Monogenean fauna of alien tilapias (Cichlidae) in south China

Tilapias are important aquaculture fishes that have been introduced widely all over the world, often carrying their monogenean parasites with them. An extensive investigation on monogeneans of invasive tilapias was conducted in 19 natural water sources in south China between July 2015 and December 2017. We found nine known species of monogeneans, i.e., Enterogyrus coronatus, E. malmbergi, Cichlidogyrus cirratus, C. halli, C. sclerosus, C. thurstonae, C. tilapiae, Scutogyrus longicornis, Gyrodactylus cichlidarum, and one unknown Gyrodactylus species. In addition to reporting ten new hosts and four new geographical records, we observed new morphological characteristics of these species. Observation on living specimens of Enterogyrus spp. demonstrated that these two species have characteristic opisthaptoral retraction capacities, while the opisthaptor glands were not observed in our specimens of E. coronatus and E. malmbergi. The morphological differences of the accessory piece of the male copulatory complex between C. cirratus and C. mbirizei (character for species differentiation) could result from the observation at different perspectives, which indicates that C. mbirizei is likely a synonym of C. cirratus. A more detailed structure of the sclerotized parts of Cichlidogyrus spp. and S. longicornis were revealed by scanning electron microscopy. As was the case for the monogeneans found on alien tilapias from other geographic regions, the present study confirmed the high potential of these monogeneans to establish populations in new habitats.

As the country with the highest tilapias aquaculture production [16], China initially introduced Mozambique tilapia (O. mossambicus Peters, 1852) from Vietnam in 1956; other tilapias were then also introduced for culture or breed improvement [9,60]. In the course of tilapias culture, the escapees gradually established wild populations in many natural waters of south China [21], which has become a great concern because they will not only damage the aquatic community, but also act as a refuge for aquaculture pathogens. In addition, they could acquire native parasites ( [8,23], and our unpublished data) and may spillback the parasites acquired to the endemic hosts [24]. However, extensive investigation of monogeneans of tilapias was lacking, although several sporadic reports indicated the existence of the alien gill parasites on tilapias in China [28,37,58,59]. To fill this gap, an extensive investigation on the monogenean fauna of feral tilapias was carried out in south China between July 2015 and December 2017. The results presented in this paper include the monogenean fauna of tilapias and a supplementary description of new morphological features for three species.

Ethics
All the experimental handling was carried out in compliance with animal safety and ethics rule issued by the School of Life Sciences, Sun Yat-sen University.

Host and parasite collection
Investigation of wild tilapias and their monogenean fauna was conducted in 19 natural waters sources in south China. These sampling locations were selected based on field study and reports [21,37,59] to cover the distribution of wild tilapias in south China (Fig. 1, Table 1). Year-round investigations were implemented monthly in three reservoirs from April 2016 to August 2017: Nanshui reservoir (24°44 0 N, 113°1 0 0 E), Gaozhou reservoir (22°08 0 N, 111°05 0 E) and Songtao reservoir (19°24 0 N, 109°33 0 E) to cover the seasonality of infection (data about seasonality of infection unpublished). Fishes were identified by morphological features according to Fish-Base (www.fishbase.org). Nile tilapia Oreochromis niloticus and its hybrids were accepted as O. niloticus due to their indistinguishable morphologies. Oreochromis niloticus samples in Guangzhou, Guangdong Province were purchased from a local fish farm or caught from a small pond in the south campus of Sun Yat-sen University. These fishes were kept in the laboratory for observation of live parasite specimens.
Fish samples caught in the wild were individually killed and examined for parasites in the gills, stomach and urinary bladder. The parasite examination for the fish samples kept in the laboratory included the body surface. For identification, live monogeneans were detached with a dissecting needle, pipetted out, and mounted in a drop of ammonium picrate glycerin (GAP) on a slide under a coverslip, which was sealed using nail polish. After identification, target slides were rinsed in distilled water for 12-24 h until the nail polish could be easily removed, and the detached specimens were stored in vessels for further studies.

Light microscopy and scanning electron microscopy
For SEM studies, worms were processed for scanning electron microscopy (SEM) according to Mo and Appleby [38] or Fannes et al. [15], sputter coated with gold and finally examined under Quanta 400 (FEI, Netherlands) in the Instrumental Analysis & Research Center, Sun Yat-sen University. For light microscopy studies, identified worms were digested following the protocol depicted by Fannes et al. [15], and later remounted in GAP on slides. Or alternatively, specimens were rinsed several times with water before being stained in Modified Gomori's Trichrome, dehydrated in an ethanol gradient, cleared in clove oil, and finally mounted in neutral Canada balsam. For Gyrodactylus species, GAP preserved specimens were digested in situ: proteinase K solution was dripped on one side of the coverslip, while a piece of filter paper was placed on the opposite side until the GAP solution was entirely replaced by proteinase K solution; later the same method was used to replace the digestive fluid by GAP solution.

Morphological analyses
The numbering of the sclerotized parts of the species in genera Scutogyrus, Cichlidogyrus and Enterogyrus was adopted from ICOPA IV [14] and the terminology followed Pariselle and Euzet [49]. For Gyrodactylus species, the measurements of sclerotized parts and terminology followed Shinn et al. [54]. The terminology was employed as follows: anchor instead of gripus or hamulus; hooks rather than marginal hook, uncinulus or hooklet, and ventral bar instead of ventral transverse bar. Additionally, the ventral bar length represents the length of one branch rather than the whole length (with that of Gyrodactylus as an exception). The metrics are shown in Figure 2.
Photographs and measurements of the sclerotized parts were taken under an Olympus DX41 microscope equipped a DP73 Olympus camera (Olympus, Japan), and processed by the software cellSens Standard 1.7.1. Illustrations were drawn freehand with the aid of an Olympus U-DA drawing attachment and then digitized and processed using Adobe Illustrator CS6 (Adobe, USA). All measurements were taken in micrometers and presented in the following order: mean ± standard deviation (minimummaximum, number of measurements). Voucher specimens were stored in the Research Center for Parasitic Organisms, School of Life Sciences, Sun Yat-sen University (SYSU) and a set of whole-mount specimens was also deposited in the Muséum National d'Histoire Naturelle, France (MNHN).
In the eight locations (Baise, Heyuan, Liuzhou, Maoming, Nanning, Quannei, Quanzhou and Xishuangbanna) where E. coronatus coexisted with the hosts O. niloticus and C. zillii, C. zillii was always found with E. coronatus infection, while O. niloticus was only found to be infected in Liuzhou (see Table 1). Especially in Maoming, where both O. niloticus and C. zillii were monthly sampled for year-round, E. coronatus was exclusively collected from C. zillii (prevalence: 32.2%; mean intensity: 2.1). In the whole investigation, E. coronatus was not collected from S. galilaeus and O. mossambicus. This species had previously been reported from Tilapia guineensis (Coptodon guineensis) [50], Tilapia dageti (Coptodon dageti) [34], and Pseudocrenilabrus philander philander [30]. In a word, E. coronatus shows host preference to C. zillii in China and it possesses the potential to infect other cichlids. The occurrence of this species in the stomach of O. niloticus and C. zillii from China provides new localities and new host records.
Remarks: The morphologies and measurements of the sclerotized parts of our specimens basically agree with the previous description of E. malmbergi, but are slightly larger [7], probably influenced by environmental conditions [11]. The in situ observation of E. malmbergi found that the wound in the stomach of hosts was larger than those caused by E. coronatus [31] and the persistence of infection will enlarge the wound. No opisthaptor glands were observed in this species, which was consistent with descriptions of other Enterogyrus species [3, 5-7, 30, 50] except E. cichlidarum which was ever described with opisthaptor glands [42]. In addition, E. malmbergi has a much less retractable opisthaptor (only cup-shaped) than E. coronatus (Figs. 4, 6).
In the ten sampling sites (Baise, Chengmai, Chaozhou, Danzhou, Liuzhou, Maoming, Nanning, Quannei, Quanzhou and Xishuangbanna), where E. malmbergi coexisted with the hosts O. niloticus and C. zillii, this parasite was always collected from O. niloticus, but not from C. zillii except in Maoming where both O. niloticus and C. zillii were found to be infected (see Table 1). Even in Maoming, year-round investigations revealed that E. malmbergi had much higher infection levels in O. niloticus (prevalence: 16.4%; mean intensity: 1.4) than in C. zillii (prevalence: 1.4%; mean intensity: 1.1). In addition, E. malmbergi was also sampled from S. galilaeus and O. mossambicus in the present study, and had previously been reported from Cichlasoma callolepis [23]. These results indicate that E. malmbergi presents host preference to O. niloticus in China, but has the potential to infect other cichlids. It is the first record of E. malmbergi in China, and with C. zillii, S. galilaeus and O. mossambicus as new host records.
Remarks: Cichlidogyrus cirratus was first described by Paperna [43] from the gills of Tilapia galilaea (Sarotherodon galilaeus) in Lake Tiberias, Israel, and redescribed by Ergens [13] from the gills of Tilapia nilotica (Oreochromis niloticus) in River Nile, Egypt. In the present study, C. cirratus was collected from O. mossambicus (new host), O. niloticus and Coptodon zillii, but was not found in its type host (S. galilaeus).
The long winding penis and the short first pair of hooks differentiates C. cirratus from all other congeneric species from cichlid hosts, except C. mbirizei Muterezi Bukinga et al., 2012 [39,49]. The measurements of C. cirratus were consistent with those of C. mbirizei, and the latter was distinguished from C. cirratus in the original description by the shape of the accessory piece of the male copulatory complex (C. mbirizei without long expansion at mid-length and with two ends of rounded outgrowth versus C. cirratus with long expansion and hooked ends) and the vagina (double pitch in C. mbirizei versus sinuous in C. cirratus). However, the thin, long and transparent expansion in the middle of the accessory piece of C. cirratus was variable due to the different perspectives (Fig. 8). Moreover, two types of accessory piece extremity (hooked versus rounded outgrowth) could transform in digested specimens, when the accessory piece turns over (Fig. 8). These morphological features suggest that C. mbirizei and C. cirratus are likely synonymous. Pending genetic study, these two species are kept valid in the present study.
Cichlidogyrus cirratus (or C. mbirizei) was also recorded from non-native tilapias in Malaysia and Thailand [27,29]. In addition, Cichlidogyrus sp. (named as C. bananensis by Xiao [59]) found in Lancang River in China was likely a misidentified C. cirratus as it shows close morphological and morphometric similarities to the latter, which was also collected from the same locality in the present study. Voucher specimens observed and deposited: 31 (SYSU-CHA1-30; MNHN HEL901).
Remarks: The morphologies and measurements of voucher specimens in the present study agree with the previous descriptions of G. cichlidarum which was firstly described by Paperna [44] in Ghana and redescribed by García-Vásquez et al. [20]. This species had also been reported from non-native tilapias in the Philippines [2] (G. niloticus was synonymized with G. cichlidarum [20]) and Mexico [46]. Remarks: Although only one specimen has been collected, its characteristics of dorsal bar (with two protuberances) and ventral bar (with two large rounded auricular processes) made it resemble Gyrodactylus yacatli García-Vásquez et al., 2011 [19], which was first described from the gills and fins of O. niloticus cultured in Mexico and also from the fins of O. niloticus and Pseudocrenilabrus philander in Zimbabwe [61]. The marginal hook sickles of the present specimen are different from former descriptions (i.e., with a larger angle) [19,61]. However, the drawings of dorsal bar in these descriptions were not consistent; the former had a straight dorsal bar but the latter possessed a dorsal bar with two protuberances. Our specimen was more like the description of Zahradníčková et al. [61], but could not be definitively identified.    [3,[5][6][7]49]. These species were initially described with two different opisthaptor features (cupor tongue-shaped), which were used for the division of Enterogyrus into two groups [6,30,50]. However, the results in the present study, based on the observation of live worms of E. coronatus and E. malmbergi in situ, did not support this hypothesis. Enterogyrus coronatus presented a variable opisthaptoral shape during anchoring on the stomach wall (opisthaptor tongue-shaped) and shifting from one location to another (opisthaptor cup-shaped). The opisthaptor of E. coronatus was more variable than that of E. malmbergi which could only present a cup-shaped opisthaptor. In addition, the opisthaptoral sclerotized parts (posterior hooks (pairs I and II), ventral anchors and ventral bar) of E. coronatus were discernibly slenderer than those of E. malmbergi. This might facilitate the extension of the opisthaptor and its penetration into the stomach wall. Pathologically, E. malmbergi caused larger wounds than E. coronatus [31] in the stomach of the host, which might be ascribed to the larger body size and less extensible opisthaptor. This inference needs further confirmation by comparing the pathologies of other Enterogyrus species with slenderer opisthaptor sclerotized parts (E. cichlidarum, E. melenensis, E. barombiensis, E. foratus and E. amieti), and that of E. crassus which possesses larger opisthaptoral sclerotized parts.

Discussion
The identification of Cichlidogyrus species was primarily based on the two-dimensional morphologies of the sclerotized parts in the whole-mount specimens, e.g., GAP preserved specimens. Based on the three-dimensional morphologies of the accessory piece terminal of the male copulatory complex of C. cirratus, considerable change was detected as a result of different view angle in the present study. This hints that threedimensional morphologies of isolated sclerotized parts by modern technical methods such as laser scanning confocal fluorescence microscopy can provide more comprehensive information for taxonomic studies [18,53].
The introduction and spread of non-indigenous tilapias could be associated with the introduction and spread of their parasites, but the parasite species richness often decreased in comparison with that in their native range [55]. For example, it was reported that the monogenean species of O. niloticus numbered 18 in its native range [26,48], while in the introduced areas they numbered between 1 and 7 although the species presented were similar [1, 2, 21, 22, 26-28, 32, 44, 50, 54, 56]. The similar monogenean species composition might be related to the similarity of tilapia strains cultured in different areas, e.g., genetically improved farmed tilapia (GIFT), which was widely introduced and cultured around the world. It was ever reported once that the monogenean species were completely lost as a result of tilapia introduction [17]. In the present study, the monogenean fauna of tilapias also shows different species loss in the different locations (see Table 1), e.g., only C. tilapiae was found in two sites (Haikou and Gengma) and even no parasites in Macau.
This study reported ten new host records of several monogeneans on tilapias (E. coronatus from Coptodon zillii and O. niloticus; E. malmbergi from Coptodon zillii, S. galilaeus and O. mossambicus; C. halli from Coptodon zillii and O. mossambicus; C. thurstonae from Coptodon zillii; C. cirratus from O. mossambicus; C. sclerosus from S. galilaeus), which demonstrated their lower host specificity. However, the host specificities of these species were basically consistent with previous reports [35,51]. In addition, the distinct host preference of Enterogyrus species (E. malmbergi prefers to infect O. niloticus; E. coronatus prefers to infect Coptodon zillii), together with the preferences of Cichlidogyrus and Scutogyrus species to O. niloticus in the present study, could be ascribed to the considerable intergeneric and parental care behavior differences of hosts [35].
In China, Cichlidogyrus levequei Pariselle & Euzet, 1996 was previously reported from O. niloticus [28], but it was not collected in the present study. However, the existence of this species in China is really doubtful because authors could not provide specimens and the description was too simple to judge the species [28]. In addition, C. levequei was recorded to be specific to the host Coptodon coffea which is endemic in West Africa [48].