Dispersal Kernel

Seedling dispersal patterns in two Mediterranean high-mountain species: fitting dispersal kernels along an altitudinal gradient.

Lara-Romero, Carlos (Universidad Rey Juan Carlos, URJC);Robledo-Arnuncio, J.J. (CIFOR-INIA); Iriondo, Jose María (URJC); Albert, María Jose (URJC); GarcíaFernández, Alfredo (URJC).

Recruitment capacity can affect considerably the composition anddynamics of plant communities and depends among other factors on environmental conditions and their macroscale and microscale variations. Therefore, a characterization in contrasting environmentsis necessary to understand natural variation in the processes underlying recruitment of particular species (e.g. population dynamics, dispersal capacity or gene flow among populations). We usedan inverse modelling approach to analyse seedling dispersal patterns in five populations of Silene ciliata and five populations of Armeria caespitosa along an altitudinal gradient in Sierra deGuadarrama (Madrid, Spain). Inverse modelling allows assessing seedling shadows without identifying the maternal origin of each recruit, providing parameter estimates of assumed probabilitydensity functions of seedling dispersal distances (dispersal kernels). Preliminary results showed that 2Dt and Log-normal dispersal kernels provided the best fits for both species in all theirpopulations. No relationship between population altitude and the shape and average of the best dispersal kernel emerged. However, the mean dispersal distance (MDD) varied among populations. Thesevalues are congruent with those estimated through direct and indirect methods in other species of the same genera. Estimated MDDs suggest that Armeria caespitosa and Silene ciliata have lowseed dispersal ranges. However, inverse modelling is not suitable to evaluate long-distance dispersal events, the magnitude and scale of which remain to be investigated in these species.