Five years monitoring of the Lake Muir-Unicup wetland system, south¬ 
western Australia 
N Gibson 1 , G J Keighery 1 & J A K Lane 2 
1 Science Division, Department of Conservation & Land Management, Wildlife Research Centre, 
PO Box 51 Wanneroo, Western Australia 6065 
2 Science Division, Department of Conservation & Land Management, 14 Queen Street Busselton, Western Australia 6280 
El neilg@calm.wa.gov.au 
(Manuscript received July 2003; accepted June 2004) 
Abstract 
The vegetation of 27 wetlands was monitored over a five year period (1997 to 2002) in the Lake 
Muir-Unicup system east of Manjimup, where salinization has been reported over the last decade. 
Initial results suggest that only three wetlands have deteriorated over this period. However given 
the lag times known to occur in groundwater movements and the appearance of new saline seeps 
in the catchment, it is unlikely that the ground water system is yet in equilibrium. There is an 
urgent need for ground water management programs to be developed to protect biodiversity 
values. Detailed hydrological mapping of the catchment is currently underway. 
Keywords: Monitoring, wetlands, salinity, Muir-Unicup catchment 
Introduction 
Dryland salinity is one of the most serious concerns 
for conservation management in the South West 
Botanical Province (George el al 1995) which has been 
recognized as one of the world's 25 biodiversity hotspots 
under threat (Myres et al. 2000). The salinity problem has 
arisen as a result of the large scale clearance of perennial 
native vegetation for annual cereal crop and pasture 
production with the subsequent rise in water tables and 
mobilization of salt that was previously confined deep in 
the soil profiles. Because of their position low in the 
landscape, wetlands are generally considered one of the 
first habitats to be impacted by rising saline ground 
water. The problems in wetlands are often exacerbated 
by drainage works. 
In 1996 the Western Australian Government released 
a Salinity Action Plan to address a number of issues 
arising from salinity (Government of Western Australia 
1996). One of the strategies was to designate recovery 
catchments which would be allocated resources to protect 
biodiversity values. The Lake Muir-Unicup system of 
wetlands was one of the first identified, on the basis of 
outstanding fauna and flora values (Australian Nature 
Conservation Agency 1996; Gibson & Keighery 2000). 
Based on these same data the Muir-Byenup system, south 
of the Muir Highway, was subsequently nominated 
(Government of Western Australia 2000) and listed 
(Wetlands International 2002) as a Ramsar wetland of 
international importance. 
In 1997 a series of 41 permanent plots was established 
in 27 wetlands to monitor vegetation change through 
time and to provide a mechanism to assess the 
effectiveness of conservation management activities 
carried out in the Lake Muir-Unicup catchment. This 
© Royal Society of Western Australia 2004 
catchment is situated in the 600 - 800 mm rainfall zone 
and wide-scale plantings of eucalypt plantations for fiber 
production was believed to provide an economically 
feasible way to control rising water tables. This paper 
reports on the changes in wetland vegetation for the five 
years commencing in 1997. 
Methods 
Vegetation cover was recorded in 41 quadrats located 
in 27 wetlands in October 1997. Most were resurveyed in 
October 2002, however five quadrats (from three 
wetlands) could not be accessed at that time and were 
instead resurveyed in March 2003 (Fig. 1). Each quadrat 
was 5 m x 10 m with comers marked by steel star bars or 
droppers. Lists were compiled of all vascular species 
recorded in each quadrat and a percentage cover class 
was visually estimated for each species (<2, 2-10, 10-30, 
30-70, >70%). In basin wetlands plots were laid out with 
their long axis parallel to the shoreline. Between one and 
three plots were established in different vegetation zones 
at each wetland. Where multiple quadrats were used, 
quadrats were placed in parallel along the littoral 
gradient. In all, five broad wetland types were sampled 
(Baumea articulata basin wetlands; flats and riparian zones 
dominated by Baumea juncea; swamps and flats with 
understorey dominated by Lepidosperma longitudinale; 
other wet flats; and one seasonally inundated claypan). 
The position of each quadrat was fixed with a GPS unit. 
To test whether a significant change in vegetation 
cover of the quadrats had occurred over the five years a 
similarity matrix using midpoint cover values was 
constructed using Bray-Curtis similarity coefficients with 
data from all quadrats at both sampling times. This 
matrix was analyzed using the non parametric ANOSIM 
routine (Primer v5.2.8 Clark & Gorley 2001) which tests 
for differences between rank similarities of quadrats 
between years using a permutation test. 
29 
