PLANT LIFE-HISTORY ATTRIBUTES - THEIR RELATIONSHIP TO DISTURBANCE RESPONSES IN HERBACEOUS VEGETATION

1 Species composition and disturbance state (level of soil disturbance , grazing and water addition) were recorded for 120 (30-m2) plots in t emperate grassland of the New England Tablelands (Australia). Three di fferent classificatory schemes were used, based on (a) life-form (sens u Raunkiaer); (b) dispersal unit morphology and (c) vegetative reprodu ction. We analysed the effects of the different disturbance types and intensities on number and proportion of species belonging to these gro ups. The results were used to describe the spectrum of life-histories likely to be present at sites in different states of exogenous disturb ance. 2 Soil disturbed sites had proportionately more therophytes, ver satile/flat rosettes and wind dispersed species. Two features are like ly to lead to success in soil disturbed sites: ability to colonize ope n space (e.g. therophytes, wind dispersal) and capacity to capture res ources effectively (e.g. flat architecture of rosettes). Non-disturbed sites had more geophytes, chamaephytes, phanerophytes, proto-hemicryp tophytes and erect rosettes, with greater numbers of vegetatively repr oducing species. 3 Heavily grazed sites had higher proportions of ther ophytes and versatile/flat rosettes and species with mobile seeds than sites with light grazing. Moderately grazed sites had increased propo rtions of versatile and erect rosettes and more species with adhesive seeds (mainly grasses). Lightly grazed sites had a greater diversity i n terms of evenness of all life-forms, dispersal morphologies and repr oductive modes. 4 Water enriched sites had fewer geophytes and phanero phytes and some chamaephytes, erect rosettes and proto-hemicryptophyte s. This loss was only partially compensated by a gain in versatile and partial rosettes but richness decreased overall. Water enrichment was unrelated to dispersal unit morphology or the potential for vegetativ e reproduction. 5 Of the three attributes examined, life-form was most useful in characterizing community response to different disturbance types. Traits related to regeneration (seed morphology and capacity of vegetative reproduction) were relevant to soil disturbance only. Our results support the use of such classifications for monitoring the eff ects of disturbance.