Zygoparity in Characidae – the first case of internal fertilization in the teleost cohort Otomorpha

Most teleosts are externally fertilizing, with internal fertilization occurring as a relatively rare event. Until now, Euteleosteomorpha is the only teleost cohort known to undergo internal fertilization. In the teleost cohort Otomorpha, it has been recorded the presence of sperm in the ovaries of some species of Characiformes and Siluriformes, but no fertilized eggs have been found so far in the female reproductive tract. It has been presumed that oocytes can be released into the water with associated spermatozoa and only there becomes fertilized, and the term insemination has been used to characterize the strategy adopted by these fish. Here, we present the discovery of the first case of internal fertilization in the teleost cohort Otomorpha, in Compsura heterura (Characiformes: Characidae). In the course of spawning, the eggs form the perivitelline space and the animal and vegetative poles within the ovaries, evidencing oocyte fertilization. The newly spawned eggs then continue to form the animal and vegetative poles and increase the perivitelline space. These eggs are in the zygotic stage. These data indicate that fertilized eggs are only retained for a short period, providing evidence that C. heterura is a zygoparous fish.


INTRODUCTION
Fertilization (sensu lato) refers to several steps leading to and resulting in sperm-oocyte fusion. These steps include oocyte activation, expulsion of the second polar body, cortical reaction, formation of the perivitelline space, hardening of the envelope (chorion) and bipolar differentiation (cytoplasmic movement). Most teleost fish have centrolecithal unfertilized oocytes at oviposition and become telolecithal only after activation or even later in the process of fertilization (Kunz, 2004).
Although no cases of internal fertilization have been shown thus far in the teleost cohort Otomorpha (Fig. 1), the presence of sperm has been verified in the female ovaries in some species of Characiformes (Kutaygil, 1959;Nelson, 1964;Burns et al., 1995;Burns et al., 1997;Malabarba, 1998;Burns et al., 2000;Castro et al., 2003;Weitzman et al., 2005;Javonillo et al., 2009;Quagio-Grassiotto et al., 2012), andSiluriformes (von Ihering, 1937;Loir et al., 1989;Burns et al., 2000;Meisner et al., 2000;Javonillo et al., 2009;Spadella et al., 2012). However, in all these cases, the exact moment of fertilization is unknown because fertilized eggs are not found within the ovaries (Burns et al., 1995;Burns et al., 1997;Azevedo et al., 2000;Burns, Weitzman, 2005;Oliveira et al., 2010). Furthermore, the presence of sperm in the ovaries does not prove internal fertilization because an oocyte can be released into the water with associated spermatozoa and only there becomes fertilized, which is known as internal gametic association (Munehara et al., 1989). Given this occurrence, the term insemination has been used to characterize the strategy adopted by these fish (Burns et al., 1997;Javonillo et al., 2009), and the exact place that fertilization takes place remains unknown in Characiformes and Siluriformes of the teleost Cohort Otomorpha.
The objective of this study is to investigate and describe in an inseminating characid when fertilization occurs. For this purpose, the Cheirodontinae characid Compsura heterura Eigenmann, 1915 was used as a model. This species is inseminating, showing spermatozoa with slightly elongated nuclei in female ovaries (Burns et al., 1997: Fig. 1e;Oliveira et al., 2010: Fig. 1-2). The description of how fertilization takes place in such a member of the Cohort Otomorpha may serve as a starting point for comparison and discovery of the reproductive strategy of other inseminating Characiformes and Siluriformes.

MATERIAL AND METHODS
Fish Sampling. Compsura heterura specimens were collected with a 1 mm-plastic mesh hand net in 2001 from the Maxaranguape stream (05°30'45.00"S, 35°19'13.97"W), Maxaranguape city, Rio Grande do Norte state (RN), Brazil. Individuals were maintained at Ichthyology Laboratory, UFRGS, into twelve 60-litre aquariums, at a maximum of seven fish per aquarium. The aquariums contained dechlorinated tap water, were equipped with an undergravel biological filter, were lined with approximately 5 cm of gravel (5 mm grain size), and were planted with Hygrophila difformis and Microsorum pteropus. The lighting consisted of 20 W white fluorescent lamps with a timer set to a photoperiod of 12 h: 12 h light-dark. Aeration was constant by submerged pumps (280 l/h). The temperature was maintained at approximately 25°C with thermostat and heater, in an air-conditioned room. Fish were fed with commercial food flakes once a day ad libitum. The offspring of this wild population was analysed in this study.
Eight spawning females were euthanized using an overdose of Eugenol and dissected for ovaries removal. Twenty-two newly spawned oocytes were collected adhered to plant leaves (< 1 min after spawn).

Light Microscopy.
Ovaries and oocytes were fixed by immersion in 2% glutaraldehyde and 4% paraformaldehyde in Sorensen's phosphate buffer (0.1 M, pH 7.2). The ovaries and oocytes were dehydrated in ethanol and embedded in historesin (Leica HistoResin). Serial sections (3 µm) were stained with periodic-acid-Schiff (PAS)/hematoxylin/metanil yellow (Quintero-Hunter et al., 1991). Voucher specimens were deposited in the fish collection of the Departamento de Zoologia at the Universidade Federal do Rio Grande do Sul (UFRGS 22200, UFRGS 22201, UFRGS 22202).  Ethical note. This study conforms to the principles outlined in the Guide for the care and use of laboratory animals (Garber et al., 2011) and was approved by the Ethics Committee on Animal Use of the Federal University of Rio Grande do Sul (Universidade Federal do Rio Grande do Sul) (Project n. 32291).

RESULTS
All spawning females had unfertilized oocytes in the follicle (follicular cavity) and fertilized eggs in the ovarian lumen (ovulated). All newly spawned eggs were fertilized. The observed oocytes (unfertilized) had a single micropyle and several cortical alveoli of different sizes at the peripheral region of the oocytes (cortical cytoplasm) ( Fig. 2A). In the ovarian lumen, eggs were observed at the first cytoplasmic movements that define the animal and vegetative poles, containing few cortical alveoli and forming the perivitelline space (Fig. 2B). In the gonoduct, the eggs decreased the number of cortical alveoli (cortical reaction), and there was a clear distinction between the animal and vegetative poles (bipolar differentiation, Fig. 2C). The newly spawned eggs formed the animal and vegetative poles, and there was an increase in the perivitelline space. In the animal pole, there were no blastomeres (embryonic cells). These eggs were still in the zygotic stage (Fig. 2D).

DISCUSSION
Oviparity is defined as the laying of fertilized or unfertilized eggs with intact eggshells or jelly coats in the environment (Blackburn, 2015). At the time of laying, the egg may be unfertilized (i.e. oocyte; Ovuliparity), in the first stages of development (e.g., zygote; Zygoparity) or even in advanced embryonic stages (Embryoparity). Viviparity has been defined as a process in which females retain developing eggs inside their reproductive tracts and give birth to their young (e.g., larvae; Blackburn, 2015). Most teleosts (nearly 97.5%) are externally fertilizing, whose gametes are released and fertilization takes place in the water (Ovuliparity).
Teleost fish have been often referred to as "internally fertilizing" based on behavioral records, possible copulatory organs (e.g., male fin modifications), presence of sperm in the ovary or egg/embryo observations under stereomicroscopy, but showing no details on the stage of egg development (Meisner, 2005). Such is the case of the Glandulocaudinae characids (now part of the Stevardiinae), initially treated as bearing internal fertilization based on the observation of spermatozoa within the ovarian cavity (Burns et al., 1995), but latter referred to inseminating based on the lack of fertilized eggs observed within the ovaries (Burns et al., 1995;Burns et al., 1997;Azevedo et al., 2000;Burns, Weitzman, 2005;Oliveira et al., 2010).
Here, in the ovaries of spawning females of C. heterura, the cytoplasmic movements, few cortical alveoli, absence of blastomeres, formation of perivitelline space and animal and vegetative poles in the eggs allowed us to infer that the spermatozoa fertilize the oocyte in the ovary and that this egg was in the zygotic stage (formation of the zygote or egg cell, before the first cleavage). This is the first case of internal fertilization in the teleost cohort Otomorpha. The zygotic stage of newly spawned eggs in C. heterura indicates that there is a short period of retention of the fertilized eggs. Fertilization likely occurs after ovulation, in the ovarian lumen, and shortly after the fertilized eggs are released into the external environment, indicating that C. heterura (Characidae -Characiformes) is a zygoparous fish.
Zygoparity has been recorded in species of the Cohort Euteleostomorpha, in which the egg retention varies from zygote to embryo (embryoparity), as for example in Kryptolebias marmoratus (Poey, 1880) Sebastolobus Gill, 1881, Sebastidae (Harrington, 1961Rosen, Bailey, 1963;Sequeira et al., 2003;Pavlov, Emel'yanova, 2013). In addition, Oryzias setnai (Kulkarni, 1940) -Adrianichthyidae -and Zoarces americanus (Bloch, Schneider, 1801), Zoarcidae, may be zygoparous because copulation and spawning egg in early stages of embryonic development have been observed in both species (Kulkarni, 1940;Pavlov, Emel'yanova, 2013). Among Characiformes, insemination is found only in Characidae sensu stricto (Malabarba, Weitzman, 2003;Oliveira et al., 2011;Mirande, 2018), but is observed in part of the species of all the three main lineages of this family (Javonillo et al., 2010;Oliveira et al., 2011): the Clade A, now Stevardiinae (44 inseminating species listed by Thomaz et al., 2015: Table II); the Clade B, in Cheirodontinae (all species of the tribe Compsurini listed by Malabarba (1998)); and the Clade C, in Stethaprioninae (species of the genera Hollandichthys and Rachoviscus listed by Quagio-Grassiotto et al., 2012). The presence in different lineages has evidenced the multiple origin of insemination in Characidae. The zigoparity of C. heterura (shown here) may be shared among all species of its tribe, but it must be tested. On the other hand, the timing and place of fertilization in the inseminating species of the subfamilies Stevardiinae and Stethaprioninae remain unknown. Since the origin of insemination in these groups is not homologous and arose separately from that of Compsurini, their strategies as inseminating ovuliparous, zygoparous or even embryoparous remains to be investigated.