{"paper_id":"61648c28-3b1c-4022-b19c-8b422d48c314","body_text":"Reproduction of marble-mouth frogfish Lophiocharon lithinostomus (Lophiiformes, 1 \nAntennariidae) and the evolution of parental care among frogfishes 2 \n 3 \nAuthor names 4 \nToshiaki Mori 1* , Risa Murai 1, Takeshi Ito 2, Seiya Okuno 2, Yuya Kobayashi2, 5 \nTomoyuki Uehara 1 and Shun Satoh 2,3 * 6 \n 7 \nAffiliation 8 \n1 Marine Science Museum, Fukushima (Aquamarine Fukushima), 50 Tatsumi-cho, 9 \nOnahama, Iwaki, Fukushima, 971-8101, Japan 10 \n2 Department of Biology and Geosciences, Graduate School of Science, Osaka City 11 \nUniversity, Sumiyoshi, Osaka, 558-8585, Japan 12 \n3 Department of Evolutionary Studies of Biosystems, The Graduate University for 13 \nAdvanced Studies, Miura, Kanagawa 240-0193, Japan 14 \n* Corresponding authors 15 \n 16 \nRunning head 17 \nReproduction of frogfish Lophiocharon lithinostomus  18 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nAbstract 19 \nHere, we observed the reproductive behavior of marble -mouth frogfish ( Lophiocharon 20 \nlithinostomus) and the morphology of newly hatched juveniles under captive conditions. 21 \nAdult males showed pursuit behavior towards females approximately seven days before 22 \nspawning; spawning and ejaculation took place almost simultaneously. An adult female 23 \ncared for a fertilized egg mass adhered to their right of the left side of the flank. The 24 \nposition of the adhered eggs on the flank was left-biased (3 on the right and 25 on the 25 \nleft). The females exhibited a proactive fanning behavior towards the egg masses using 26 \ntheir dorsal fins; the fanning frequency increased over time after spawning. Meanwhile, 27 \nthe males did not display any form of parental care behavior. The eggs had hook-like 28 \nstructures that might enable the eggs to attach to the flank of the parent. Hatching 29 \noccurred between 22 and 28 days after spawning. Newly hatched juveniles already had 30 \na full set of fin rays ; their morphology was similar to that of adult fish.  Moreover, we 31 \nreviewed the evolution of parental care behaviors and egg types among frogfish family. 32 \n 33 \nKeywords 34 \nAntennariidae, aquarium experiment, early ontogeny, egg care, parental care35 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nIntroduction 36 \n 37 \nAntennariidae, the frogfish, belongs to the order Lophiiformes and consists of two 38 \nsubfamilies: Antennariinae and Histiophryninae, with 14 genera (Pietsch and Arnold 39 \n2020). Frogfishes are found in all tropical and subtropical oceans and seas except for the 40 \nMediterranean (Jordan and Richardson 1908; Pietsch 2004; Pietsch and Arnold 2020), 41 \nand they are predatory fishes (Arnold et al. 2014). Almost all frogfishes resemble 42 \ncertain environmental objects, such as sponges, rocks, and coral (Arnold et al. 2014). 43 \nCryptic colorations can help frogfishes avoid predation and misidentification by prey 44 \n(Arnold et al. 2014). Although the fishes in the family Antennariidae have diverse egg 45 \ntypes and reproductive behaviors (Pietsch and Grobecker 1980; Pietsch and Grobecker 46 \n1987; Kuiter 1993; Liem 1998; Pietsch et al. 2009; Arnold et al. 2014; Arnold and 47 \nPietsch 2018; Pietsch and Arnold 2020; Bray and Gomon 2020), there are few detailed 48 \nstudies on their reproduction. According to the few reports in the literature (Mito 1960; 49 \nFries 1973; Pietsch and Grobecker 1987), the fishes in the family Antennariidae are 50 \ndivided into two types of species that lay an aggregated pelagic egg (e.g., Antennarius 51 \nspp. and Histrio spp.) and demersal eggs (e.g., Histiophryne spp. and Lophiocharon 52 \nspp.) (Friese 1974; Molter 1983; Fujita and Uchida 1985; Arnold et al. 2014).  53 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 54 \nSome species of Antennariidae demonstrate egg care behavior. For instance, 55 \nLophiocharon spp. cares for a demersal egg mass adhered to the flank of the parents 56 \n(Pietsch and Grobecker, 1980). Nevertheless, the sex of the parent engaging in egg care 57 \nremains unclear. For example, Pietsch and Grobecker (1980) reported that only male L. 58 \ntrisignatus could care for their eggs; however, Pietsch and Arnold (2020) later observed 59 \nthat Lophiocharon trisignatus females cared for their eggs. This confusion is likely due 60 \nto the difficulty in determining the sex of frogfishes based on their external 61 \nmorphologies, although the females are usually are bigger than the males (Pietsch et al. 62 \n2013). Therefore, detailed studies of reproduction are needed to fully understand the 63 \nearly-life history of frogfishes exhibiting various reproductive strategies.  64 \n 65 \nHere, we report the sequences of spawning and egg care behavior of marble-mouth 66 \nfrogfish (L. lithinostomus). This species inhabits coastal reefs of Indonesia and the 67 \nPhilippines and resembles algae-covered rocks (Arnold and Pietsch 2012; Arnold et al. 68 \n2014). There have been no reports on the species’ reproductive behavior. In addition to 69 \nthe analysis of the reproductive behavior, we report the unique egg structure adapting to 70 \nthe egg care behavior by female parents and the morphology of newly hatched juveniles. 71 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nMoreover, we review the evolution of parental care behaviors and egg types, pelagic or 72 \ndemersal eggs, among frogfish families using limited available reports. 73 \n 74 \nMaterials and methods 75 \nStudied Fish 76 \n 77 \nTen adult L. lithinostomus, including four males with a standard length (SL) of 105.7 ± 78 \n15.6 mm and six females with an SL of 120.0 ± 10.1 mm, were obtained from the 79 \nornamental fish company (Kamihata Fish Industries LTD, Hyogo, Japan) and identified 80 \naccording to Pietsch (2004). The sexes of these individuals were determined by 81 \ndissection when they died. Moreover, we used one female with an SL of 89.5mm bred 82 \nby the Marine Science Museum, Fukushima. Thus, a total of 11 L. lithinostomus 83 \nindividuals were used in this study. They were individually identified by observers 84 \nusing their unique coloration and body morphology. The individuals used here have 85 \nbeen registered as voucher specimens in the Marine Science Museum, Fukushima 86 \n(registration number: AMF0053 and 0115).  87 \n 88 \nFishkeeping and reproduction 89 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 90 \nWe used two water tanks in this study. Four fish (two males and two females) were held 91 \nin a 1-m3 water tank (1.0 × 1.0 × 1.0 m) at 26.0℃ from 2016 to 2019, and six fish (two 92 \nmales and five females) were held in a 0.45-m3 water tank (0.7 × 1.0 × 0.6 m) at 93 \n25.0 ℃ from 2018 to 2019. The fish were fed silver-stripe round herrings 94 \n(Spratelloides gracilis) once daily. Between 2016 and 2019, 28 spawning events were 95 \nobserved.  96 \n 97 \nBehavioral observation 98 \n 99 \nWhen tracking behavior by the males toward the females was observed, we videotaped 100 \ntheir reproductive behaviors and continued throughout the day until spawning was 101 \nobserved. Then, we recorded individuals that cared for eggs and the lateral surface of 102 \nthe body to which the eggs were attached.  103 \n 104 \nWe quantified the investment in offspring by parents by conducting detailed behavioral 105 \nobservation of four reproductive events (n = four females). Female parents attached 106 \nspawned eggs to the left or right flank and exhibited fanning behavior toward the eggs 107 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nusing their dorsal fin. We count the fanning behavior for 10 min per day. This 108 \nbehavioral observation was conducted between 8:00 and 17:00. Lastly, statistical 109 \nanalysis was performed using R (R development core team 2020). 110 \n 111 \nMorphology of Eggs and newly hatched juvenile 112 \n 113 \nBecause one female (137.8 mm SL) renounced an egg mass twice during observation, 114 \nwe used these the egg masses for measurements of egg morphology. The number of 115 \neggs per clutch was estimated by comparing the weight of fifty eggs and the weight of 116 \negg mass. Egg size was measured using a stereomicroscope (ZEISS SteREO 117 \nDiscovery.V12, ZEISS Research Microscopy Solution, Ltd., Germany). The egg mass 118 \nwas not consumed by other fish because we sampled them right after it was renounced.  119 \n 120 \nWe observed the morphology of newly hatched juveniles by fixing them in 10% 121 \nformalin (n = 10). Their standard lengths were measured to the nearest 0.1 mm under a 122 \nmicroscope, according to Hubbs and Lagler (1958), after they were overdosed with an 123 \nanesthetic (MS-222, Wako Pure Chemical Industries). The number of fin ray of all fins 124 \nof the newly hatched juveniles were counted and compared with those of adult fish. The 125 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nobserved individuals were deposited as voucher specimens (registration number: 126 \nAMF0127). We classified the newly hatched fish as \"juveniles,\" according to the 127 \ndevelopmental classification by Kendall et al. (1984) because their fin rays already 128 \nattained the full complement right after hatching.  129 \n 130 \nReview of the evolution of egg care in the frogfish family 131 \n 132 \nTo understand the functional role of egg care behaviors and morphology of eggs in L. 133 \nlithinostomus, we performed a phylogenic analysis to examine the evolution of egg care 134 \nbehaviors in the frogfish family. The data on the reproductive characteristics, such as 135 \nthe size and number of eggs, egg types such as pelagic or demersal, and egg care 136 \nbehaviors, were collected from the primary literature. Eventually, we found a total of 14 137 \nreports on the reproduction of 14 frogfish species. The egg care behaviors were 138 \nclassified into four types according to Arnold et al. (2014): 1) Attaching to the body of 139 \nthe parent (A-type); 2) carrying within a pocket of parent (P-type); 3) guarding in the 140 \nnest (N-type); 4) no egg care. In the species with A-type behavior, an egg mass is 141 \nattached to the lateral side of the parent’s body and nurtured (Pietsch and Grobecker 142 \n1980; Pietsch and Arnold 2020). In the species with P-type behavior, the parents 143 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nembrace their egg masses by curling their dorsal and anal fins (Pietsch and Grobecker 144 \n1987; Pietsch et al. 2009; Arnold and Pietsch 2018; Pietsch and Arnold 2020). Finally, 145 \nin species with N-type behavior, the parents spawn specific substrates and care for their 146 \negg mass (Kuiter 1993; Liem 1998; Arnold et al. 2014; Bray and Gomon 2021). We also 147 \nrecorded the flank that the A-type and P-type parents carried their egg mass because we 148 \nfound a left-bias in egg caring on the flank in L. lithinostomus. 149 \n 150 \nIn addition, the presence or absence of egg care and egg types (pelagic or demersal) in 151 \nmajor species relative to frogfishes were recorded. Reports on the presence or absence 152 \nof egg care and egg types for coffinfish Chaunax abei (Mimori 2015), anglerfish 153 \nLophius litulon (Ishikawa et al. 2022), and batfish Ogcocephalus nasutus (Christie 154 \n2016) were obtained. Although footballfishes (e.g., Himantolophus appelii) are also a 155 \nmajor species relative to frogfishes (Rabosky et al. 2018), we could not find reports on 156 \ntheir reproduction because they inhabit the deep sea. 157 \n 158 \nFor the phylogenetic analyses, the phylogenetic tree of frogfishes and related species 159 \nwas obtained or modified from a previously published tree (Rabosky et al. 2018). This 160 \ntree was estimated using RAxML and dated using treePL, and non-target species were 161 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nexcluded using the “droptip” function in R ver. 4.0.3 (R Core Team. 2014), package 162 \n“ape” (Paradis et al. 2004). Ancestral states of the egg were assessed by using Mesquite 163 \nVersion 3.61 (Maddison and Maddison 2019) with the likelihood method in a Markov, 164 \nk-status, 1-parameter model, using the modified tree. The reconstruction of ancestral 165 \nstates was conducted using 11 frogfish species and their related species because of their 166 \nclear phylogenetic relationships (Rabosky et al. 2018). 167 \n 168 \nResults 169 \nSpawning behavior 170 \n 171 \nDuring observation, we successfully videotaped three spawning behaviors 172 \n(Supplementary S1). Adult males and females were usually solitary in the water tank; 173 \nhowever, multiple males began to follow females approximately seven days before 174 \nspawning. When spawning began, a male ejaculated when a female released an egg 175 \nmass from its gonopore. Before the egg mass was attached to the side of the female, the 176 \nmale left the female (Fig. 1a). A female attached the spawned egg mass to its side using 177 \nits caudal fin to wrap it around the lateral surface of its body. The egg mass was not 178 \nreleased into the water but transferred from the gonopore to the lateral surface of the 179 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nbody (Fig. 1b). A female protected an egg mass by covering it with her dorsal, anal, and 180 \ncaudal fins and occasionally fanned it with her dorsal fin (Supplementary S1). 181 \n 182 \nEgg care and laterality of female parents 183 \n 184 \nAll egg care behaviors (n = 28) were performed only by the female parents, and no 185 \nmales cared for the eggs, based on the spawning events by seven females. While one of 186 \nthe females cared for its eggs on its right flank three times, 89.2 % of the egg masses 187 \n(25 out of 28 spawning events) were cared for on the left flank of the parent, suggesting 188 \na significant bias towards egg care on the left flank (Pearson’s chi-squared test χ21 189 \n=37.605; p < .001). The number of fanning behaviors towards attached eggs varied 190 \nconsiderably, ranging from 0 to 537 per 10 min with a mean of 201.9 ± 148.6; it 191 \nincreased from the day after spawning until fifteen days after spawning. However, after 192 \nthat, the number of fanning behaviors remained unchanged (Fig. 2). Hatching began 22 193 \nto 23 days after spawning; almost all juveniles hatched by days 27 to 28. The juveniles 194 \nwere scattered throughout the tank after hatching.  195 \n 196 \nMorphological characteristics of eggs and newly hatched juveniles 197 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n  198 \nThe egg mass had an elliptic shape with a mean major axis of 91.2 mm and a mean 199 \nminor axis of 62.5 mm. The eggs were round; their average length was 2.9 ± 0.3 mm, 200 \nranging from 2.4 to 4.1 mm (n = 50; from two egg masses from one female). The 201 \ncoloration of eggs was semi-transparent white just after spawning. Eye pigmentation 202 \nwas confirmed 11 days after spawning. Interestingly, demersal eggs had a unique 203 \nstructure, an s-shaped hook (Fig. 3). Although juveniles (n = 10) had yolk sacs just after 204 \nhatching, all their fin rays already attained the full complement compared with the 205 \nadults (Fig. 4). The standard length (SL) of the newly hatched juveniles (n = 10) was 206 \n5.70 ± 0.20 mm, ranging from 5.40 to 6.09 mm. They already had esca, which was the 207 \ndistinctive protuberance of this group used for predation as a lure; in addition, they 208 \ncould swim freely and often they could attach to substrates in the water tank.  209 \n 210 \nEvolution of egg care among frogfish family 211 \n 212 \nWe obtained 14 anecdotal or qualitative reports in the field associated with egg care 213 \nbehaviors of frogfishes (Table 1). The species in Antennariinae do not display egg care 214 \nbehaviors and have pelagic eggs. In contrast, the species in Histiophryninae have three 215 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\ntypes of egg care behaviors, i.e., attaching to the body of the parent (A-type), carrying 216 \nwithin a pocket (P-type), and guarding at the nest (N-type). The species in 217 \nHistiophryninae have demersal eggs. Any form of egg care behavior has not been 218 \nreported from the species related to frogfishes. Ancestral state reconstruction revealed 219 \nthat P-type and N-type egg care behaviors had been evolved independently from the 220 \nno-care species in Histiophryninae (Fig. 5). In addition, A-type care had been evolved 221 \nfrom a group within the P-type egg care (Fig. 5).  222 \n 223 \nDiscussion  224 \n 225 \nIn this study, we investigated the reproductive behavior of marble-mouth frogfish 226 \n(Lophiocharon lithinostomus). We found that multiple males stalked breeding females 227 \nbefore spawning, although only one male could participate in the spawning event. 228 \nSimilar behavior by adult males before spawning has been observed in several fishes 229 \nand functions as mate guarding (Yokoi et al. 2016). Moreover, female filefish (Rudarius 230 \nercodes) move around and choose any male following them as their reproductive 231 \npartners (Akagawa et al. 1998). Thus, the behavior of L. lithinostomus will be regarded 232 \nas a process of mate choice or mate guarding. For L. lithinostomus, fertilization 233 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\noccurred immediately after the onset of spawning, and the male left the female before 234 \nthe entire egg mass was laid. The time difference between fertilization and spawning did 235 \nnot affect the fertilization rate, and most of the eggs were fertilized (Mori personal 236 \nobservation). Fertilized eggs were covered with a gelatinous curtain. Since the activity 237 \ntime of the sperms in the ovarian cavity fluid of females plays an important role in 238 \nfertilization, the gelatinous membrane covering the eggs may play a role similar to the 239 \novarian cavity fluid (Hayakawa and Munehara 2001). 240 \n 241 \nWe found that only females attached egg masses to their flanks and began to fan the egg 242 \nmasses after spawning. Also, regarding L. trisignatus, a related species to L. 243 \nlithinostomus, only female parents care for egg masses (Pietsch and Arnold 2020). In 244 \nmany species in the frogfish family (Antennariidae), the females are larger and have 245 \nlarger abdomens than the males (Pietsch et al. 2013). Moreover, according to the reports 246 \non the sexual dimorphism in frogfishes, there are sexual differences in the number of 247 \ndermal spinules in striated frogfish (Antennarius scaber) (Breder and Rosen 1966). 248 \nHowever, no definitive sex differences have been found among frogfishes. Therefore, 249 \nthe study of the sex of the frogfishes caring for egg mass should be reconsidered.  250 \n 251 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nMore generally, the morphologies of newly hatched and developing teleost fishes are 252 \nincomplete compared with adults (Mosher 1954; Ehrlich et al. 1976; Martin and Drewry 253 \n1978; Kendall et al. 1984; Satoh et al. 2017). In some species, the mouth and anus of 254 \nthe newly hatched larvae do not open, making feeding impossible (Satoh et al. 2017). 255 \nEven with an open mouth, it would be difficult for the larvae to catch small prey with 256 \nfin folds due to their low swimming ability (Satoh et al. 2017). Therefore, 257 \nmorphological ontogeny can highly influence early-life history, especially foraging 258 \necology (Satoh et al. 2017). For example, the newly hatched larvae of frogfishes in the 259 \nAntennariinae with pelagic eggs, such as Histrio histrio, do not have advanced 260 \norganogenesis (Mosher 1954). In contrast, the newly hatched juveniles of L. 261 \nlithinostomus already have open mouths, and all their fin rays attain the full 262 \ncomplement, suggesting that they already have had robust foraging abilities 263 \nimmediately after hatching.  264 \n 265 \nThe dispersal of many marine fishes is commonly concentrated in egg or early larval 266 \nstages (Barlow 1981; James et al. 2002). However, the dispersal ability of L. 267 \nlithinostomus would be even lower than the pelagic egg group, such as Histrio histro 268 \nand Antennarius striatus, because L. lithinostomus eggs do not float and are cared for by 269 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nthe female parents. Indeed, frogfishes in Histiophryninae have a smaller distribution 270 \narea than those in Antennariinae and highly depend on the specific local area (Arnold 271 \nand Pietsch 2012). In general, species with a low dispersal ability are more susceptible 272 \nto local environmental changes (Baguette et al. 2012; Gonçalves et al. 2016). Recently, 273 \nthe deterioration of the coastal environment, including coral reefs, has been reported 274 \n(Kennish 2002). Trends in species such as L. lithinostomus, which are presumed to be 275 \nless tolerant of environmental changes, should be all eyes. 276 \n 277 \nAccording to the ancestral state reconstruction, three types of egg care had evolved only 278 \namong the species of Histiophryninae, which have demersal eggs, from non-egg care 279 \nspecies. These findings suggested that egg care behaviors had co-evolved with demersal 280 \neggs. Demersal eggs are generally larger than pelagic eggs, and the fry can hatch with 281 \nadvanced organogenesis (Sargent et al. 1987; Araujo‐Lima 1994). Indeed, we found 282 \nnewly hatched juveniles of L. lithinostomus had highly advanced organogenesis. 283 \nHowever, demersal eggs stick to substrates and thus, cannot change location. Therefore, 284 \ndemersal eggs will be exposed to high predation pressure unless they have a chemical 285 \ndefense or camouflage coloration (Eisner et al. 2000; Ruxton et al. 2004; Skelhorn 286 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n2015). The frogfishes of Histiophryninae may have evolved egg care as a 287 \ncounter-strategy to avoid the high predation pressure applied on demersal eggs. 288 \n 289 \nThe N-type egg care is thought to have evolved independently of these lineages. Also, 290 \nwe found that A-type egg care may have evolved from P-type egg care. It would be very 291 \nunusual for diverse modes of egg care to evolve among closely related species. In 292 \nspecies with P-type egg care, female parents embrace their egg masses by curling dorsal 293 \nand anal fins into a pocket-like structure (Arnold et al. 2014). In contrast, egg masses 294 \nare adhered to the flank for fishes with A-type egg care, such as L. lithinostomus. We 295 \nfound an s-shaped hook structure in the egg masses of L. lithinostomus. This structure 296 \nmay function as adhesives to attach the egg masses to the flank of female parents so that 297 \nthe eggs are tightly glued to the parental body. Indeed, when a female parent abandoned 298 \nthe glued eggs, she shook her body vigorously to remove the eggs. (Mori personal 299 \nobservation). 300 \n 301 \nThe number of eggs varies between egg care types. For example, speceis with N-type 302 \nhas the highest number of eggs at approximately 5000 eggs (n = 1 species), followed by 303 \nA-type with 320 to 1311 eggs (n = 2 species) and P-type with 105 to 220 eggs (n = 3 304 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nspecies) (Table 1). The number of eggs by N-type fish will not be limited because the 305 \nfish spawn egg masses on a substrate (Arnold et al. 2014). However, in A-type and 306 \nP-type, the size of egg masses will be restricted by the flank of the parents because the 307 \nfish attach or entrap their eggs to the body (Arnold et al. 2014). Although they are 308 \nconstrained by the number of eggs they can spawn in one reproductive event, the 309 \nparents with A- and P-type egg care can migrate while raising their eggs. Additionally, 310 \nthe species with A-type egg care have more eggs than those with P-type egg care (Table 311 \n1). In general, oxygen demand, hence the frequency of fanning, increases with 312 \nembryonic development (Abe and Munehara 2005). Unlike a P-type species, an A-type 313 \nspecies can use its dorsal fins for fanning behavior, likely allowing it to hold more eggs, 314 \neven though the number of eggs is limited by body size (Bagenal and Braum 1978; 315 \nWarner 1984). Although these comparisons are constrained by the simplicity of the 316 \nevolution of egg care modes among frogfishes, they suggest that the number of eggs 317 \nwill vary depending on the egg care mode. 318 \n 319 \nFinally, we found that L. lithinostomus females preferred to use the left side of their 320 \nbodies for egg care. Interestingly, they also appeared able to use their right flank for egg 321 \ncare because one female used the right flank. Lateralized (left-right asymmetry) 322 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nmorphology and behavior among fishes have been reported in many species (Takeuchi 323 \net al. 2016; Hori et al. 2017). Because left-biased egg care was not detected in species 324 \nwith A-type egg care (L: R = 3:2 from 5 reports for 3 species, Table 1) but L. trisignatus 325 \nalso displayed left-biased egg care similar to L. lithinostomus (Table 1), this tendency 326 \nmay have co-evolved with A-type egg care. However, we could not determine why L. 327 \nlithinostomus females used their left side for egg care; this interesting phenomenon 328 \nrequires further morphological or evolutionary validation.  329 \n 330 \nAcknowledgments 331 \n 332 \nWe are grateful to Mr. Yoshitaka Abe and Mr. Takeshi Furukawa of Aquamarine 333 \nFukushima for cooperation for fishkeeping. We would also like to thank Dr. Toshiro 334 \nSaruwatari and Dr. Toru Miura of University of Tokyo, Dr. Takashi Asahida and Dr. Go 335 \nKatayose of Kitasato University, and Dr. Atsushi Sogabe of Hirosaki University for the 336 \ncomments on our study. We would also like to thank anonymous reviewers for their 337 \nhelpful comments. 338 \n 339 \nReferences 340 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nAbe T, Munehara H (2005) Spawning and maternal‐care behaviours of a copulating 341 \nsculpin, Radulinopsis taranetzi. Journal of Fish Biology 67(1): 201–212. 342 \nhttps://doi.org/10.1111/j.0022-1112.2005.00728.x 343 \n 344 \nAraujo‐Lima CARM (1994) Egg size and larval development in Central Amazonian 345 \nfish. Journal of Fish Biology 44(3): 371–389. 346 \nhttps://doi.org/10.1111/j.1095-8649.1994.tb01219.x 347 \n 348 \nArnold RJ, Pietsch TW (2018) Fantastic beasts and where to find them: a new species of 349 \nthe frogfish genus Histiophryne Gill (Lophiiformes: Antennariidae: Histiophryninae) 350 \nfrom Western and South Australia, with a revised key to congeners. Copeia 106(4): 351 \n622–631. https://doi.org/10.1643/CI-18-112 352 \n 353 \nAkagawa I, Kanda T, Okiyama M (1998). Female mate choice through spawning parade 354 \nformed by male-male competition in filefish Rudarius ercodes. Journal of Ethology 355 \n16(2): 105–113. https://doi.org/10.1007/BF02769289 356 \n 357 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nArnold RJ, Harcourt R, Pietsch TW (2014) A new genus and species of the frogfish 358 \nfamily Antennariidae (Teleostei: Lophiiformes: Antennarioidei) from New South Wales, 359 \nAustralia, with a diagnosis and key to the genera of the Histiophryninae. Copeia 360 \n2014(3): 534–539. https://doi.org/10.1643/CI-13-155 361 \n 362 \nBagenal TB, Braum E (1978) Eggs and early life history. In: Bagenal TB (ed) Fish 363 \nreproduction in freshwaters. Blackwell, Oxford, 165–201. 364 \n 365 \nBaguette M, Benton TG, Bullock JM (2012) Dispersal ecology and evolution. Oxford 366 \nUniversity Press, 498 pp. 367 \n 368 \nBray DJ, Gomon MF (eds) (2021) Fishes of Australia. Museums Victoria and 369 \nOzFishNet. http://fishesofaustralia.net.au/ [Accessed on 19 January 2021] 370 \n 371 \nBarlow GW (1981) Patterns of parental investment, dispersal and size among coral-reef 372 \nfishes. In Ecology and ethology of fishes. Springer, Dordrecht, 65–85. 373 \n 374 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nBreder GMJ, Rosen DE (1966) Modes of reproduction in fishes. Natural History Press, 375 \nGarden City, NY, 941 pp. 376 \n 377 \nChristie BL, Monyoya PZ, Torres LA, John WF (2016) The natural history and 378 \nhusbandry of the walking batfishes (Lphiformes: Ogcocephalidae). Drum and Croaker 379 \n47: 6–40. 380 \n 381 \nEhrlich KF, Blaxter JHS, Pemberton R (1976) Morphological and histological changes 382 \nduring the growth and starvation of herring and plaice larvae. Marine Biology 35(2): 383 \n105–118. 384 \n 385 \nEisner T, Eisner M, Rossini C, Iyengar VK, Roach BL, Benedikt E, Meinwald J (2000) 386 \nChemical defense against predation in an insect egg. Proceedings of the National 387 \nAcademy of Sciences 97(4): 1634–1639. 388 \n 389 \nFriese UE (1973) Anglerfishes. Marine Aquarist 4(5): 29–36. 390 \n 391 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nFujita S, Uchida K (1959) Spawning habits and early development of a sargassum fish, 392 \nPterophryne histrio (‘Linne’). Science Bulletin of the Faculty of Agriculture Kyushu 393 \nUniversity 17(3): 277–282. 394 \n 395 \nGonçalves J, Honrado JP, Vicente JR, Civantos E (2016) A model-based framework for 396 \nassessing the vulnerability of low dispersal vertebrates to landscape fragmentation 397 \nunder environmental change. Ecological Complexity 28: 174–186. 398 \nhttps://doi.org/10.1016/j.ecocom.2016.05.003 399 \n 400 \nHayakawa Y , Munehara H (2001) Facultatively internal fertilization and anomalous 401 \nembryonic development of a non-copulatory sculpin Hemilepidotus gilberti Jordan and 402 \nStarks (Scorpaeniformes: Cottidae). J Exp Mar Biol Ecol 256 (1): 51–58. 403 \nhttps://doi.org/10.1016/S0022-0981(00)00303-8 404 \n 405 \nHori M, Nakajima M, Hata H, Yasugi M, Takahashi S, Nakae M, Kosaku Yamaoka, 406 \nMasanori Kohda, Jyun-ichi Kitamura, Masayoshi Marhata, Hirokazu Tanaka, Norihiro 407 \nOkada, Takeuchi Y (2017) Laterality is universal among fishes but increasingly cryptic 408 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\namong derived groups. Zoological Science 34(4): 267–274. 409 \nhttps://doi.org/10.2108/zs160196 410 \n 411 \nHubbs CL, Lagler KF (1958) Fishes of the great lakes region. Bull. Carbrook Inst. Sci 412 \n26: 1–213. 413 \n 414 \nI.O.P. diving news (2001) 139, V ol.12, No.10, pp.1. 415 \n 416 \nIshikawa T, Nakaya M, Takatsu T (2022) Embryonic development and effect of water 417 \ntemperature on hatching of Lophius litulon. Environmental Biology of Fishes 105, 77–418 \n86. https://doi.org/10.1007/s10641-021-01195-4 419 \n 420 \nJames MK, Armsworth PR, Mason LB, Bode L (2002) The structure of reef fish 421 \nmetapopulations: modelling larval dispersal and retention patterns. Proceedings of the 422 \nRoyal Society of London. Series B: Biological Sciences 269(1505): 2079–2086. 423 \nhttps://doi.org/10.1098/rspb.2002.2128 424 \n 425 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nJordan DS, Richardson RE (1908) Fishes from islands of the Philoppine Archipelago. 426 \nBulletin of the United States Bureau of Fisheries 27: 233–287. 427 \n 428 \nKendall AW Jr, Ahlstrom EH, Moser HG (1984) Early life history stages of fishes and 429 \ntheir characters. In: Moser HG, Richards WJ, Cohen DM, Fahay MP, Kendall AW Jr, 430 \nRichardson SL (eds) Ontogeny and systematics of fishes, Am Soc Ichthyol Herpetol 431 \nSpec Publ No 1. Allen Press, Lawrence, 11–22. 432 \n 433 \nKennish MJ (2002) Environmental threats and environmental future of estuaries. 434 \nEnvironmental Conservation 29(1): 78–107. 435 \nhttps://doi.org/10.1017/S0376892902000061 436 \n 437 \nKuiter RH (1993) Coastal Fishes of South-Eastern Australia. University of Hawaii Press, 438 \nHonolulu, Hawaii, 437 pp. 439 \n 440 \nLiem KF (1998) Introducing fishes. In: Paxton JR, Eschemeyer WN (ed) Encyclopedia 441 \nof Fishes, 2nd edition. Academic Press, San Diego, California, 14–19. 442 \n 443 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nMartin FD, Drewry GE (1978) Development of fishes of the Mid-Atlantic Bight. An 444 \natlas of egg, larval and juvenile stages. VI. Stromateidae through Ogcocephalidae U S 445 \nFish Wild Serv Biol Serv Prog. FWS/OBS-78/12: 1–416. 446 \n 447 \nMolter TJ (1983) A spawning of the Atlantic anglerfish Antennarius scaber. Cuvier. 448 \nFreshwater and Marine Aquarium 6(1): 34–35, 66, 69. 449 \n 450 \nMosher C (1954) Observations on the behavior and the early larval development of the 451 \nSargasum fish Histrio histrio (Linnaeus). Zoologica 39: 141–152. 452 \n 453 \nMimori R (2015) 454 \nhttps://www.tokyo-zoo.net/topic/topics_detail?kind=news&inst=&link_num=22733 455 \n[Accessed on 20 February 2022] 456 \n 457 \nParadis E, Claude J, Strimmer K (2004) APE: Analyses of phylogenetics and evolution 458 \nin R language. Bioinformatics 20(2): 289–290. 459 \n 460 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nPietsch TW (2004) A new species of the anglerfish genus Lophiocharon Whiteley 461 \n(Lophiiformes: Antennaridae) from Australian waters. Records of the Australian 462 \nMuseum 56(2): 159–162. https://doi.org/10.3853/j.0067-1975.56.2004.1418 463 \n 464 \nPietsch TW, Grobecker DB (1980) Parental care as an alternative reproductive mode in 465 \nan antennariid anglerfish. Copeia 1980: 551–553. 466 \n 467 \nPietsch TW, Grobecker DB (1987) Frogfishes of the world: systematics, zoogeography, 468 \nand behavioral ecology. Stanford Univ Press, Stanford, CA, 420 pp. 469 \n 470 \nPietsch TW, Arnold RJ (2020) Forogfishes: biodiversity, zoogeography, and behavioral 471 \necology. Johns Hopkins University Press, Baltimore, 624 pp. 472 \n 473 \nPietsch TW, Arnold RJ, Hall DW (2009) A bizarre new species of frogfish of the genus 474 \nHistiophryne from Ambon and Bali, Indonesia. Copeia 2009: 37–45. 475 \n 476 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nPietsch TW, Ross SW, Caruso JH, Saunders MG, Fisher CR (2013) In-situ observations 477 \nof deep-sea goosefish sladenia shafersi caruso and bulis (Lophiiformes: Lophiidae), 478 \nwith evidence of extreme sexual dimorphism. Copeia 2013(4): 660–665. 479 \n 480 \nRabosky DL, Chang J, Cowman PF, Sallan L, Friedman M., Kaschner K, Garilao C 481 \nNear TJ, M Coll, Alfaro ME (2018) An inverse latitudinal gradient in speciation rate for 482 \nmarine fishes. Nature 559: 392–395. https://doi.org/10.1038/s41586-018-0273-1 483 \n 484 \nRuxton G, Sherratt T, Speed M (2004) Avoiding attack: The evolutionary ecology of 485 \ncrypsis, warning signals and mimicry. Oxford, U.K.: Oxford University Press, 278 pp. 486 \n 487 \nSargent RC, Taylor PD, Gross MR (1987) Parental care and the evolution of egg size in 488 \nfishes. The American Naturalist 129(1): 32–46. 489 \n 490 \nSatoh S, Tanoue H, Ruitton S, Mohri M, Komatsu T (2017) Morphological and 491 \nbehavioral ontogeny in larval and early juvenile discus fish Symphysodon 492 \naequifasciatus. Ichthyological Research 64(1): 37–44. 493 \nhttps://doi.org/10.1007/s10228-016-0530-y 494 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 495 \nSkelhorn, J (2015) Masquerade. Current Biology 25: 643–644. 496 \nhttps://doi.org/10.1016/j.cub.2015.02.069 497 \n 498 \nMaddison WP, Maddison DR (2019) Mesquite: a modular system for evolutionary 499 \nanalysis. Version 3. http://www.mesquiteproject.org. 500 \n 501 \nTakeuchi Y , Hori M, Tada S, Oda Y (2016) Acquisition of lateralized predation behavior 502 \nassociated with development of mouth asymmetry in Lake Tanganyika scale-eating 503 \ncichlid fish. PLOS ONE. https://doi.org/10.1371/journal.pone.0147476 504 \n 505 \nYokoi S, Ansai S, Kinoshita M, Naruse K, Kamei Y , Young LJ, T Okuyama Takeuchi H 506 \n(2016) Mate-guarding behavior enhances male reproductive success via familiarization 507 \nwith mating partners in medaka fish. Frontiers in Zoology 13(1): 1–10. 508 \nhttps://doi.org/10.1186/s12983-016-0152-2 509 \n 510 \nWarner RR (1984) Mating behavior and hermaphroditism in coral reef fishes. American 511 \nScientist 72: 128–136. 512 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 513 \n 514 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nTable 1. Comparison of reproduction modes for frogfish family Antennaridae. 515 \nSpecies Egg size (mm) The number of eggs Type of egg Parental care mode Laterality of body side used for egg care Reference(s)\nAntennariinae\nAntennarius striatus 0.6 89635 pelagic No care － Fujita and Uchida 1959\nAntennarius nummifer 0.7 73000–288000 pelagic No care － Piestch and Grobecker 1987\nHistrio histro 0.7 48800 pelagic No care － Ray 1961\nHistiophryninae\nHistiophryne cryptacanthus 3.6–4.2 115 demersal P-type  L:R = 1:0 (n  = 1 report) Piestch and Grobecker 1987\nHistiophryne bouganivilli 3.2–3.9 105 demersal P-type  L:R = 1:1 (n  = 2 reports) Piestch and Arnold 2020\nHistiophryne psychedelica 3.0–4.0 220 demersal P-type  L:R = 1:1 (n  = 2 reports) Piestch et al. 2009; Piestch and Arnold 2020\nHistiophryne narungga No data No data demersal P-type No data Arnold and Piestch 2018\nLophiocharon trisignatus  3.2–3.6 650 demersal A-type  L:R = 4:0 (n  = 4 reports) Piestch and Grobecker 1980; Piestch and Arnold 2020\nLophiocharon lithinostomus 2.1–4.1 320–1311 demersal A-type L:R = 26:3 (n  = 3 reports) I.O.P Diving News 2001; Piestch and Arnold 2020; Present study\nEchinophryne crassispina No data 150 demersal N-type － Liem 1998; Bray and Gomon 2021\nEchioophryne reynoldsi No data No data demersal N-type － Bray and Gomon 2021\nRhycherus filamentosus  5.0 5000 demersal N-type － Kuiter 1993; Bray and Gomon 2021\nPorophryne erythrodactylus No data No data demersal N-type － Arnold et al. 2014\nPhyllophryne scortea No data No data demersal N-type － Bray and Gomon 2021\n 516 \nA-type: attaching to the body of the parent, P-type: carrying within a pocket of parent, N-type: guarding in the nest. 517 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nFigure and legends 518 \n 519 \nFigure 1. A Photograph of a gravid female (dark brown color) marble-mouth frogfish 520 \nLophiocharon lithinostomus) surrounded by two courting males (green color). B Female 521 \nmarble-mouth frogfish with egg mass attached on the left side of bodies. 522 \n 523 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\nFigure 2. Relationship between days after spawning and the frequency of fanning 524 \nbehavior by female marble-mouth frogfish Lophiocharon lithinostomus. A solid line 525 \nand a grey area indicate the loess regression and 95% Confidence interval, respectively. 526 \n 527 \nFigure 3. Photograph of fertilized eggs and gelatinous curtains with S-shaped hooks 528 \nsurrounding the eggs. White scale bar: 2 mm. 529 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 530 \nFigure 4. Photographs of a marble-mouth frogfish Lophiocharon lithinostomus juvenile 531 \njust after hatching. A Overhead view. B Lateral view. White scale bar: 1 mm. 532 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895\n\n 533 \nFigure 5. Evolution and phylogenetic distribution of the forms egg care in 534 \nAntennariidae and related species. The ultrametric tree was modified from Rabosky et 535 \nal. (2018). Egg care behaviors were classified into four types: 1) Attaching to the body 536 \nof the parent (blue); 2) carrying within a pocket of a parent (green); 3) guarding in the 537 \nnest (black); 4) no egg care (white).  538 \n 539 \nESM S1 540 \nSpawning and fanning behavior of marble-mouth frogfish Lophiocharon lithinostomus. 541 \n 542 \nAuthor-formatted, not peer-reviewed document posted on 06/04/2022. DOI:  https://doi.org/10.3897/arphapreprints.e84895","source_license":"CC-BY-4.0","license_restricted":false}