Print PageMany invasive plant species propagate clonally, so that dispersal is strongly localized. This growth pattern implies that an invader will likely advance spatially as a roughened, but discernable front. Spatial correlations in the extent of advance along the front induce relationships between the length of the front L and invasion statistics. Applying kinetic roughening theory, the velocity of invasion should decline with L, and both the variance of advance about the front’s mean position (termed "roughening") and relative position of the maximal incursion (the “front runner”) should show a positive power law scaling with L. We tested these predictions in a field experiment where white clover (Trifolium repens) and perennial ryegrass (Lolium perenne) competed for space preemptively. Both species reproduce via vegetative growth when mowed regularly, and soil nutrients levels assured that clover would be the superior competitor. We established plots with L =1, 2, 4, 8, and 16 meters, each replicated four times. Plots were initially divided lengthwise, separating monocultures of the two competing species. After dividers were removed, spatial advance was monitored monthly by recording the position of the most advanced clover leaf in each 1 cm section of plot length.
Results/Conclusions
Kinetic roughening theory predicts that the front’s velocity and variance about the average frontal position increase initially, until equilibrating at a level scaling with the length of the front. During the six months of experimental growth, clover advanced within each plot with mean velocities ranging from 20 to 32 cm/month. Variance about the front’s mean advance increased during the first 4-5 months and then equilibrated statistically. As our model for invasive growth predicts, at equilibrium the mean position of the "front runner" relative to the average height scaled as a power law, with slope 0.52 +/− 0.13 (p < 0.001). Asymptotic roughening of the front also scaled with front length L as a power law function, with slope 0.61 +/− 0.28 (p < 0.001). Invasion velocity did not exhibit a significant relationship to front length. This effect was expected to be small over the experimental range of plot sizes, and as a result our study may have lacked statistical power sufficient to detect an effect of L on velocity. This study indicates that kinetic roughening theory may be a useful way to predict and describe the advance of ecological fronts.
