In clonal plant populations, a number of genetically identical ramets form a genet. While coexisting ramets potentially perform independently, their behaviours not only depend on ages and sizes but are also constrained by genetic background. In this study, genet dynamics and its variability among neighbouring genets were investigated based on the ramet demography of each genet in Convallaria keiskei. Genet dynamics were first formulated as a matrix model with the two components of clonal growth (clonal reproduction) and survival-transitions between ramet size classes. Then, a statistical estimation of the matrix elements was established using three datasets: aboveground demographic censuses, belowground directional rhizomatous connections and genetic identification of ramets. Finally, genet growth rates reflecting both the changes of clonal growth and ramet size growth were estimated and compared for fundamental demographic elements among genets. Over three years of aboveground annual censuses of a 28 × 2 m plot, 2,021 ramets were identified as belonging to 28 genotypes. Belowground excavation detected 515 clonal fragments. Genet growth rate of three dominant genets varied with medians of 1.13, 1.02 and 1.05; 95% credible intervals of the posterior distributions did not overlap between the genet with the largest median and the others. The variation was caused primarily by differences in clonal growth rather than survival-transitions between size classes. Clonal growth by branching was rarer than at the tips but contributed to the maintenance of the genet. Therefore, both clonal growth frequencies and connecting patterns of ramets caused the variation of genet dynamics and established genets persist for a long time through the positive growth rates, which would contribute to maintaining a population. We also conclude that fundamental demographic elements relating to clonal growth traits (the features of individual genets) strongly impact genet dynamics.

In 2005, a long rectangular plot (28 × 2 m) was established for the present study within the original plot (coordinates X = 20–48, Y = 50–52 in Araki et al. 2007). Fifty-six 1 × 1 m subplots were also laid out in the plot. Aboveground demographic censuses were performed in June and July from 2005 to 2007. In the first year, all aerial shoots observed in the 28 × 2 m plot were carefully marked and mapped. For each shoot, the number of leaves was counted and the length of the longest leaf (leaf size) was measured. The fate of leaves and successive changes in leaf number and size at marked shoots were recorded and re-measured until 2007.

Leaf tissues of all ramets appearing in 2005 and newly emerging in 2006 and 2007 within the 28 × 2 m plot were collected after they were measured. Genotypes were detected by six microsatellite markers.

Data description

• Birth year: The first year when ramets were identified. "2005" means 2005 or earlier. For some ramets, the year was adjusted by below-ground information.
• Genotype: e-1, e-2, e-3, and e-4 show dominant genotypes; 1001-3007 are minor genotypes; "-" means genotypes were unidentified.
• Leaf no: "-" means an unemerged ramet after it was identified once or the ramet was not born.
• Leaf size: "-" means that the size couldn't be measured.

Additional information

For belowground information, we show the summary data as an Appendix of the manuscript. The raw data is available by contacting the corresponding author.