Grassland responses to multiple disturbances on the New England Tablelands in NSW, Australia

Disturbances have played a significant role in shaping vegetation patterns following European settlement and agricultural development in Australia, particularly over the last 100 years. However, little is known about the temporal dynamics of plant communities in relation to disturbances and their interactions. In this study we examined the response and recovery of temperate grassland communities to disturbance, using a multi-factorial experiment involving grazing exclusion (absent and present), fire (absent and present), soil cultivation (none, 5 and 20 cm) and soil amendment (none, fertiliser and fertiliser plus clover seeds) on the New England Tablelands in NSW, Australia. Temporal patterns of plant communities were analysed using detrended correspondence analysis for eight surveys over 24 months. Vegetation patterns at each survey were analysed using principal components analysis. The effects of treatments on malleability (Bray-Curtis dissimilarity) of plant communities were analysed using a linear mixed model, and the relationships between malleability and plant species groups were depicted using a generalised additive model and further analysed using a linear model. Perennial native grasses and a non-native forb (Hypochaeris radicata) initially dominated the vegetation, but after the disturbances H. radicata and other non-native species dominated. Compositional changes from the initial states were greatest in the first spring (7 months after treatment application), and then the vegetation tended to recover towards its original state. Soil cultivation resulted in the greatest deviation in community composition, followed by soil amendment, fire, with grazing exclusion the least. The recovery process and rate of recovery varied with treatment reflecting the dominance of soil cultivation and its interactions with other forms of disturbances. Soil amendment and grazing exclusion tended to reduce the effects of soil cultivation. Malleability was negatively related to perennial native grasses; positively to other non-native species, annual native grasses and perennial native sedges/rushes; and negatively to H. radicata when its cover was below 18%, but positively when above 18%. The degree of malleability reflected the high resilience of the vegetation to disturbance, and was mainly due to the recovery of perennial native grasses and H. radicata. This resilience demonstrated that the small-scale disturbances did not cause vegetation to cross an ecological threshold and that the present vegetation is resilient to common disturbances occurring at small scales. The results also suggest (1) that the present vegetation has developed mechanisms to adapt to these disturbances, (2) the importance of disturbance scale or (3) that the ecological threshold had already been crossed and the present vegetation is in a degraded state compared with its original state before the end of the 19th century.