Introduction

Model organisms are representative species used for research in various biological fields. Numerous studies using model organisms have been conducted in laboratories, contributing to our understanding of complex biological phenomena in simple environments by controlling complicated conditions that are difficult to regulate. However, natural environments in which organisms live are not as simple as those in the laboratory [1]. Surprisingly, relatively little is known about the ecology, behavior, reproduction, and life histories of some model organisms in their natural environments, and the extent to which laboratory observations are consistent with their ecology and behavior in nature remains unclear. To comprehensively understand an organism, it is essential to elucidate its ecology and behavior in natural habitats [2]. Field research provides insights into the ecology and behavior of the studied species. It contributes to understanding the adaptive significance and mechanisms of their genomic, morphological, and physiological traits. Understanding ecology and behavior in the wild also enables the maintenance, breeding, and experimentation of research species under conditions closer to their natural environment without inducing stress [3], consequently leading to improved animal welfare. Furthermore, this is particularly important to ensure the validity of the research outcomes, as the poorly understood welfare of the study species can affect the reliability of the findings.

The freshwater fish medaka (Oryzias latipes) is widely used as a model organism in various research fields, including physiology, genetics, developmental biology, behavioral science, and medicine. This is due to its unique combination of characteristics, such as small body size (3–4 cm standard length), ease of laboratory breeding, sexual dimorphism, short generation time, large and transparent eggs, and a relatively compact genome [4, 5]. However, despite using medaka as a model organism for over a century, its ecology and behavior in the wild, particularly its reproductive ecology, remain unclear [6]. Medaka (O. latipes) are distributed throughout Japan, inhabiting still or slow-moving water bodies such as rice paddies, ponds, and agricultural waterways [4]. The preferred habitat for medaka is shallow waters near the shore with abundant vegetation, where they feed on microorganisms, including algae and zooplankton [7]. Their lifespan in the wild is approximately one year [8]. The breeding season of medaka spans from early summer to autumn [9–12]. To our knowledge, mating behavior in the wild has only been reported by Kobayashi et al. (2012), in which, based on visual observations, mating behavior was observed primarily from early morning after sunrise to mid-morning [13].

In the laboratory, medaka mating is observed every morning, with males mating multiple times and females spawning once daily [14, 15]. Spawning typically occurs in pairs, and eggs are fertilized externally. After spawning, the female carries fertilized eggs on her abdomen and brushes against the aquatic plants to attach the eggs after a few hours. Many laboratory experiments on mating in medaka have been conducted because of the ease of observing their mating behavior. Examples include female mate choice [16–19], male mate-guarding behavior [20–22], alternative male reproductive tactics [23–25], and male sperm allocation [26, 27]. All studies were conducted under bright lighting conditions in the morning.

Interestingly, in the laboratory, females complete ovulation at night and are ready to mate in the morning [28]. Medaka activity levels increased significantly 4–5 h before the lights were turned on, which could be associated with behaviors such as searching for mating partners [29]. Furthermore, studies have reported that medaka can spawn in the dark [30] and mate without relying on vision [31]. Indeed, in our previous aquarium observations, females carrying eggs after spawning were frequently observed 3 h before the lights were turned on (Y. Kondo, personal observation). Accumulating evidence suggests that the onset of spawning in female medaka occurs before sunrise. If mating starts before light, the medaka breeding behavior observed under bright conditions in the laboratory may only cover a part of the breeding period (i.e., the latter part).

Therefore, this study aimed to elucidate the reproductive behavior and ecology of medaka during the breeding season in their natural habitat in Gifu, Japan. Based on the evidence of medaka reproduction, we used video cameras in streams during the night to observe behavioral changes. We determined the spawning onset time using egg-carrying females as indicators. Furthermore, we investigated changes in activity levels, courtship behaviors, and other mating behaviors throughout the night.

Methods

Results

Initiation time of spawning

To determine the initiation time of spawning, video recordings were conducted over four days, from July 25 to 28, 2023. During the observations, multiple individual medakas were visible in all videos. Egg-hanging females were observed at 23:54 on July 25, 0:47 on July 27, and 0:01 on July 28 (Table 1). Post-spawning females were consistently observed in all videos at 0:01 on July 28. On other recording days, post-spawning females were visible in nearly all videos from 1:40 to 1:50 am onwards (Table 1; S1 Movie).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Spawning time of medaka (Oryzias latipes) during the breeding season in Gifu, Japan.

https://doi.org/10.1371/journal.pone.0318358.t001

Nocturnal activity and courtship behavior

To examine the nocturnal activity and courtship behavior of the medaka, video analyses were conducted from 21:00 to 5:00 for ten days, from August 1 to 10, 2023. In all the videos analyzed, medaka appeared at least once in each video, with a total of 2,474 times (mean ± SE per one video = 10.36 ± 0.46, n = 239, S2 Table). The fewest appearances were recorded at 22:00 to 23:00, at 247 (mean frequency per video ± SE = 8.23 ± 1.20, n = 30), and the most at 1:00 to 2:00, at 377 (mean frequency per video ± SE = 12.57 ± 1.04, n = 30). The GAMM analysis revealed that the number of medaka appearances was similar from 21:00 to 23:00, gradually increased from 23:00, peaked from 0:00 to 2:00, and decreased after 2:00 (GAMM, deviance = -17.66, df = -5.60, p = 0.0054, n = 239; Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Number of medaka (Oryzias latipes) appearances from 21:00 to 5:00 during the breeding season in Gifu, Japan.

Each plot signifies the observed values from the analyzed videos. The regression curve was based on the generalized additive mixed model (GAMM), and the gray shading indicates the 95% confidence intervals. 21:00–4:00: n = 30 for each, 4:00–5:00: n = 29 video samples.

https://doi.org/10.1371/journal.pone.0318358.g001

Changes in individual swimming and staying times, which are indicators of activity, were examined between 21:00 and 5:00. Individual swimming was observed a total of 2,512 times (mean ± SE per one video = 10.51 ± 0.51, n = 239). The lowest frequency of swimming was recorded from 22:00 to 23:00, at 263 times in total (mean frequency per video ± SE = 8.80 ± 1.56, n = 30), and the highest from 2:00 to 3:00, at 376 times (mean frequency per video ± SE = 12.53 ± 1.64, n = 30). Individual swimming gradually increased from 21:00 and remained high after 1:00 (GAMM, deviance = -11.50, df = -4.95, p = 0.0409, n = 239; Fig 2A). Staying was observed a total of 304 times (mean ± SE per one video = 1.27 ± 0.67, n = 239). The highest frequency was from 21:00 to 22:00, at 62 times in total (mean frequency per video ± SE = 2.07 ± 0.51, n = 30), and the lowest from 3:00 to 4:00, at 13 (mean frequency per video ± SE = 0.43 ± 0.19, n = 30). Staying decreased from 21:00 to 1:00 and remained low from 1:00 to 5:00 (GAMM, deviance = -27.61, df = -4.02, p < 0.0001, n = 239; Fig 2B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Changes in nocturnal behaviors from 21:00 to 5:00 in medaka (Oryzias latipes) during the breeding season in Gifu, Japan.

(a) individual swimming, (b) staying, (c) following, and (d) quick circle. Each plot signifies the observed values from the analyzed videos. The regression curve was based on the generalized additive mixed model (GAMM), and the shading indicates the 95% confidence intervals. 21:00–4:00: n = 30 for each, 4:00–5:00: n = 29 video samples.

https://doi.org/10.1371/journal.pone.0318358.g002

Following and quick circles, both courtship behaviors were observed from 21:00 to 5:00. Following was observed a total of 260 times (mean ± SE per one video = 1.09 ± 0.11, n = 239). The lowest frequency was from 21:00 to 22:00, at 9 times in total (mean frequency per video ± SE = 0.30 ± 0.12, n = 30), and the highest from 2:00 to 3:00 at 50 times (mean frequency per video ± SE = 1.67 ± 0.37, n = 30). Following gradually increased from 21:00 to 22:00 and remained high after 1:00 (GAMM, deviance = -23.12, df = -4.13, p = 0.0001, n = 239; Fig 2C). Quick circle was observed a total of 179 times (mean ± SE per one video = 0.75 ± 0.14, n = 239). The lowest frequency was from 00:00 to 22:00, at 1 time in total (mean frequency per video ± SE = 0.03 ± 0.03, n = 30), and the highest from 2:00 to 3:00, at 48 times (mean frequency per video ± SE = 1.60 ± 0.55, n = 30). Quick circle gradually increased from 23:00 and remained high after 1:00 (GAMM, deviance = -20.86, df = -3.09, p = 0.0001, n = 239; Fig 2D). No spawning behavior was observed during the video recording period.

Discussion

Many aspects of the ecology, behavior, reproduction, and life history of model organisms, including medaka, remain unexplored in the wild. Spawning in medaka begins an hour before or after the lights are turned on [31, 35–37] Although spawning is thought to start from early to mid-morning in the wild [13], no studies have been conducted on the behavior of medaka in the wild. Our quantitative assessment of detailed behavior using video recordings revealed, for the first time, that medaka commenced spawning around midnight, several hours earlier than previously thought. Furthermore, we found that medaka activity levels increased around midnight, peaking at approximately 2:00, and courtship behavior increased around midnight and was maintained at a high level thereafter.

Previous studies have suggested that medaka spawning occurs within one hour before or after the lights are turned on [31, 35–37]. However, these studies did not directly observe spawning in the dark and instead estimated spawning times based on the developmental stages of eggs after spawning. Consequently, the actual spawning onset time could not be identified. Our nocturnal observations using videos revealed that despite sunrise occurring at approximately 5:00, females with eggs adhered to their abdomens were observed around midnight (23:54, 00:47, and 00:01) for four consecutive days. Therefore, at least at the study sites, spawning started around midnight, much earlier than previously documented.

Video analysis elucidated the transitions in nocturnal medaka behavior. From 21:00 to midnight, the activity levels were low, with many individuals sleeping or swimming slowly. The rate of slow swimming decreases as midnight approaches. In contrast, individuals engaged in courtship gradually increased from 21:00, with quick circles increasing from 23:00 and maintaining high levels until sunrise. These results suggest that medaka engage in courtship behavior from midnight onward. Previous laboratory experiments observing 24-hour activity levels in medaka reported a gradual increase in activity, particularly 4–5 h before the lights were turned on [29]. In this study, sunrise time was approximately 5:00, and our findings of increased activity of medaka in the wild at approximately 23:00 are generally consistent with previous laboratory results showing changes in activity. In addition, video observations revealed for the first time that the increase in activity before lighting was attributable to mating behavior.

Medaka may initiate mating behavior at night to avoid predation on eggs. Some species of wrasses and Japanese ornate dragonets spawn around sunset or at night, which is thought to be due to the lower activity levels of plankton predators that consume eggs and piscivorous predators that target adult spawning during this period [38, 39]. This rationale may also apply to medaka. Although detailed studies on the trophic position of medaka are limited, they are mainly preyed upon by dragonfly larvae, birds, and predatory fish [40–42]. While medaka face the risk of predation by birds and piscivorous predators, the lower activity levels of these predators at night compared to daytime allow them to engage in mating behavior without needing vigilance against predators.

It has been reported that medaka can engage in mating behavior without relying on visual cues [31] and that olfaction plays a crucial role in mating [34]. Medaka also spawns in the dark [30]. Based on these insights, it is conceivable that male medaka in the wild can identify ovulated females using olfactory cues without relying on visual cues. Consistent with these findings, in the present study, we observed a temporal correlation between increased courtship around midnight and the subsequent detection of post-spawning females. It is assumed that medaka initiates courtship behavior around midnight and begins spawning with post-ovulatory females shortly after ovulation. Further research is required to determine the precise onset of ovulation in wild female medaka. Furthermore, fish use chemical cues to communicate with their surroundings and conspecifics [43–45]. Investigating the cues that male medaka use to identify spawning-ready females is expected to deepen our understanding of medaka ecology and mating behavior.

This study has two limitations. First, although many females with eggs that spawned on their abdomens were observed in our study field, no eggs were found on any of the substrates. In aquaria and artificial ponds, medaka females suspend spawned eggs on their abdomens and rub their bodies against the aquatic plants and artificial spawning beds to attach them [46]. Thus, future studies should aim to identify medaka spawning sites in the wild. By identifying spawning sites, the differences between courtship and egg attachment locations can be determined, and a deeper understanding of the reproductive behavior of medaka can be obtained. Second, we used red light for nocturnal video recordings because it is less likely to disturb the nocturnal activity of fish, including medaka [13]. It is widely used for underwater recordings at night [47]. On the contrary, however, optomotor response studies have demonstrated that several fish species, including medaka, can perceive red light [48–51]. According to a previous study on medaka, photoperiod-reversed conditions under white light cause a shift in spawning initiation time, but this effect appears 4–5 days after reversal [30]. Therefore, if red light has a large effect on medaka spawning and mating behavior as the white light, it would be expected that a similar shift in spawning time would occur in this field. In this study, however, despite the light being turned on at approximately 21:00, spawning occurred 3 h later on all four consecutive days, and no shift in the spawning time was observed. Even on the first day, when the influence of the red light was considered minimal, individuals initiated spawning at 23:54. Additionally, behavioral observations conducted in semi-natural environments using infrared illumination and cameras that are imperceptible to medaka fish indicated that medaka begin spawning and courtship at midnight (Kondo and Awata, submitted to a journal), which is consistent with the results of this study. Although no studies have quantitatively investigated the effects of red light on medaka behavior and reproduction, we speculate that red light has little effect on medaka behavior and reproduction. Further studies are required to confirm these hypotheses.

Many aspects of the ecology, behavior, reproduction, and life history of model organisms, including medaka, remain unexplored in the wild. The lack of comparative verification studies between laboratory observations and field data has led to a renewed recognition of the importance of conducting systematic field surveys and understanding their results. For example, zebrafish Danio rerio are widely used as model organisms in the laboratory; however, little is known about their life in the wild [52]. Consequently, for a long time, the reproductive behavior and life history of zebrafish in the wild could only be inferred from laboratory observations. Recent advancements in field research have led to the accumulation of knowledge regarding the factors that influence the survival and reproduction of zebrafish in the wild [53–55]. These studies have revealed that morphology, boldness, aggression, and shoal size differ between laboratory and natural environments [56–58]. Furthermore, substantial variations in these traits have been observed in wild populations. Cichlid fish have attracted attention as models for interdisciplinary research in ecology, behavior, evolution, genomics, genetics, developmental biology, and behavioral biology [59]. The advantage of cichlids is the extensive literature available on their ecology and behavior in the wild. This has led to the accumulation of knowledge in diverse research fields, including behavior and ecosystem ecology, enabling an integrated understanding of biological organization at various levels, from genes to phenotypes, individuals, populations, and communities [60, 61].

In conclusion, this study showed that medaka in the wild initiate spawning during late nocturnal hours and exhibit vigorous courtship behavior at midnight. Future research should focus on comprehensively examining the reproductive behavior of medaka over 24 h in their natural habitat and other areas to elucidate when courtship and reproduction peak. These findings will contribute to understanding the reproduction of medaka in the laboratory and progress in different fields of study, such as physiology, genetics, and conservation. Although medaka and other model organisms are invaluable in laboratories, their ecology in the wild remains largely unknown. Further field observations are essential to compare the behavioral patterns observed in natural environments with those from laboratory experiments. This approach is expected to deepen our understanding of the mechanisms underlying the biological phenomena observed in controlled settings and their adaptive significance in nature. Furthermore, understanding the ecology and behavior of the subject species in the wild will contribute to improving animal welfare, such as reducing stress, and an appropriate research setting that considers the welfare of the species should dramatically increase the reliability of research findings.

Supporting information

Acknowledgments

We thank the members of the Laboratory of Animal Sociology, Osaka Metropolitan University, for their assistance throughout the study. We also acknowledge Editage (www.editage.jp) for the English language editing.