INTRODUCTION
Selection of oviposition sites is important for amphibians because amphibian larvae have to develop without parents and spend the larval stage in a limited space (Matsushima and Kawata, 2005). Therefore, the various characteristics of an oviposition site have a major impact on the success of egg hatching, larval development, and larval survival (Resetarits and Wilbur, 1989; Mousseau and Fox, 1998), and amphibians take into consideration a variety of environmental factors when selecting an oviposition site (Skelly et al., 1999; Halverson, 2003; Semlitsch and Bodie, 2003; Porej et al., 2004; Marsh et al., 2005).
The factors influencing the selection of amphibian oviposition sites are diverse. Biological factors include the distribution of conspecifics and predators (Howard, 1978; Resetarits and Wilbur, 1989; Crump, 1991; Dillon and Fiano, 2000; Halloy and Fiano, 2000; Blaustein et al., 2004), whereas non-biological factors include water temperature, humidity, water depth, water volume, soil, and vegetation (Howard, 1978; Seale, 1982; Huk and Kuhne, 1999; Reich and Downes, 2003). Although a site in which conspecifics have previously spawned provides a strong indication of the quality of the habitat (Matsushima, 2005), it also represents a negative factor, in that it increases competition and decreases the growth rate of offspring (Skelly and Kiesecker, 2001). Generally, oviposition sites preferred by amphibians are sites where the temperature is adequate, the risk of drying out is low, food sources are available, and predators can be avoided (Petranka et al., 1994; Viertel, 1999; Binkley and Resetaris, 2003; Ficetola and De Bernardi, 2004; Resetarits, 2005; Rudolf and Rodel, 2005).
Rana huanrensis (Fei, Ye and Huang, 1990) was identified as a species distinct from the similar species Dybowski's Brown Frog (Rana dybowskii) based on the results of morphological, ecological, and genetic analyses (Yang et al., 2000). Morphologically, the diameter of the eardrum is one-half or less the length of the eye, the finger and toe tips are not rounded, and the flippers are very well developed (Song et al., 2005). The eggs are spawned in an egg clutch (a cluster of eggs) around rocks submerged in shallow water or small gravel areas on the edges of valleys. The eggs within a clutch are attached to each other and notably smaller and harder than those of R. dybowskii (Seo, 2011).
R. huanrensis is one of the earliest spring spawning species in Korea, spawning upstream of valleys between February and April. This species is classified as an explosive breeder because it spawns explosively in a short period in early spring (Waldman, 1982; Yoo and Jang, 2012), and can be classified as a stream breeder because it spawns upstream of valleys. Amphibians spawning in early spring maximize the transformation potential of tadpoles before pools dry out. Furthermore, because many potential predators are not active in early spring, sufficient larval development time is guaranteed (Calef, 1973; Heyer et al., 1975).
Spawning in early spring in valleys can, however, be a very risky breeding strategy. Temperatures in early spring can drop rapidly and egg clutches can occasionally freeze (Wright, 1914;Wright and Wright, 1949; Licht, 1971; Waldman, 1982). Furthermore, the current is strong in the upstream regions of valleys, and the habitat is prone to flooding when snow melts or there is rain. In addition, the water noise in valleys is a significant limiting factor for frogs, the males of which court females using croaking sounds.
R. huanrensis is widely distributed in the mountain valleys of Korea; however, there have been few ecological studies on this species and studies on the environmental characteristics of oviposition sites are nonexistent. Moreover, there have been no studies on how these frogs overcome the cold weather of early spring and avoid the loud ambient water noise that interferes with courtship sounds. Therefore, the goal of this study was to investigate the general physical characteristics of oviposition sites of R. huanrensis and, in particular, elucidate the water temperature and sound environment of oviposition sites in noisy valleys in early spring.
MATERIALS AND METHODS
1.Study Site
The study sites were located within the major valleys of Chiak National Park in Gangwon-do (Figure 1). Chiak National Park is a mountainous national park in central Korea with altitudes reaching 1,288m and a total area of 175.7km2. Because the valleys in Chiak National Park are well developed and water is plentiful in the spring, R. huanrensis spawns in most of the valleys. In this study, we investigated 75 R. huanrensis oviposition sites: 25 sites in Guryongsa valley in the north, 32 sites in Bugok-ri valley in the east, 9 sites in Geumdae-ri valley in the south, 6 sites in Sangwonsa valley and 3 sites in Heungyang-ri valley in the west. The study sites were a restricted access area. And there were no artificial facilities. Therefore, there were no impacts of anthropogenic noise.
2.Survey Methods
For the oviposition site survey, the status of locations where R. huanrensis egg clutches were found was surveyed by searching the mainstream of the valley and adjacent water pools on bedrock while walking along the valley. Eggs having strong viscosity and tightly adhere to bedrock or tree branches were identified as being those of R. huanrensis (Seo, 2011). The field survey was conducted for 7 days between March 23 and April 24, 2015, and the survey time was between 9AM and 5PM. The survey was conducted by two teams. 1 team consisted of two persons.
The survey items of R. huanrensis oviposition sites were the number of egg clutches, location, water temperature, surface area, water depth (the water depth of spawning point), floor type, vegetation, and sound environment. The total number of egg clutches within an oviposition site was surveyed. The oviposition sites were divided into three types: isolated pool, mainstream, and waterfall. Water temperature was measured at the depth where egg sacs were attached, and the water temperature of the adjacent mainstream was also measured. For the waterfall and mainstream types, the temperatures of oviposition sites and the mainstream were recorded as being the same. Water temperature was measured using a Testo 05601113 thermometer (accuracy ±1°C). The area of oviposition sites was calculated using the ellipsoid area equation (area = π × radius of the long axis × radius of the short axis) based on measurements of the length of the long and short axes of an oviposition site obtained using a tape measure. Water depth was measured from the central point of an egg sac using a tape measure. The floor types of oviposition sites were recorded as being foliage, rock (the presence or absence of the small rocks such as gravel in the pool), and soil types.
The type of vegetation overhanging oviposition sites was recorded as being coniferous, deciduous, fully open, rocks. The crown density of vegetation was, however, not measured since the spawning season was early spring and leaves had yet to sprout.
For each oviposition site, we recorded sounds in the immediate pool area (water surface) and in the area surrounding the pool (1.5m from the water surface) for 1 minute each using an Idam PRO U11 digital voice recorder. MP3 320 kbps was used as the recording file format. We used two 10-mm directional microphones built into the recorder, each fitted with wind screens. For analysis of the sound environment, the sound volume (decibels) was measured using Adobe Audition CC (version 6.0) (Ki and Sung, 2014; Ki et al., 2014).
3.Statistical Analysis
The survey items for the physical environment of oviposition sites were analyzed using descriptive statistics, frequency analysis, T test, and correlation analysis. Descriptive statistics were calculated for the number of egg clutches; the area, water depth, and water temperature of the habitat; and the sound volume of oviposition sites. Frequency analysis was performed for the location and floor type of oviposition sites, and Pearson’s correlation analysis was conducted for the number of egg clutches, water temperature, water depth, and sound environment. A paired sample t-test was performed to compare the water temperature and sound environment between the oviposition sites and the mainstream. IBM SPSS Statistics (Ver. 21) was used for statistical analysis.
RESULTS AND DISCUSSION
1.Physical Characteristics of Oviposition Sites
A minimum of 1 to a maximum of 55 R. huanrensis egg clutches were observed in oviposition sites with an average of 10.3. The average area of oviposition sites was 4.7 m2, ranging from 0.03 m2 to 45.5 m2. The average water depth was 10.1 cm, ranging from 1 cm to 30 cm. The average water temperature of oviposition sites was 9.3°C, ranging from 5°C to 20.9°C.
The average sound volume of oviposition sites was -45.3 dB, ranging from -65.3 dB to -16.4 dB. Most of the oviposition sites were found in the vicinity of rocks, which tend to reduce the noise from water in the mainstream, particularly in large deep pools. This contrast with eggs laid in the waterfall and mainstream sites, where the noise of rushing water is very prominent.
The oviposition site location types included 67 isolated pools (89.3%), 5 mainstream sites (6.7%), and 3 waterfall sites (4%). The results of this study indicated that R. huanrensis favors small pools in which the risk of eggs being swept away is low, even in valleys with a rapid current. The waterfall type refers to sites where water has accumulated. Even in the 5 mainstream oviposition sites, the water flow was stagnant due to the surrounding microenvironment.
The floor types of oviposition sites included 24 sites (32%) covered only with foliage; 38 sites (50.7%) with foliage and rocks; 12 sites (16%) with foliage, rocks, and soil; and 1 site (1.3%) with foliage and soil. Previous studies have determined that the amount of foliage and aquatic plants are important factors in the selection of an oviposition site (Laurila, 1998; Marsh and Borrell, 2001; Wang et al, 2007). Foliage and rocks were found in all oviposition sites of R. huanrensis. This is probably because R. huanrensis chooses oviposition sites where the adult frogs can hide from predators and which provide larvae with food. The frogs did not spawn at sites that lacked hiding places on the floor.
The vegetation overhanging oviposition sites included conifers at 15 sites (20.0%), deciduous trees at 52 sites (69.3%), and mixed conifer and deciduous trees at 3 sites (4.0%). Two sites (2.7%) were overlain by rocks, whereas 3 sites (4.0%) were in the open. The vegetation of the conifer-type sites consisted of 1 species of pine tree, whereas a variety of deciduous trees were distributed in the deciduous tree-type sites. The reason why the proportion of deciduous trees was high is that the oviposition sites were located near valleys primarily inhabited by wetland deciduous trees. Since the survey was conducted in early spring, the vegetation was still not in leaf, and this may have allowed sufficient sunlight to filter through to the oviposition sites, thereby raising the temperature.
2.Correlation between the Number of Egg Clutches and Physical Factors of the Oviposition Sites
Correlation analysis of the number of egg clutches and physical factors of the oviposition sites revealed that the number of clutches was positively correlated (p < 0.05) with the water temperature and negatively correlated with the sound volume of oviposition sites (p < 0.05)(Table 1).
Oviposition site noise showed no correlation with the water temperature and water depth. The Huanren brown frog appears to spawn lots of eggs when there is less noise than the surroundings, for effective delivery of mating calls, and when the water temperature is higher. This indicates that R. huanrensis favors a higher water temperature and a lower sound volume when selecting an oviposition site.
No correlation was found between the number of egg clutches and the area or depth of water pools. This finding contrasts with previous studies on frogs, which found that the area of water pools was an important factor in the selection of oviposition sites (Laurila, 1998; Marsh and Borrell, 2001; Wang, 2008). For R. huanrensis, various factors, including accessibility of the location, safety of the floor type, higher water temperature, and nature of the sound environment, played a more important role than the area of a water pool in the selection of oviposition sites.
In this research, the correlation between the clutch number and water temperature was not high. This was due to large variation in water temperature because measurements were taken within a wide time interval, from 9am to 5pm.
3.Characteristics of the Water Temperature of Oviposition Sites
Water temperature is a core factor that affects the development of amphibian eggs and larvae (Sjögren et al., 1988). The water temperature of R. huanrensis oviposition sites was on average 2.2°C higher than that of the mainstream (Table 2, Figure 2). There are a variety of reasons for this difference. First, while the mainstream maintains a constant temperature because the water is flowing, the water temperature in oviposition sites rises as the atmospheric temperature rises because the water is shallow and stagnant. In addition, some of the water pools are located in the recesses of large rocks and these sites provide favorable conditions for larval development because bedrocks are easily warmed by sunlight, which raises the water temperature of the pools at the same time.
The depth of water where R. huanrensis spawned varied depending on the area and floor type, but the depth where egg clutches were located was approximately 10 cm. Seale(1982) reported that Rana sylvatica benefitted from a warmer temperature by selecting medium depths in pools, where water is warm, as the oviposition site because water temperature decreases as the depth increases. R. huanrensis benefits from a warmer temperature by spawning in shallow locations regardless of the area of an oviposition site or the water depth.
R. huanrensis tended to spawn eggs tightly clustered together. Waldman (1982) reported that eggs are spawned in clusters for protection from the low temperature in early spring. This is because eggs in the center are protected from the low water temperature at the periphery and are warmer than eggs at the periphery. Although eggs at the periphery of an egg cluster are colder than those in the center, they still tend to be warmer than single eggs (Waldman, 1982). Therefore, it is thought that R. huanrensis spawns eggs in clusters to maximize the thermal insulation effect.
4.The Sound Environment Characteristics of Oviposition Sites
Male amphibians use a variety of strategies to court females in environments with heightened natural or artificial noises. The Chinese concave-eared torrent frog (Amolops tormotus) uses ultrasound to overcome the sound of water in valleys (Feng et al., 2006). Borneo tree-hole frogs use empty tree holes as a megaphone to amplify the courtship sound (Lardner and Bin Lakim, 2002). The Taiwanese Mientien tree frog makes croaking sounds in sewage pipes in cities to amplify its courtship sound. The sound volume of frogs in sewage pipes is 4 dB higher than that in the outside environment, and the sound persists for 10% longer (Tan et al., 2014). In addition to audible signals, the Bornean rock frog (Staurois parvus) uses visual signals in a noisy valley environment. The visual signals include the behavior of spreading the hind and front legs and opening the mouth (Grafe et al., 2012).
When exposed to artificial noises, frogs sometimes change the courtship sound so that it is not masked by artificial noises (Given, 1999; Bee and Swanson, 2007; Glenn and Fahrig, 2010). Species that are exposed to artificial noises expend considerable energy in producing a sound that can be heard against the background noise. Because the sound energy is proportional to the square of the sound amplitude, certain species change the sound to a higher frequency instead of increasing the sound amplitude (Glenn and Fahrig, 2010). On the other hand, some frogs in noisy highway areas croak where the vehicle noise is blocked off to avoid the acoustic interference (Ki and Sung, 2014).
Compared to the mainstream, the sound volume of R. huanrensis oviposition sites was 6.9 dB quieter (Table 2). This is a high value considering the 4 dB of sound amplification achieved by Taiwanese Mientien tree frog in sewage pipes (Tan et al., 2014). The breeding sites of R. huanrensis are located upstream of the valley where the water noise is very loud. Noisy water interferes with the courtship calls of frogs, making it difficult for males to convey courtship calls to females. The results of this study suggest that R. huanrensis males select water pools that are relatively quieter than the surrounding area, thereby enabling them to effectively transmit courtship calls to females.
Figure 2 is a box plot that compares water temperature and sound volume between the oviposition site and the main current. For water temperature at the oviposition site, there were 3 outliers at around 20°C, and this was because we measured water temperature in the afternoon at a time when the air temperature is high. In other words, we could know that, compared to the main current, whose water temperature is constant, the water temperature at the oviposition site rises higher in the daytime (Figure 2-A).
In the case of oviposition site sound volume, the maximum and minimum volumes at the oviposition site and the main current volume were similar but we can see that the 1st quartile, 3rd quartile, and median for the oviposition site are lower than those for the main current. Because the oviposition site’s noise source is the main current, we can say its noise is dependent on the main current. However, because noise is reduced, since the oviposition site is surrounded by bedrock, the mean sound volume was measured to be lower than that of main current. An outlier was measured in 1 place for the main current, and this place was an area where there was a small waterfall and the sound volume was measured to be higher than that of other places (Figure 2-B).
5.Conclusions
Collectively, the results of this study suggest that R. huanrensis selects micro-habitats in valleys optimized for breeding, which avoid rapid currents, sudden increases in water flow, low temperatures in early spring, and loud water noise. More specifically, most of the oviposition sites were located in isolated water pools, in which the probability of eggs being swept away by flood water is low and the water temperature is higher than that of the mainstream, thereby shortening the egg-hatching period. The egg clutches found in the oviposition sites were laid in shallow areas on the edge of pools at depths of approximately 10 cm depth, even in large deep water pools, possibly due to the advantage of earlier hatching, since water temperature rises faster in shallow water. The bottom of oviposition sites was covered with foliage and rocks, which may offer effective escape from predators during the breeding season. The boundaries of the oviposition sites were mostly surrounded by rocks that impede the loud water noise of the mainstream. R. huanrensis males are thought to have chosen these sites because they are relatively quieter compared to the mainstream, and therefore they can effectively transmit their courtship calls to females without them being mask by the sound of water in the valley.
This study is significant because, to the best of our knowledge, this is the first time that the sound environment of amphibian oviposition sites has been characterized. Previous studies on amphibian oviposition sites have mainly analyzed location, size, water temperature, water depth, floor type, and vegetation type. Moreover, studies on amphibian behaviors in environments where the ambient noise is extreme, such as valleys, have mainly focused on how frogs overcame the noise through croaking and visual actions. However, in this study, we found that R. huanrensis breeds after finding relatively quiet locations that are free from ambient noise, thereby enabling males to more effectively convey their courtship calls to females.
This paper can provide data to support future habitat conservation of the Huanren brown frog. When creating Huanren brown frog oviposition sites, we must create puddles instead of flowing water, and the water depth should be shallow, at around 10cm, in order to maintain a high water temperature. In particular, we should create a space that can block the sound of water and other noise from the surrounding environment when creating or managing a Huanren brown frog breeding environment.
