Journal of the Royal Society of Western Australia, 86(1), March 2003 
analysis using a gas phase mass spectrometer (VG Isogas 
SIRA 10), and were analysed using the procedure 
outlined by Coleman et al. (1982). Hydrogen/deuterium 
ratios (5 2 H) were expressed according to standard 
notation in parts per thousand (%o) relative to the V- 
SMOW (Vienna Standard Mean Ocean Water) standard 
as 5 2 H =((R Mmplt /R 5la „ djrd )-D x 1000 where R Hmp and 
^standard are rat i° s of the heavy to light isotope of the 
sample and standard water respectively. 
8 2 H was only measured once for all water samples, 
except for free water samples where duplicate samples 
were analysed. For soil water 5 2 H, soil samples collected 
at 0.4 m intervals were analysed. 
Data analysis 
Analyses of variance were calculated using the 
SuperANOVA (Abacus Concepts, Berkeley, CA) 
computer package. Homogeneity of variances was 
checked by residual plots and data were transformed as 
necessary by log or square roots, with data presented as 
untransformed means. Tukey's HSD test was used for 
comparisons between means at the 0.05 significance level. 
Figure 2. Climate and groundwater data for Perth and Lexia wetlands. A: Rainfall and temperature data for Perth meteorological 
station from Jan 2000 - July 2001. Arrows represent sampling times for plant water relations; B: groundwater levels in situ at the three 
study sites from Oct 2000 to Oct 2001; C: monthly variation in groundwater levels (Jan 2000 - Sept 2001) from two permanent 
monitoring bores near the Lexia wetlands (see Fig 1). Bores chosen were those that had similar October 2000 (spring) and March 2001 
(summer) groundwater depths to the dampland and midslope study sites. 
34 
