Journal of the Royal Society of Western Australia, 86(1), March 2003 
- 0 . 1 - 
- 0 . 2 - 
-0.3- 
Dampland 
GW _ 
Af § 
EpE I- 
EpMl- 
Ha I- 
Spring 
Embank. 
Midslope 
GW 
Pe 
i 
T I-1-1- 
"35 -30 -25 -20 -15 -10 
Xylem 5 2 H (%o) 
Figure 7. Predawn water potential (WP d ) in relation to xylem 
water 5 2 H on a seasonal basis. Shaded bars represent the range 
of groundwater 8 2 H for the three sites, except for spring data 
where dampland and embankment/midslope sites are 
represented by different bars (unshaded and shaded 
respectively). Species abbreviations are provided in Fig 1. Values 
are mean ± se for 3 different individuals per species. 
that of A. fascicidaris (-27.54 ± 2.16 %). A. fascicularis was 
utilising the same soil moisture source (0.8-1.0 m) as P. 
ellipticum in winter. 
At the embankment site, H. angustifolium was 
accessing soil moisture in the top 1 m of the soil profile 
in spring and summer (Fig 6) and probably in winter. 
Winter soil water 5 2 H data coincided with xylem water 
5 2 H at 0.6, 2.2 and 3.0 m; however, H. angustifolium was 
unlikely to be accessing soil moisture at depths > 1 m, 
due to its shallow-rooted nature (Dodd el al. 1984). £. 
pauciflora was accessing moisture from both shallow (< 1 
m) and deeper (< 2 m) soil depths in spring and summer 
at the embankment site, and possibly groundwater. 
Xylem water 8 2 H data suggest that £. pauciflora was not 
accessing groundwater when winter data was collected. 
It appears that at the midslope site £. pauciflora 
accessed groundwater during spring and summer, but 
not in winter. £. pauciflora was also accessing moisture 
from depths up to 2 m in the soil profile in summer and 
between 1-2 m in winter. 
Discussion 
Summer drought is one of the major environmental 
factors affecting species survival in south-western 
Australia (Hnatiuk & Hopkins 1980), with severe or 
prolonged drought periods considered an extension of 
the normal summer drought scenario (Hobbs & Mooney 
1995). Past studies on plant water relations of Perth's 
sandplain species have shown that shallow rooted (< 1 m 
rooting depth) species tend to display more negative 
water potentials and have substantially lower 
transpiration rates than co-occurring deeper rooted shrub 
species during mid-late summer (Grieve 1956; Dodd et al 
1984; Dodd & Bell 1993). On the other hand, the present 
study shows that shallow-rooted species were the least 
water stressed, compared with the deeper rooted £. 
pauciflora species. The difference between these results is 
because none of the earlier studies included wetland 
sites, or sites where summer groundwater depth was < 3 
m, whereas all the shallow-rooted species in this study 
occurred within or fringing a wetland. 
The response of myrtaceous shrub species to the 
extended summer drought of 2000/2001 differed 
according to their location within the sandplain 
landscape, as shown by an overall decline in WP pd with 
increased groundwater depth. A decline in water 
potential and stomatal conductance was expected during 
mid-late summer, as species respond to increasing soil 
moisture deficits. This was the case for H. angustifolium 
and £. pauciflora occupying the embankment and 
midslope sites, and resulted in substantial decreases in 
soil-leaf hydraulic conductivity, but not for the dampland 
species. For H. angustifolium and £. pauciflora the seasonal 
trends between stomatal conductance and WP pd or 
hydraulic conductivity suggests a close association 
between the root/soil system and stomatal response, but 
does not explain H. angustifolium's relatively large 
hydraulic conductance in winter. Differences in stomatal 
responses to soil and atmospheric water deficits between 
species may be explained by differences in soil-leaf 
hydraulic conductance and a species threshold water 
potential (Bond & Kavanagh 1999), the latter being a 
minimum water potential beyond which the chance of 
38 
