Physical disturbance has often been invoked to control genotypic diversity in sessile clonal organisms, yet experimental evidence is lacking. I studied the effects of physical disturbance on genet dynamics and genotypic diversity in a clonal marine angiosperm, Zostera marina (eelgrass). In replicated plots of 1 m2, the vegetation canopy was removed in gaps of zero (control), 25%, 50% and 75% of the area (n = 6 replicates). Before removal and during two consecutive years, the genotypic composition was determined using genetic markers (DNA microsatellites) in a 5 x 5 pixel grid per plot. An aggregate index of genet dynamics summarizing recruitment, increase, loss and decrease of clones was maximal at intermediate disturbance levels (quadratic polynomial P = 0.02). Physical disturbance also increased the occurrence of new genotypes, possibly reflecting recruitment (linear model, P < 0.05). Contrary to expectations, there was no competitive advantage of more heterozygous genotypes over less heterozygous ones. In the absence of disturbance, in particular, clones with lower individual heterozygosity were more likely to increase in area over a 1-year time period than more heterozygous ones, while there was no such correlation in plots with disturbance (logistic model, P(disturbance x heterozygosity) = 0.036). Undisturbed plots revealed background recruitment independent of canopy gaps, suggesting that Z. marina exhibits a strategy of continual recruitment. Effects of experimental disturbance (linear or quadratic) on clonal diversity were not detectable. Instead, initial (pretreatment) clonal diversity accounted for between 68% and 91% of the variance in diversity, indicating remarkable resilience of genotypic diversity in the face of physical disturbance.