Groom: Water relations of Myrtaceae shrubs during drought 
xylem cavitation is greatly increased. Based on the data 
presented in this paper, it would be expected that the 
threshold water potential of £. pauciflora would be more 
negative than that of H. angustifolium. 
The relatively low (approximately -3.3 MPa) WP pd of 
£. pauciflora at the embankment and midslope sites 
suggests that this species did not have access to 
groundwater sources by late summer, and is in contrast 
to the interpretation obtained from comparing xylem 5 2 H 
with that of groundwater and soil water 5 2 H data. A lack 
of discrimination between groundwater 5 2 H and soil 
water 5 2 H, as occurred at both £. pauciflora sites in spring 
and the embankment site in summer, makes it difficult to 
ascertain the main sources of water used (see Burgess et 
al. 2000). In the present study, water potential data 
suggest that £. pauciflora had access to groundwater in 
October (spring), when groundwater levels were at their 
highest (2.40 and 3.34 m respectively for the embankment 
and midslope sites), but did not have access to 
groundwater in July (mid-winter) when groundwater 
levels were approximately 0.1 m higher than in March 
(late summer). It is therefore unlikely that £. pauciflora 
had direct access to groundwater in summer at either 
site, and this is supported by WP pd data. This suggests 
that the minimum accessible groundwater depth for £. 
pauciflora at the embankment and midslope sites is 
between 2.4-3.0 m and 3.4-4.2 m respectively. Based on 
water potential data, Dodd & Bell (1993) showed that £. 
pauciflora was unable to access groundwater during 
summer at a depth of 7.2 m, but accessed groundwater 
during the spring months when the water depth was 6.3 
m. 
If £. pauciflora is not accessing groundwater during the 
summer period at the embankment site, then it stands to 
reason that H. angustifolium with its shallower root 
system is also not accessing groundwater. So why is it 
that H. angustifolium had a less negative WP pd ? H. 
angustifolium is restricted to and often fringes winter-wet 
depressions (Marchant et al. 1987), and is reliant on 
shallow soil moisture or groundwater sources when 
available. Differences in WP pd may be a result of £. 
pauciflora being restricted to the top end of the 
embankment near where the piezometer was installed 
(Fig 1), whereas H. angustifolium was sampled from the 
lower edge of the embankment where groundwater 
levels would have been < 3 m in summer. The negative 
correlation of summer WP pd with xylem water 8 2 H 
between species (Fig 7) suggests that species reliant on 
shallow soil moisture sources (< 1 m soil depth) were 
less water stressed than those utilising deeper soil 
moisture reserves ( i.e . £. pauciflora). 
Of ten myrtaceous shrub species examined for their 
long-term (30 years) response to decreasing groundwater 
levels on Perth's Swan Coastal Plain, the species 'tolerant 
of excessive wetness' (see Havel 1968) displayed the 
greatest reduction in population size (Groom et al. 2000a). 
This included A. fascicularis, P. ellipticum and H. 
angustifolium. It was expected that as A. fascicularis and P. 
ellipticum are species restricted to winter-wet depressions 
(Marchant et al. 1987) such as damplands, they would be 
dependent on groundwater all year round, particularly 
during the dry summer period. 
This study suggests that during a prolonged summer 
drought, neither A. fascicularis or P. ellipticum were 
accessing groundwater but were relying on soil moisture 
reserves. This also suggests that shallow soil moisture 
reserves (gravimetric content of 10-20%) were sufficient 
to sustain their summer water-use requirements, as both 
species were able to maintain summer stomatal 
conductances similar or greater than values obtained 
during the previous spring. The stable hydrogen isotope 
data show that P. ellipticum may have had access to 
groundwater in July 2001 (early winter) as a result of a 
small rise in winter groundwater levels, implying that 
the minimum accessible groundwater depth for P. 
ellipticum occurs between 1.39 to 1.45 m at this dampland. 
Wetland water levels on Perth's coastal plain fluctuate 
seasonally, and some wetlands, including damplands, 
dry out by the end of summer. Damplands by definition 
are seasonally waterlogged basins (Semeniuk 1987); 
however, Perth's damplands do not always retain surface 
water, and was the case in 2000/2001 for the dampland 
studied in this paper. This may be due to below-average 
winter rainfall, reduced seasonal recharge and/or 
declining trends in regional groundwater level (Muir 
1983; Townley et al. 1993). The trend for decreasing 
average rainfall currently being experienced by Perth 
may be part of a longer cycle (Davidson 1995). It has 
been predicted that by 2030 Perth may experience up to 
10% less winter rainfall, and 5% more, or less, summer 
rainfall; by 2070, winter rainfall may decrease by 20% 
with summer rainfall increasing/decreasing by 10% 
(Anon 1996). Consecutive years of summer drought/poor 
winter rainfall can trigger extensive mortality of 
sandplain vegetation (ITnatiuk & Hopkins 1980; Groom 
et al. 2000b). If reduced winter rainfall translates into a 
lowering of summer groundwater levels and soil 
moisture content, then shallow-rooted species occurring 
in low-lying areas (e.g. A. fascicularis, P. ellipticum) will be 
particularly susceptible. 
Acknowledgments: This research was supported by a Postdoctoral 
Research Fellowship Scheme funded by the Ministry for Commerce and 
Trade, Western Australia. Gary Ogden, Sandra Zencich and Ray Froend 
are thanked for their assistance in the field. Lidia Bednarek supervised 
the water extraction procedure and performed the mass spectrometry. 
Tim Perkins assisted in the production of the study site map. Access to 
the land containing the Lexia wetlands was granted by the Water and 
Rivers Commission. Direct access to the study sites was provided by 
Rocla Quarry Products. 
References 
Anon 1996 Climate Change Scenarios for the Australian Region. 
Climatic Impact Group, CSIRO Division of Atmospheric 
Research, Aspendale, Victoria. 
Arnold JA & Sanders CS 1981 Wetlands of the Swan Coastal 
Plain. In: Groundwater Resources of the Swan Coastal Plain 
(ed BA Carbon). CSIRO, Perth, 126-137. 
Beard JS 1984 Biogeography of the Kwongan. In: Kwongan: Plant 
Life of the Sandplain (eds JS Pate & JS Beard). University of 
Western Australia Press, Perth, 1-26. 
Bond BJ & Kavanagh KL 1999 Stomatal behaviour of four woody 
species in relation to leaf-specific hydraulic conductance and 
threshold water potential. Tree Physiology 19:503-510. 
Burgess SSO, Adams MA, Turner NC & Ward B 2000 
Characterisation of hydrogen isotope profiles in an 
agroforestry system: implications for tracing water sources of 
trees. Agricultural Water Management 45:229-241. 
Coleman MC, Shepherd TJ, Durham JJ, Rouse JD & Moore GR 
39 
