Spatial and Temporal Variations in Mortality of the Biennial Plant, Lysimachia rubida: Effects of Intraspecific Competition and Environmental Heterogeneity

• Published in: The Journal of ecology 2003-02, Vol.91 (1), p.114-125
• Main Authors: Suzuki, Ryo O., Kudoh, Hiroshi, Kachi, Naoki
• Format: Article
• Language: English
• Subjects: Animal and plant ecology; Animal, plant and microbial ecology; Biennials; Biodiversity; Biological and medical sciences; Coastal ecology; Demecology; Demography; density‐dependent mortality; density‐independent mortality; Ecological competition; Ecology; Flowers & plants; Fundamental and applied biological sciences. Psychology; ground surface conditions; growth stage; Human ecology; Mortality; Plant growth; Plants; Plants and fungi; Population ecology; Seedlings; spatial scale of mortality
• ISSN: 0022-0477; 1365-2745
• DOI: 10.1046/j.1365-2745.2003.00745.x

1 We studied mortality, growth and spatial distribution of individuals of Lysimachia rubida to elucidate causes and consequences of temporal and spatial variations in mortality. This short-lived monocarpic biennial grows in open dry habitats on rocky coastal cliffs of the subtropical Bonin (Ogasawara) Islands, Japan. 2 All individuals of two successive cohorts that appeared in a 8 × 8-m quadrat were mapped, and their fate and size were recorded at 2-3 month intervals from March 1998 to July 2000. We analysed spatial and temporal changes in the relative importance of plant size, local densities of neighbouring plants and ground surface conditions to determine plant mortality. We also examined spatially non-random mortality and its spatial scales. 3 Patchy seed dispersal from scattered mother plants resulted in a clumped distribution of seedlings. During the growth cycle of L. rubida, we recognized two growth stages that differed in the relative importance of the factors affecting plant mortality. 4 At the earlier growth stage, after germination in March until July, mortality was caused mainly by local density. Smaller plants within areas of high local densities were more likely to die. The spatial patterns of surviving and dead plants tended to be regular and aggregated, respectively, in comparison with those expected from random mortality. The spatial scale of aggregation of the dead plants was 5-25 cm, corresponding to the distance of direct interactions between neighbours. 5 At the later growth stages, in October until May of the following year, mortality was affected mainly by ground surface conditions. Both surviving and dead plants were spatially aggregated at spatial scales of 20-100 cm. Plants tended to survive in micro-habitats of smaller gravel, which had a greater water-holding capacity. 6 This study also showed the consequence of spatial variations in mortality on population structures of L. rubida. Patchy seed dispersal and subsequent temporal and spatial variations in mortality are likely to contribute to the scattered distribution of L. rubida individuals observed at the study site.