Effects of Temperature, Nutrients and a Beetle on Branch Architecture of the Floating Weed Salvinia Molesta and Simulations of Biological Control

Publisher Website
Google Scholar
Abstract
(1) Salvinia molesta is a free-floating fern which sometimes forms thick mats on the surfaces of tropical freshwaters. The species is sterile and changes in population density result solely from production and loss of ramets on branched rhizomes. Ramets can survive independently but they are held together in colonies until rhizomes break. (2) Maps were drawn of rhizomes collected from the field and at the start and end of growth experiments performed in the laboratory and in the field. Data from maps were used to calculate two indices of branching, percentage of buds in particular positions which had grown into new ramets, rates at which ramets were added to the main axis and relative growth rates for total number of ramets. (3) In the laboratory, temperature had no effect on branching but had a strong effect on the rate at which ramets were added to the main axis which peaked at 0.43 ramet d-1 near 30.degree.C. Growth of buds at 30.degree.C declined on excised ramets older than 7 ramet-generations, but still occurred at 10 generations; at all ages growth increased with the availability of nutrients. (4) Maps were analysed from 416 measurements of growth in the absence of self-crowding at seven field sites located from near the equator to 33.degree.S. Branching was significantly correlated with the nitrogen and phosphorus content of the plant but not with the potassium content or temperature. The rate at which ramets were added to the main axis was significantly correlated with both temperature and nitrogen content, and a maximum of 0.41 ramets d-1 was observed. Fitted curves suggested that when the plant contained 2.65% N dry weight, all first-rank lateral buds would grow and the rate of extension of the main axis would be limited by factors other than N. (5) Crowded plants branched less than uncrowded plants, probably because crowded plants contained less N. Shading had no effect on branching, while destruction of buds by the weevil Cyrtobagous salviniae resulted in compensatory growth by increased branching. (6) In a 28-week study at one site, labelled ramets survived for a mean of 15.3 weeks and were last seen twenty ramets from the apex of the main axis. In a broader survey, the first signs of senescence were observed at between three and eight (mean 5.7) ramets from the apex. There was no effect of NPK content but senescence was found closer to the apex in crowded plants and further from the apex when rates of growth were high. (7) A difference equation model was extended to include variable probabilities of bud growth, death of ramets due to old age, colony fission, destruction of buds and ramets by C. salviniae and growth in compensation for damage. A stochastic simulation model was built with the added ability to modify probabilities of growth or death for each bud and ramet according to age and position within a colony. (8) The simulation model accurately mimicked branching growth of plants in the field and predicted that a population density of 300 adults + 900 larvae of C. salviniae m-2 should be sufficient to achieve biological control of most infestations of S. molesta. This agreed well with cases of successful biological control observed in the field.