Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 2288-1115(Print)
ISSN : 2288-1123(Online)
Korean Journal of Ecology and Environment Vol.32 No.3 pp.253-260

Comparisons of Temperature-Related Life Histories in Two Ephemerid Mayflies (Ephemera separigata and E. strigata: Ephemeridae, Ephemeroptera, Insecta) from a Mountain Stream in Korea

Sung Jin Lee, Yeon Jae Bae1*, Il Byong Yoon, Naoshi C. Watanabe2
Department of Biology, Korea University, Seoul 136-701, Korea
1Department of Biology, Seoul Women’s University, Seoul 139-774, Korea
2Faculty of Agriculture, Kagawa University, Takamatsu 760-8522, Japan
*Corresponding author: Tel: +82-2-970-5667, Fax: +82-2-970-5669, E-mail:


Temperature-related life history adaptations in two ephemerid mayflies, Ephemera separigata and E. strigata, were studied from a mountain stream in Korea. Three study sites were chosen based on the altitudinal distributions of the Ephemera populations (upstream and midstream sites: E. separigata; downstream site: E. strigata). Samplings were taken monthly from Febru- ary 1997 to January 1998, but semimonthly during the emergence period. Water temperature was monitored hourly from May 1997 to January 1998. As results, three sites represented their own temperature regimes (mean temperature upstream: 9.39°C; midstream: 10.38°C; downstream: 12.70°C). Both E. separigata and E. strigata showed evident univoltine life history pattern and two cohorts were recognized in each population based on the analyses of body length and larval developmental stages. The emergences took place from late May to late June, but earlier at the lower sites. Larval recruitments occurred during the emergence period (summer cohort) and 5~6 months later from the emergence period (winter cohort). It was hypothesized that either egg diapause or parthenogenesis could be the cause of the two periodically different early instar recruitments. Total accumulated degree day of E. separigata was 2380 and 2668 at the upstream and midstream site, respectively, and that of E. strigata was 3220 at the downstream site. It was estimated that the egg hatching and average larval growth of E. separigata could be delayed to ca. 22 days and ca. 15 days, respectively, as the annual average water temperature drops to 1°C. Ephemera separigata showed a cold adapted life history pattern, and more discussions on the intraspecific and interspecific life history adaptations were provided with other Korean and Japanese congeners.

강원도 치악산 계류에서 하루살이 두 종 (가는무늬하루살이, 무늬하루살이: 하루살이과, 하 루살이목, 곤충강)의 온도와 연관된 생활사

이 성진, 배 연재1*, 윤 일병, Naoshi C. Watanabe2
고려대학교 생물학과
1서울여자대학교 생물학과
2Faculty of Agriculture, Kagawa University


강원도 치악산 계류에서 하루살이과에 속하는 가는무늬하루살이와 무늬하루살이의 온도와 연관된 생활사 적응에 대하여 연구하였다. 하루살이 개체군의 고도분포에 따라 3개의 조사지점(상류와 중 류지점: 가는무늬하루살이; 하류지점: 무늬하루살이)을 선정하였다. 채집은 1997. 2~1998. 1 동안 월 1회 실시하였고, 우화기에는 월 2회 실시하였다. 수온은 1997. 5~1998. 1 동안 시간대로 모니터링 하였다. 결과로서, 3개의 조사지점은 서로 다른 온도역(평균온도 상류: 9.39°C; 중류: 10.38°C; 하류: 12.70°C)을 나타냈다. 체장과 유충단계의 분석 결과 두 종 모두 명백히 연 1세대였고, 각 개체군은 2개의 무리(cohort)를 포함하였다. 우화는 늦은 5월부터 늦은 6월까지 일어났지만 하류로 갈수록 일렀다. 유충의 충원은 우화기(여름형)와 그 5~6개월 이후(겨울형)에 시작되었다. 이처럼 2회에 걸친 유충의 충원에 대하여 알의 휴면 또는 무성생식이 원인으로 작용하는 것으로 가정되었다. 상 류와 중류에서 가는무늬하루살이의 총 적산온도는 각각 2380DD와 2668DD였고, 하류에서 무늬하 루살이는 3220DD였다. 가는무늬하루살이의 경우 연평균수온이 1°C 내려감에 따라 알의 부화는 약 22일, 평균 유충성장은 약 15일 지연되는 것으로 추정되었다. 가는무늬하루살이는 냉온에 적응 된 생활사 양상을 나타냈고, 다른 한국산 및 일본산 근연종과 비교하여 종내 및 종간 생활사 적응 에 대하여 추가로 논하였다.


    Temperature has been known one of the major factors affecting temporal and spatial distributions of aquatic insects (Ward, 1992). Larvae of the mayfly genus Ephemera, known as burrowing mayflies, are common in northern temperate streams, including Korea and Japan. Among the species of the genus, Ephemera separigata Bae, E. strigata Eaton, and E. orientalis McLachlan are widespread in Korea (Bae, 1995;Lee et al., 1995, 1996), whereas E. japonica McLachlan occurs in Japan instead of E. separigata (Kuroda et al., 1984;Watanabe, 1985;Bae, 1995;Lee et al., 1995, 1996). When they inhabit in the same watercourse, they show a stratif ied pattern of altitudinal distribution occupied by E. separigata or E. japonica in the uppermost section of the stream, and followed by E. strigata and E. orientalis (Kuroda et al., 1984;Watanabe, 1985;Bae, 1995;Lee et al., 1995). It has been a keen interest to know the factors affecting the altitudinal distributions, which may be involved in the adaptation, speciation, and evolution of the species.

    The purpose of this study is to elucidate the intraspecific and interspecific life history adaptations of E. separigata and E. strigata from a mountain stream in Korea with relation to temperature as the major determinant.


    Study area

    Field study was conducted from a mountain stream in the Chiaksan National Park, located in the central area of Korea, from February 1997 to January 1998. The stream originates from the Namdaebong peak (37°18ʹ29ʺN and 128°3ʹ55ʺ E; alt. 1181.5 m) and its upper section runs through the Sangwonsa valley where the forest has been well preserved.

    Three sampling sites were chosen in the upper section of the stream based on altitude as well as the stratified distributions of the species of Ephemera. The upstream site (alt. 800 m) is a small headwater stream (width 2 m) completely canopied by the riparian forest. Only E. separigata in the genus inhabited in the site. The midstream site (alt. 470 m) is a mid-sized mountain stream (stream order 3; width 10 m) partially canopied by the forest. E. separigata was predominant at the site throughout the study period, but minor population of E. strigata also became to occur. Below the midstream site, E. separigata was gradually disappeared and E. strigata was predominant to the downstream site. The downstream site (alt. 370 m) is a mid-sized plain stream (stream order 4; width 20 m) flowing through a small village. E. strigata was predominant at the site, but minor population of E. orientalis also occurred which is more and more abundant below the site. Detailed habitat information and distributions of general aquatic insects were reported in Park et al. (1997) and Bae et al. (1998).

    Sampling and field experiment

    Qualitative samplings were monthly taken during the study period, but semimonthly taken from May to August in 1997 when the populations of Ephemera were expected to emerge (Kuroda et al., 1984;Lee et al., 1996). Larvae were collected by a hemispheric-shaped wire net with handle (diameter: 200 mm; mesh size: 0.5 mm) because they were found within 1cm depth in the sand dunes sometimes mixed with gravel and pebbles. Approximately thirty or more individuals were collected at each sampling trial except for the upstream site where the population was limited. Larvae were killed with 80% ethanol in the f ield, fixed with Kahle’s fluid in the lab for about two days, and then preserved in 80% ethanol. Larval body length and wingpad length were measured by a micrometer on the dissect microscope (10~80x).

    Water temperature was monitored at intervals of two hours at each study site from May 1997 to January 1998 by an automatic temperature recorder made by Onset Computer Corporation (Model: Optic StowAway). Thermometers were installed on the bottom of the running water area (20~50cm deep) avoiding the direct sun light. Thermometers at the upstream and midstream sites, however, were missing during the rainy season (Fig. 1-a). To calibrate this data loss, a tertiary polynomial regression was used in which indicated relatively high correlation (upstream site: r2 = 0.92; midstream site: r2 = 0.94; downstream site: r2 = 0.89) (Fig. 1-b). ANOVAs also showed no significant differences (p = 0.520) between the real temperature data and the model data at the downstream site. Because the seasonal f luctuation pattern of the water temperature would be similar in a stream, we assumed that each regression values could represent the characteristics of water temperature at each site.

    Developmental stages and degree day calculation

    Since mayfly instars have not been able to be discriminated, arbitrary larval developmental stages mainly based on wingpad development as well as genital development have been used in the analysis of mayfly populations (Clifford, 1970; McCafferty and Huff, 1978;Heise et al., 1987). In this study, we slightly modified previous devel- opmental stage analysis methods as follows. Stage I: wingpad absent, stage II: wingpad length shorter than head width; external sex organ absent, stage III: wingpad length shorter than head width; external sex organ visible, stage IV: wingpad length longer than head width; wingpad color white to gray, and stage V: wingpad color black. In the analysis of the life cycle, larval body length in a detailed scale (1 mm) was also used because it could be more informative. When two cohorts were existing in a same sampling date, they were separated from each other by Cassie’s probability paper method (Cassie, 1954). Date of egg hatching in each species was estimated from extrapolating of the mean larval size and developmental stage information in each sampling date.

    Water temperature was converted to degree day (DD) by the rectangle method shown in the following equation. Rectangle DD = (Tmax+Tmin)/2 -Tb (where Tmax and Tmin are daily maximum and minimum water temperature, respectively, and Tb is the thermal threshold for development of a given population). In this study, Tb for each population was assumed to 0°C, because it has not been studied in detail for the populations of E. separigata and E. strigata. The Julian day was also assumed from the regression of the larval development data.


    Daily water temperature

    The patterns of the yearly water temperature change between the study sites were relatively similar, with the highest temperature in late July and the lowest in early January which are typical in northern temperate region (Fig. 1). Daily water temperatures calculated from the model ranged 0~15.93°C (mean 9.39°C), 0~ 16.40°C (mean 10.38°C), and 0.12~19.24°C (mean 12.70°C) at the upstream, midstream, and downstream site, respectively. ANOVAs also indicated that the temperature differences between the sites were significant (p<0.05). In the summer time, the temperature difference between the midstream and downstream sites was larger than that between the upstream and midstream sites.

    Life cycle

    All the populations of E. separigata and E. strigata in the study sites represented a univoltine life cycle with two distinct size groups (cohorts) (Figs. 2, 3), but differed in the timing of emergence and larval recruitment between the sites.

    In the population of E. separigata in the upstream site, last instar larvae occurred from early June to late June, although the adults were actually found for some more extended period. Early instar larvae (stage I) were recruited twice for one year period. One of the early instar groups (summer cohort) occurred during the emergence period, from early June to early August, while the other group (winter cohort) occurred from late March to late April. In the midstream site, the last instar larvae occurred only in early June, and the summer and winter cohorts were recruited from late May to mid-July and from mid-January to late April, respectively.

    In the population of E. strigata at the down- stream site, last instar larvae occurred from late May to late June, but majority of them was found in late May. Summer and winter cohorts were recruited in late May and late March, respectively.

    Larval growth

    As shown in the body length and larval developmental stage analyses (Figs. 2, 3), two distinct larval growth types were indicated. Larval growth was more or less gradual in the summer cohorts, but it was relatively abrupt in the winter cohorts. Earlier instar larvae grew more rapidly, and stage III larvae were most abundant throughout the year at all sites. The body length was significantly different between the summer and winter cohorts, but the difference became less significant as closing to the emergence time (Table 1).

    Based on the regression of the degree day accumulation associated with the average body length at each larval stage in the summer cohorts (Fig. 4), the first egg hatching day was estimated around April 19, March 27, and May 5 in the population of the upstream, midstream, and downstream site, respectively. The degree day accumulation also indicated that it took 2380 DD, 2668 DD, and 3220 DD for the full growth of the larvae in the order of the sites above (Fig. 4).

    In the populations of E. separigata, the amount of cumulative DD was almost same to grow to the stage I and stage II larvae between the upstream and midstream sites. To grow to the stage III larvae, however, the amount was somewhat larger at the upstream site comparing with that at the midstream site. The population of E. strigata at the downstream site needed more cumulative DD to grow to the stage III larvae comparing with those of E. separigata. Estimated days to grow to the stage I, stage II, and stage III lar- vae at the sites in above order were 30, 59, and 185 days, 49, 77, and 194 days, and 23, 34, and 135 days, respectively (Fig. 4).


    Intraspecific life history adaptation

    Our results showed that all the populations of Ephemera in the study stream were univoltine with two distinct cohorts. Previous studies on Japanese E. strigata also showed similar results (Ban et al., 1988). Lee et al. (1996), however, was not aware of the two different cohorts in E. separigata because winter and early spring samplings were lacking in that study.

    Possible hypotheses have been presented to explain the cause of two periodically different early instar recruitment in a mayfly population. One of them is the effect of egg diapause, which may cause the delay of egg hatching in part of the egg population. Clifford (1982) argued that approximately 10% of the univoltine mayfly species, which he examined, took summer diapause, although summer diapause is not common in aquatic hemimetabolous insects. Giverson and Rosenberg (1992) also reported that an ephemerid mayfly, Hexagenia limbata, produced an egg mass about a half of which underwent diapause under cold temperature (<4°C) condition.

    Parthenogenesis could be alternative hypothesis, which may cause the delayed egg hatching in the Ephemera populations, because non-fertilized eggs hatched less successfully and developed more slowly in most cases of parthenogenesis (Humpesch, 1980). Both Hunt (1953) and Friesen and Flannagan (1976) reported the cases of non-obligatory parthenogenesis in ephemerid mayflies, Hexagenia limbata and H. rigida, respectively. Brittain (1982) reviewed many cases of non-obligatory parthenogenesis in mayflies. Watanabe and Ishiwata (1997), in a comprehensive study on the parthenogenesis in a Japanese ephemeroid mayf ly, Ephoron shigae, found that the sex ratio was the major clue in determining the bisexual and parthenogenetic populations, and that the parthenogenetic mayfly eggs would be produced in the end of the emergence period. In the populations of E. separigata and E. strigata in this study also produced more females in the end of the emergence period (unpublished data), which may indicate a possible parthenogenesis.

    Our study sites represented different thermal regimes each other (Fig. 1). The average temperature difference between the upstream and midstream sites was about 1°C and that between midstream and downstream sites was about 2.3°C. In the populations of E. separigata, the life history features such as egg hatching time, larval growth, and emergence time were similar, but those were delayed to a certain period at the upstream site. Based on our temperature and life history data (Figs. 1, 3 and 4), it can be estimated that the egg hatching and average larval growth from the stage I to the stage III may be delayed to ca. 22 days and ca. 15 days, respectively, as the annual average water temperature drops to 1°C.

    In comparison with the population of E. strigata in the downstream site, Japanese populations of E. strigata showed somewhat widespread larval recruitment and emergence time although their general life history features are similar (Kuroda et al., 1984;Ban et al., 1988). This could be caused by the geographic or altitudinal differences that may effect on the different temperature regimes.

    Interspecific life history adaptation

    In comparison with the larval growth between E. separigata at the upstream site and E. strigata at the downstream site, E. strigata required more degree days to grow to the stage I, stage II, and stage III larvae (Fig. 4). This means, if the temperature regime is same, the larvae of E. separigata may grow more rapidly than the larvae of E. strigata. Early instar larvae of E. separigata in the winter cohort were recruited earlier (January), even in the lower temperature condition (near 0°C), than those of E. strigata (late March) (Fig. 3). This also may indicate that E. separigata is a cold-adapted species comparing with E. strigata.

    Based on some previous studies on the life history of E. japonica, the uppermost stream representative among Japanese Ephemera, its general life history features are similar to those of E. separigata, but it has wider range of altitudinal distribution and longer period of emergence time (Kuroda et al., 1984;Watanabe, 1985). This may indicate that E. japonica has adapted to a wider range of thermal regime.


    This study was supported by GRANT No. 981- 0512-061-2 from the Korea Science and Engineering Foundation.



    Yearly water temperature change and its regression at the upstream (Y = 6E-6X3-0.0029X2+ 0.3332X+4.7244; R2 = 0.92), midstream (Y = 4E- 6X3-0.002X2+0.275X+6.5256; R2=0.93), and downstream (Y = 5E-6X3-0.0025X2+0.3031X+ 8.0995; R2 = 0.90) site in the study stream.


    Larval body size frequency distributions of the populations of E. separigata (a: upstream site; b: midstream site) and E. strigata (c: downstream site). (▨: stage I; ▩: stage V).


    Life cycle patterns of the populations of E. separigata (a: upstream site; b: midstream site) and E. strigata (c: downstream site). (○: summer cohort; ●: winter cohort).


    Comparison of degree day accumulation required for the larval growth of the populations of E. separigata (upstream and midstream sites) and E. strigata (downstream site).


    Comparisons of body size between the summer and winter cohorts of E. separigata and E. strigata when both of the cohorts occur.


    1. Bae, Y.J. 1995. Ephemera separigata, a new species of Ephemeridae (Insecta: Ephemeroptera) from Korea. Korean J. Syst. Zool. 11: 159-166.
    2. Bae, Y.J., S.Y. Park, S.J. Park, J.M. Hwang and J.M. Hur. 1998. Aquatic insect community in a mountain stream of Chiaksan. J. Nat. Sci. Inst., Seoul Women’s Univ., Seoul 10: 7-20.
    3. Ban, R. and KINKI aquatic insects research group. 1988. The life cycle and microdistribution of Ephemera strigata Eaton (Ephemeroptera: Ephemeridae) in the Kumogahata River, Kyoto Prefecture, Japan. Verh. Internat. Verein. Limnol. 23: 2126-2134.
    4. Brittain, J.E. 1982 Biology of mayf lies. Ann. Rev. Entomol. 27: 119-147.
    5. Clifford, H.F. 1982. Life cycles of mayflies (Ephemeroptera), with special reference to voltinism. Quaest. Entomol. 18: 15-90.
    6. Cassie, R.M. 1954. Some use of probability paper in the analysis of size frequency distribution. Aust. J. Mar. Freshwat. Res. 5: 513-522.
    7. Friesen, M.K. and J.F. Flannagan. 1976. Parthenogenesis in the burrowing mayfly Hexagenia rigida (Ephemeroptera). Can. Entomol. 108: 1295.
    8. Giberson, D.J. and D.M. Rosenberg. 1992. Egg development in Hexagenia limbata (Ephemeroptera: Ephemeridae) from Southern Indian Lake, Manitoba: temperature effects and diapause. J. N. Am. Benthol. Soc. 11: 194-203.
    9. Heise, B.A., J.F. Flannagan and T.D. Galloway. 1987. Life histories of Hexagenia limbata and Ephemera simulans (Ephemeroptera) in Dauphin Lake, Manitoba. J. N. Am. Benthol. Soc. 6: 230-240.
    10. Humpesch, U.H. 1980. Effect of temperature on the hatching time of parthenogenetic eggs of five Ecdyonurus spp. and two Rithrogena spp. (Ephemeroptera) from Austrian streams and English river and lakes. J. Animal Ecol. 49: 927-937.
    11. Hunt. B.P. 1953. The life history and economic importance of a burrowing mayfly, Hexagenia limbata, in southern Michigan lakes. Bull. Inst. Fish. Res., Michigan Dept. Conser., An Arbor. No. 4.
    12. Kuroda, T., T. Fujimoto and N.C. Watanabe. 1984. Longitudinal distribution and life cycle of the three species of Ephemera in the Kazuradani River, Kagawa Prefecture. Kagawa Seibutsu 12: 15-21.
    13. Lee, S.J., I.B. Yoon and Y.J. Bae. 1995. Altitudinal distribution of Ephemera strigata Eaton and E. orientalis McLachlan (Ephemeroptera: Ephemeridae). Korean J. Entomol. 25: 201-208.
    14. Lee, S.J., Y.J. Bae and I.B. Yoon. 1996. Life history aspects of Ephemera separigata Bae (Ephemeroptera: Ephemeridae) from a mountain stream in central Korea. Entomol. Res. Bull., Korean Entomol. Inst., Seoul 22: 73-76.
    15. McCafferty, W.P. and B.L. Huff, Jr. 1978. The life cycle of the mayfly Stenacron interpunctatum (Ephemeroptera: Heptageniidae). Great Lakes Entomol. 11: 209-216.
    16. Park, S.J., S.Y. Park, J.S. Kim and Y.J. Bae. 1997. Altitudinal distribution of Ephemeroptera, Plecoptera, and Trichoptera in a mountain stream of Chiaksan. Entomol. Res. Bull., Korean Entomol. Inst., Seoul 23: 37-42.
    17. Ward, J.V. 1992. Aquatic Insect Ecology. 1. Biology and Habitat. John Wiley & Sons, New York.
    18. Watanabe, N.C. 1985. Distribution of Ephemera nymphs in Kagawa Prefecture, Japan, in relation to altitude and gradient. Kagawa Seibutsu 13: 1-7.
    19. Watanabe, N.C. and S. Ishiwata. 1997. Geographic distribution of the mayfly, Ephoron shigae in Japan, with evidence of geographic parthenogenesis (Insecta: Ephemeroptera: Polymitarcyidae). Japanese J. Limnol. 58: 15-25.