeding [10, 12, 17]. But additionally, they probably increase the friction/interlocking during attachment on rough substrates [6, 11, 13, 14, 17]. This, together with the mucus, increases attachment strength as in other fish taxa [18, 82, 98].
With regard to interspecific variation of papilla size and morphology in Loricariidae, there are huge lacks of knowledge. In the here examined species, we were able to recognize four papillae types (Fig. 1), which differ in their height and range of motion, and eight unculi types (Fig. 2). With regard to the systematic position of the species (Fig. 3), it seems that there are high levels on convergent evolution, which was previously also proposed for mandibles and body shapes in suckermouth armoured catfishes [2, 3]. High levels of convergences are also determined for foot adhesive pads in animals [46].
The flat papillae (detected in Pseudacanthicus pitanga, Ancistrus ranunculus, Crossoloricaria cephalaspis, Pseudohemiodon almendarizi, Loricaria luciae, L. simillima, Spatuloricaria puganensis and Pterosturisoma microps) seem to be composed of rather rigid material, which is embedded in the softer and more flexible lip (Fig. 4A). In some species (A. ranunculus, C. cephalaspis, P. almendarizi, L. luciae and L. simillima), no unculi were detected and the papilla surface seems to be rather bulky. During attachment, the papillae are probably capable of interlocking with rather the stiff and rough substrate (A. ranunculus), which is facilitated by the soft embedment. We, however, expect these species to underperform on stiff and smooth substrates and to hardly attach to soft substrates (plant covers), since contact areas are reduced (Fig. 4A). This could potentially explain, why this pattern is mostly found in species living on sand or mud (C. cephalaspis, P. almendarizi, L. luciae and L. simillima), where attachment will probably not play such an important role. Here, the species probably only temporarily attach to substrate (e.g., wood) and might not need a tight and more continuous attachment. However, thick mucus, covering the bulky surface, might compensate these shortcomings, which should be investigated in the future. However, in some species (P. pitanga and S. puganensis), these papillae are covered by flexible long and thin filaments or by small mushrooms. Here, we expect the unculi to adapt to the substrate increasing either adhesion (by filaments) or interlocking (by mushrooms) (Fig. 4E,G).
The short papillae were detected in Hypoptopoma inexspectatum, Scobinancistrus aureatus, Ancistrus sp. L464, Chaetostoma formosae, and Hemiloricaria melini. These papillae seem to have limited range of motion as well, but could potentially function as bolsters, when the unculi interact with the substrate, or support rearrangement during attachment, because the papillae tips seem to be flexible. They were either covered by unculi of the mushroom type (S. aureatus, C. formosae), hooks (A. sp. L464), or short mushrooms (H. melini). Here, unculi together with the flexible papillae could enable a tight interaction with the stiff and rough or with the soft substrate by interlocking and with the stiff and smooth one by adhesion (Fig. 4C,D,E). However, mucus could also be potentially distributed between the unculi and additionally support adhesion under water.
The medium sized papillae were detected in most studied species. Due to the length of the papillae we expect this type to have a higher range of motion, which presumably enables them to adapt to rather challenging surfaces. The bases of the papillae seem to be stiffer and the tips more flexible, and therefore we expect high attachment forces, as the flexible papilla tips (which make unculi bases flexible) can easily adapt to corrugated substrates and to interact with them. They were usually covered with unculi, either with mushrooms (Otocinclus cocama, Ancistrus sp. L519, A. sp. L107, A. cirrhosus, A. luzia, Chaetostoma dorsale, C. lineopunctatum, Lamontichthys filamentosus, Sturisomatichthys aureus, Farlowella platorynchus, F. oxyrryncha, Cteniloricaria platystoma), hooks (L. stibaros), folds (Panaqolus sp. L271, P. sp. L351, Panaque nigrolineatus, Scobinancistrus aff. pariolispos, Baryancistrus xanthellus L81, L177, B. aff. niveatus), long filaments (Pseudacanthicus spinosus, P. sp. L97, P. sp. L65, P. sp. L185, P. sp. L273, Sturisomatichthys festivus) or short mushrooms (Hypancistrus contradens, Hemiodontichthys acipenserinus). The species bearing mushrooms, hooks, or short mushrooms can probably interact with the stiff and rough substrates and the soft substrate (plants, biofilm, etc.) by interlocking and with the stiff and smooth substrate by adhesion (Fig. 4C,D,E). However, on stiff and smooth substrates, contact points might be reduced, since less unculi are in contact. For the species bearing long filaments, we expect a high degree of interlocking on stiff and rough or soft substrates and of adhesion on stiff and smooth substrate, as these soft structures seem flexible enough to establish contact on most surfaces (Fig. 4G). The unculi of the fold type are potentially rather used for establishing contact by interlocking, since their tips seem to be rather stiff (Fig. 4H). We expect this type to underperform on soft substrates; however, mucus between these structures might increase their attachment ability. In one species (Acanthicus adonis), we determined medium papillae with a suction cups surface pattern. Due to morphology and material property estimation, we expect this type to adhere tightly with stiffer surfaces (both smooth and rough), but underperform on soft substrates (Fig. 4B), similar to the disc margins of clingfish (Ditsche et al., 2014). Only few species (Rhinotocinclus isabelae, Leporacanthicus joselimai, L. sp. L240, Oligancistrus immaculatus, Parancistrus nudiventris, Hypancistrus sp. L174, H. zebra and Dekeyseria picta) did not have unculi but rather a bulky surface (Fig. 4A). For these species, we expect an interlocking mechanism to be present. In this case, the relatively flexible papilla bases probably adapt to the target surface and the stiffer bulky surface enables interlocking with the rough and stiff substrate.
Long papillae were detected in some species (Scobinancistrus sp. L82, Peckoltia sp. L76, Parancistrus aurantiacus, Hypancistrus sp. L333, Ancistrus claro, A. megalostomus, Aphanotorulus cf. emarginatus, Hypostomus bolivianus, H. laplatae, Pseudorinelepis cf. genibarbis L95, L152, Rhinelepis aspera). Because of their length, we expect these papillae to adapt best to very rough surfaces, due to an increased range of motion. On top of these papillae, we detected either mushrooms (S. sp. L82, A. claro, P. cf. genibarbis L95, L152), folds (P. sp. L76, A. cf. emarginatus, H. bolivianus, H. laplatae), bulky surface without unculi (P. aurantiacus), long filaments (H. sp. L333, R. aspera) or honey-combs (A. megalostomus). Since the mushrooms, long filaments and honey-combs seem to be flexible, we expect these structures to adhere to any surface either by interlocking or by adhesion (Fig. 4D,F,G). For the folds, which seem to be stiffer at their tips, we expect an underperformance on soft plant surfaces (Fig. 4H) due to their limited ability to adapt their shape to the target substrate, which would hinder the oral disc to form an effective seal, similar to gobies underperforming on rougher surfaces [22, 62]. The bulky surfaces (Fig. 4A) potentially also underperform on soft plant surfaces, due to the limited flexibility, and on stiff and smooth surfaces, because interlocking is here not facilitated. However, mucus might increase adhesion performance, which awaits further investigations in the future.