Livestock can adversely affect stream health through grazing (riparian vegetation) and trampling (soil condition) leading to the expectation that fencing the near riparian zone and providing off-river watering points will improve stream health. Vegetation recovery is a long-term process, and varies (Lunt et al 2007) with the degree of floristic degradation and site productivity. In 2008-2009, during the Millenium Drought, Western CMA established a long-term monitoring program to track the effects of fencing the riparian zone. This had 32 sites on 5 rivers (Barwon, Darling, Culgoa, Narran, Bogan), with locations determined by the distribution of landholders participating in the fencing incentives scheme. The design was unbalanced (9 Control, 3 Reference, 20 Treatment sites) and targeted just one type of riparian vegetation (coolibah woodlands). All sites were re-sampled after 3-4 years, in October 2012, following major floods. Data analysis used ANOSIM and SIMPER routines in PRIMER, and a mixed-model ANOVA with Fencing, Time and River as factors.
Principal findings are that: Fencing has had little effect, despite improved growing conditions, with no significant differences in species composition, number of recruits and only one of five variables describing vegetation structure. In contrast, species composition, number of recruits, tree cover and litter did change with Time (interpreted as major flooding). Differences in species composition between Rivers suggest an unacknowledged regional-scale variability in riparian coolibah vegetation. Overall, the flooding response appears to be transient, with the vegetation retaining much of its baseline characteristics.