Population-biomass dynamics and the absence of - 3/2 self-thinning in the clonal weed Salvinia molesta

Abstract In the field, the population density (n) of crowded, living ramets of Salvinia molesta ranged from 2500 m ^?2 in nutrient-poor water to more than 30 000 m^?2 in nutrient-rich water. Biomass of living shoots (B) in some localities exceeded 600 gm^?2 dry weight (150 tonnes ha^?1 fresh weight) and shoots plus roots of living and floating dead material exceeded 1600 gm^?2 dry weight (400 tonnes ha^?1 fresh weight). The ultimate limit B= 10⁵n^0.5 suggested by previous authors was not exceeded. The highest n and smallest ramets occurred in nutrient-rich water and B was a linear function of n. Field experiments confirmed no effect of n, but a strong effect of nutrient availability, on ramet weight. In a glasshouse experiment, ramet populations did not ‘self-thin’ along the log log B-n trajectory of slope -0.5, which has been suggested for aclonal plants, but followed trajectories having slopes closer to +1.0 before halting at the ambient B-n limit. Nutrient concentrations in the plants increased during the experiment, causing ramet size to decrease and slopes of log log B-n trajectories to depart from +1.0 by amounts which reflected the degree to which nutrient concentration had changed. Under constant conditions, mean ramet size is expected to remain constant and log log B-n trajectories to travel in a straight line of slope +1.0 to the ambient B-n limit. This behaviour of salvinia, whose ramets consist of a single structural module, is contrasted with that expected of clonal plants whose ramets are made up of multiple structural modules. A logistic model explained the increases in n and biomass of living and dead ramets when ramets were assumed to have constant longevity, constant time to decay, and when both the B-n limit and size of ramets were dependent on nitrogen content of the plant. There was no indication that physiological integration within phenets halted production of new ramets when carrying capacity was reached and new ramets continued to be produced on top of older ramets. Rates of ramet mortality were equal to, and dependent on, rates of ramet natality.