Abstract:
Bracken (Pteridium spec.) is one of the most wide-spread weeds, especially where fire has been used for forest clearing or maintenance of agricultural areas, in particular of pastures. Taxonomically, it is considered an aggregate that separates into a northern hemispherical and a southern, tropical complex. The taxonomic ranks of the members of these complexes is still a matter of debate. Different from the extensively studied northern bracken, the knowledge of ecology and control measures of the tropical species is still fragmentary. A research team funded by the German Research Foundation (DFG) is working in an ecosystem study in a narrow valley of the eastern range of the South Ecuadorian Andes. Part of this ecosystem are pastures, replacing the natural rain forest. These pastures are in part heavily infested by bracken and many of them have been already abandoned for this reason. Several groups of the research team work on various aspects of the bracken invasion, e.g. on ecophysiology of bracken, soils and microclimate of the study area, mycorrhiza, and on insect interactions with bracken. The main aims of my thesis were (1) identification and population structure of bracken, (2) ecology of tropical bracken with particular emphasis on its survival of bushfires, and (3) development of an effective bracken control strategy and subsequent re-pasturisation of abandoned areas. The bracken vegetation of the study area consists of mixed stands of Pteridium arachnoideum (KAULF.) MAXON and P. caudatum (L.) MAXON with a proportion of 3:2, and of a hybrid (ca. 2%). Identification was by leaf morphology, allozyme analysis, comparison of particular chloroplastic DNA sequences, and analysis of four genomic and one plastidic microsatellites. Dominance of P. arachnoideum was explained by the fact that P. caudatum, as a lowland species, reaches its upper altitudinal limit in the research area. Analysis of heterozygosity indicated a higher genetic stability of the diploid P. arachnoideum population as compared to the allotetraploid P. caudatum population. Spatial extension of the individual clones is apparently much smaller than reported for members of the northern bracken, indicating higher significance of sexual reproduction for the tropical fern in comparison to vegetative propagation by rhizome fragmentation. Four weeks after burning the natural rain forest, vigorously sprouting bracken sporophytes were observed. These developed from gametophytes, which germinated from the wind dispersed spores. Fast growth of the young sporophytes established the fern in the areas. After planting pasture grass, bracken was supported by repeated burning of the areas. In the long run, the grass was outcompeted by the fern possibly due to weakening of its vitality by burning and grazing, and the areas have been abandoned. The density of bracken fronds in a settled bracken area of our research site remained constant over years with only small deviations caused by particular weather situations. Since this balance holds also for patchy fern canopies, it is assumed that this is due to nutrient shortage of the soil. Most probably, a new leaf can only develop from the nutrients remobilized from a senescing old leaf. Two to three months after a fire, an explosive emergence of new leaves was observed at rates, which substantially exceeded those under undisturbed growth. The newly formed leaves showed an extended life-span, which was attributed to a better nutrient supply from the ash. Subsequent self-thinning reduced the density of the leaves to a stable level within two years. In a laboratory experiment, the effects of heat pulse by a simulated bushfire on the bracken rhizomes were investigated. Separated long and short shoots were heated for a short time either in a water bath or embedded in soil. Subsequent to this heat pulse, they were cultivated in original soil. Short shoots showed a significantly higher heat resistance (up to 80°C) than long shoots (up to 60°C). In addition, the short shoots showed elongation growth and an enhanced frond production, whereas long shoots were not stimulated by the heat pulse. In a bracken control experiment, thirteen control measures (cutting of the fronds, several herbicides, covering with plastic foil and combinations thereof) were applied over a time-period of 23 months. Each treatment was repeated six times and the effects were recorded monthly. Quarterly cutting of the leaves as well as treatment with a customary herbicide mixture (picloram and metsulforon methyl) were the most effective treatments resulting in a reduction of the standing biomass by 65%. Monthly records of the resprouting bracken was necessary to work out the minimum number of treatments required for a clear control effect. For the five most efficient treatments among two to four applications were necessary. However, complete eradication of bracken was not possible. For re-pasturisation, the common pasture grass Setaria sphacelata was planted on the treated areas within a long-term experiment. After nearly two years of observation, the system had stabilized with a cover of S. sphacelata of 75% and of bracken of below 40%. This result demonstrated that the competitive strength of S. sphacelata was sufficient to control bracken once weakened by control treatments. The long-term experiment and, in addition, an experiment in which a gradually bracken-infested area is subjected to controlled burning, are continued.
