Phoretic association between Hisonotus chromodontus (Loricariidae: Hypoptopomatinae) and Ichthyocladius sp. (Diptera: Chironomidae) larvae in Amazonian streams

Larvae of non-biting midges in the family Chironomidae can be found in association with several species of fish in the family Loricariidae. In this study, we describe the first record of phoretic interaction between larvae of Ichthyocladius sp. and the fish Hisonotus chromodontus in streams in the Amazon basin. Between July 2010 and March 2019, fish were collected from three streams of the Teles Pires River basin in the state of Mato Grosso, Brazil. We investigated the attachment site of chironomid larvae on the body of fish and the frequency of attachment. A total of 1.241 specimens of H. chromodontus were collected, among which nine hosts were found carrying in the ventral region a single Ichthyocladius sp. specimen. All Ichthyocladius sp. were attached to a spicule of fish between the pectoral and pelvic fins, in which they were observed at the fourth developmental stage at the beginning of the cocoon construction (1.50 mm). In addition, a pre-pupa larva (2.60 mm) and an empty cocoon (2.50 mm) were reported. The Loanda stream had the highest occurrence of this type of association (of 12 fish examined, three had larvae). The other two streams reported a lower occurrence: the Baixada Morena stream had three fish associated with larvae amongst 1105 fish examined, whereas the Selma stream had three out of 124 fish infested by larvae. The occurrence of only one larva per host can be related to the small body area of the host fish (average standard length = 26.60 mm and average weight = 0.31 g). We observed that the larvae prefer fixation sites in the vicinity of the pectoral and pelvic fin of the fish, which is presumably associated with the fact that chironomids feed on algae and debris suspended by the movement of loricariids. These streams differ with respect to the presence of riparian forests, which may affect resource availability and thereby influence ecological interactions between the species.


Study area
Between July 2010 and March 2019, we collected fish from three first-to third-order streams, the Loanda (11°25ʹ33.1ʹʹS and 55°16ʹ39.3ʹʹW), Baixada Morena (11°29ʹ43.9ʹʹS and 55°21ʹ44.0ʹʹW) and Selma (11°36ʹ03.2ʹʹS and 55°25ʹ41.5ʹʹW), which are tributaries of the right bank of the Teles Pires River and differ among them in relation to preservation. The Loanda stream is characterized by an intact riparian forest, sandy substrate, and an absence of macrophytes (Cabeceira 2014), whereas the Selma stream lacks riparian forest and has areas of rapids, sand, and pebble substrate, and supports the growth of macrophytes. The Baixada Morena stream is also devoid of riparian forest in the area sampled and is considered to have undergone considerable environmental damage.

Fish collection and recording of fish-insect interactions
Fish were collected by two or three individuals using 0.5-mm-mesh nets. The collected specimens were euthanized with Eugenol anesthetic, fixed in formalin solution (10%), and subsequently transferred to 70% ethanol (modified from Mendonça et al. 2005). Voucher specimens were deposited in the fish collection of the Tropical Ichthyology Laboratory (LIT) of the Federal University of Mato Grosso, Sinop University Campus (voucher specimens with fixed larvae: LIT 3143, LIT 3144 and LIT 3145), and at the Ichthyological Collection of the Center of Research in Limnology, Ichthyology and Aquaculture (NUPELIA) of the State University of Maringá (voucher specimens: NUP 22680). The fish were weighed (g), measured (mm), examined for the presence of chironomid larvae, and photographed with the aid of Leica S9 stereomicroscope (×50 magnification) to record development (larval and pupal phases) (Trivinho-Strixino 2014) and the sites of attachment of phoretic organisms on fish.
The larvae were identified to the genus level by using morphological character-based identification keys presented by Trivinho-Strixino (2014). According to Trivinho-Strixino (2014) and Luiz Carlos de Pinho (personal communication) the identification of immature chironomid species (larvae and pupae), is only possible through the observation of the male genitalia adult.

Results
A total of 1.241 specimens of H. chromodontus, with mean standard lengths and weights ranging from 27 to 38 mm and 0.19 to 0.38 g, were collected. Among these, nine fish were found to be carrying a single Ichthyocladius sp. larva. The Loanda and Selma streams had the highest (of 12 fish examined, three had larvae) and the smallest (of the 1105 fish examined three fish were infested) occurrence of the interaction, respectively (see Table 1). All chironomids were observed in the ventral region between the pectoral and pelvic fins to the host (Figure 1 a), attached by fixing the distal part of the abdomen (prolegs) to the fish spicule, leaving the larval head free (Figure 1b). Attached Ichthyocladius sp. were observed at the fourth developmental stage at the beginning of the cocoon construction ( Figure 2 a and b). In addition, we found a pre-pupa larva (Figure 2 c) and an empty cocoon (Figure 2 d).

Discussion
We report here the first record of phoresy between insect-fish for the South of the Brazilian Amazon, involving a loricariid host with restricted distribution in the Tapajós River basin. Free-living larvae of chironomidae ( Figure S1) has limited dispersal capacity and the interactions with a fish can benefit it with no detrimental effects on their fish hosts. The phoresy recorded in this study represents a commensal interaction that benefits chironomid larvae (White et al. 2017, Silknetter et al. 2020. The findings of the present study, in which we detected only a single larva attached to each host, contrast with those of other studies that have reported means of 12.21 (Sydow et al. 2008) and 13.2 (Mattos et al. 2018) Ichthyocladius larvae associated with loricariid fish species. These studies have described loricariid hosts with relatively large body lengths of 104 mm (Sydow et al. 2008), 93 mm (Dala-Corte & Melo 2018), and 63 mm (Mattos et al. 2018), compared with the maximum length of 43 mm SL for the Hisonotus chromodontus specimens collected in the three streams we investigated. Sydow et al. (2008) examined the relationship between the total length of three species of loricariid (Ancistrus cf. multispina (104.2 mm), Pareiorhaphis hypselurus (62.5 mm), and Pareiorhaphis nodulus (37.6 mm)) and the quantity of associated chironomids. The larval density is higher in larger fish; for example, P. nudulus (a host with two larvae) had a mean length three times shorter than that of the other two species, as Pareiorhaphis hypselurus and Ancistrus cf. multispina, which had the density of larvae of 2.57 and 12.21, respectively. Similarly, Dala-Corte & Melo (2018) examined the distribution of Ichthyocladius larvae on the loricariid Pareiorhaphis hypselurus and found that smaller (36 mm) and larger (93 mm) fish had correspondingly small and large chironomid infestations, respectively. Large-sized hosts can carry a larger number of chironomid larvae on their body because they have a greater surface area, as well as a greater swimming range, thereby providing phoretic organisms with access to the larger amounts of suspended food particles (Sydow et al. 2008, Dala-Corte & Melo 2018. The Ichthyocladius larvae associated with fish gain advantages with respect to diet (continuous access to debris and algae available on the host body), mobility (avoiding unfavorable habitat sites), protection against predators (larvae are camouflaged by the host) and environmental disturbance (commensal species vulnerable to environmental disturbance exploit hosts to evade areas exposed to such disturbance) (Tokeshi 1993). As observed herein, larvae prefer attachment sites in the vicinity of pectoral and pelvic fin of fish, which is presumably associated with the fact that chironomids feed on algae and debris suspended by the movement of loricariids (Henriques-Oliveira et al. 2003, da Silva et al. 2008, Sydow et al. 2008, Mattos et al. 2018. Ichthyocladius larvae have thick posterior pseudopods bearing strong claws that facilitate attachment to the host, and construct cocoons shaped like fish scales attached to the spines of host fish fins (Trivinho-Strixino 2014). In this regard, it can be speculated that the symbionts have preferred sites for attachment on the surface of the host body and that there is competition for these sites. Alternatively, larvae may not show any site-specific preference and colonize all sites equally but remain attached to those sites that offer better resources or protection against predators (Dala-Corte & Melo 2018). The dispersal of an organism by phoretic interaction involves locating, fixing and detaching from the host (Bartlow & Agosta 2021). The success of this interaction is clear from our findings, as we found both larvae in the initial fixation phase and an empty cocoon trapped in the host.
Hisonotus species feed on periphyton, with a diet composed of diatoms, Chlorophyceae, and organic matter (Casatti 2002, Teresa & Casatti 2012, and numerous studies have shown that the availability of periphyton in streams is dependent on riparian coverage and may influence the growth of loricariid fish species with this type of diet (Bojsen & Barriga 2002, Teresa & Casatti 2012, Fernandes et al. 2013, Gomes et al. 2020). Therefore, it is possible that the degradation of the vegetation cover has a negative influence on the ecological interactions of the streams studied here.

Supplementary Material
The following online material is available for this article:   Figure S1 -Free-living larvae of Ichthyocladius sp. (indicated by arrow) fixed in a submerged log near two specimens of Hisonotus chromodontus in an Amazonian stream.