Habitat Characteristics and Ecotypic Differentiation in Cicuta virosa, an Endangered Species in Korea

Abstract

Cicuta virosa L. (water hemlock) is a perennial herb native to northern and central Europe, northern Asia, and northwestern North America. Its population size has decreased in Korea which is a southern marginal area of distribution and C. virosa listed as an endangered species in Korea. Only four habitats of C. virosa are known in South Korea. Pyeongchang (PC) was a fenced wet meadow with water shortage in spring and Hoengseng (HS) was about 12 year old abandoned paddy field and narrow streamlet holding shallow water level through the year. Gunsan (GS) was a 1m deep reservoir with floating mat composed of organic matter in 2° south of HS and PC and very low altitude. Last, Daegi-ri (DG) was an agricultural waterway recorded as a C. virosa habitat for the first time in South Korea but now, there are less than 10 individuals of C.virosa. To conserve and restore habitats of this plant, I investigated water and soil environmental characteristics and vegetation at four habitats during the growing season. The C. virosa habitats differed in community structure, water and substrate properties, and water regime. Although the total distribution ranges of the water and soil environments for C. virosa were wide and overlapped with the optimal environmental range of distribution of accompanying species, the optimal water level range for C. virosa was defined as 7 ± 3.5 cm. Water level was adjusted by substrate structure such as a mound of P. japonica and a floating mat comprised of accompanying species. A floating mat could be an aid to maintain an optimal and stable water level in deep or fluctuating water and to prevent strong competition with prolific macrophytes. The GS sampling site, which had floating mats, could be a good model for C. virosa conservation in a warm temperate region, whereas the PC sampling sites, which experienced a water shortage in spring, provided a clue about the decline in C. virosa population size. Based on the first survey result, I surveyed seedling distribution in GS and HS habitats and examined growth increment in mesocosm by planting C. virosa on artificial mats and soil to investigate the function of floating mat in seedling establishment and population sustenance of C.virosa. Seedling density and coverage on floating mats and their edges, with high level of solar radiation and sufficient water level, were significantly higher than those on soil. In the mesocosm experiment, shoot and root dry-weight in soil was double to those in artificial mats, and these significant difference reflected nitrogen availability in the substrate. However, there was no significant difference in the numbers of flower stalks, or tiller or stem width, despite the difference in nutrients level of the three substrates. I also observed life history variation between GS and HS populations. These results showed that floating mat was advantageous to the establishment of seedlings and to the expansion of C. virosa population, and high allocation to reproduction of C. virosa sustained coexistence with prolific macrophytes on floating mat even though the overall growth on floating mat was worse than that on soil. Using floating mat could be a way for C. virosa population sustenance and conservation. Through the advanced two habitat surveys, I found out differences not only in habitat environments (e.g. altitudes, climates, water regimes and nutrient availability) but also in leaf morphology and flowering time among populations. I considered that those three populations (PC, HS and GS) could be classified as ecotypes and figured out intraspecific variations of C. virosa populations as ecotypic differentiations in morphological and physiological features of seed and seedling by tests of germination responses to light and temperature and the sensitivity of seedlings to dryness and shade. I also analyzed which traits were critical to separate three populations and interpreted relationships between discriminating traits and habitat environments. The seed length of GS (2.04 ± 0.03 mm) was significantly shorter than that of PC (2.44 ± 0.05 mm) and HS (2.60 ± 0.03 mm). Seed weight of GS (0.83 ± 0.01 mg) was significantly lighter than that of the others (PC: 1.47 ± 0.02 mg, HS: 1.33 ± 0.02 mg). PC and HS seedlings had larger root: shoot ratio values and GS had significantly higher relative growth rate (RGR) through 8 weeks of growth. GS seedlings were more sensitive to dryness than PC and HS in leaf chlorophyll contents, specific leaf area (SLA) and RGR. The pattern of germination responses was similar among the three populations but germination rates were very different. Maximum germination rates were 63.0%, 23.9% and 96.9% in PC, HS, and GS, respectively, under 28/18°C/14h photoperiod. The nonviable seed rate of GS (18.05±0.70%) was significantly higher than that of the others (PC: 3.07±1.63%, HS: 3.44 ± 0.62%) in the tetrazolium tests result. Seed mass which was correlated to other plant traits (i.e., root: shoot ratio, RGR) and germination rate were the most discriminating variables in the discriminant analysis. Seed weight was negatively correlated with temperature in January and water level and germination rate was significantly related with water level and water temperature in August. Based on these results, I suggest that these three C. virosa populations from different habitats are ecotypes.