MEDIDAS DE VISCOSIDAD

Figure 8-11 Source-sink model

8.3 Conclusions

8.3.1 Study weaknesses

Field work in the study region was conducted during an extended period of drought, negatively influencing the condition and hence percentage cover of perennial shrub species.

The study was carried out in a remote area of arid Australia where vehicular access was very limited. These attributes, combined with the spatial extent of the study area, (approximately 120 kilometres north to south), restricted the total number of sites that could be assessed in the time available. Consequently some vegetation communities and species were under sampled. This restricted the number of species available for generalized linear mixed modelling. The predictive power of the models subsequently produced for these species would have been greatly improved by additional sampling. In hindsight, the study was probably too ambitious in aiming to cover the full range of floodplain communities in the limited time available to sample both soils and vegetation.

Inevitably, given the available time and resources, confounding factors, such as grazing, road, track and seismic survey infrastructure, all of which could exert additional influence upon plant distribution and abundance, were not included in the study. To a certain extent the focus

on very long lived and predominantly unpalatable perennial species may have alleviated this weakness. However, the cover estimates for a number of species considered palatable and vulnerable to grazing, such as Chenopodium auricomum, Atriplex nummularia, and Sporobolus mitchellii are likely to have been affected as a result of past grazing pressure and need to be viewed in this light.

While the scope of inference of the findings reported here must generally be restricted to the study area, the range of most of the perennial floodplain species considered extends well beyond it.

8.3.2 Future research

8.3.2.1 Vegetation

Arid ecosystems have long memories. Rare flooding or extensive rainfall events with long return periods are highly significant in structuring arid ecosystems. The ecology of long lived perennial plant species depends upon these major events and hence any monitoring of these species must involve long term studies.

8.3.2.1.1 Autecology and long-term monitoring

While there is remarkably little evidence of detailed autecological studies of Australian floodplain perennial species, particularly those of arid zone environments, there have been two significant studies of Eucalyptus coolabah in the far north-east of South Australia. The first examined the demography of Eucalyptus coolabah in the lower reaches of the Cooper near the Birdsville track (Roberts, 1993).The second examined the water sources accessed by Eucalyptus coolabah at sites on the floodplain of the Diamantina and Neales Rivers (Costelloe et al., 2008). Both authors refer to the dearth of knowledge regarding the ecology of this species. The potential of Eucalyptus coolabah as a „keystone ecosystem engineer‟ species (Corenblit et al., 2007) should be investigated. The study area is currently experiencing a rare La Niña driven, extensive, clustered flooding event (Puckridge et al., 2000). This situation provides a prime opportunity for the establishment of a long-term autecological study of the species. The floods will drive the dispersal and germination of Eucalyptus coolabah across the extent of the floodplain. This cohort should be followed through time across a recognized mesic to xeric gradient and range of soils, assessing attributes such as growth rates, establishment and persistence.

8.3.2.1.2 Investigation of the hypothetical recharge model

The proposed hypothetical conceptual model of soil and ground water recharge and its role in the persistence of Eucalyptus coolabah in more xeric sections of the floodplain should be investigated (Figure 8-9). The research approach of Costelloe et al. (2008) could be extended to Cooper Creek to test the proposed hypothesis. Radiometric data have been usefully employed in

the past to detect buried palaeodrainage lines in Australia (English, 1998). The availability of data for the Cooper should be investigated and the potential for its use in mapping sections of the Cooper floodplain determined. Additionally SANTOS Ltd has extensive bore log data associated with numerous drill sites across the floodplain. Interpretation of this data in combination with radiometric survey could provide useful insights into little known surface water-ground water interactions of arid zone floodplain systems (Costelloe et al., 2007, Jolly et al., 2008), particularly in the context of persistence of Eucalyptus coolabah across the landscape.

8.3.2.1.3 Investigation of spatial distribution

The data collected along each line intercept transect during the study was detailed. Quantitative measurements recorded width, and height for all intercepted plant individuals from all strata. These data could provide potentially valuable insights into possible shifts in dimensions and spatial distribution of species across soil and moisture gradients within the floodplain, which may reflect inter or intra-species interactions such as competition or facilitation.

8.3.2.2 Soils

During this study, in addition to the single 50 cm soil profile sample collected along each transect (138 in total), soil samples were collected for three 10 cm profiles (414 profiles in total). These samples were intended for investigation of biogeochemical characteristics and potential seed germination trials. Time and financial restrictions prevented the intended examination of these samples.

8.3.2.2.1 Mid infrared (MIR) calibration

There is little information on the MIR spectral characteristics of dryland river floodplain soils (Neil McKenzie pers comm.). The 414 profiles would provide a valuable calibration data set against more orthodox laboratory analyses.

8.3.2.2.2 Nutrient dynamics

There is a dearth of information on biogeochemical cycles and nutrient dynamics of dryland river floodplain soils (Robertson et al., 1999, Baldwin and Mitchell, 2000, Scholz et al., 2002, Ogden et al., 2007). The importance of the concentration of biologically mobile C, N and P in the upper soil profile and the role of microbial activity in nutrient fluxes, particularly the top 5-10 cm in semi-arid and arid soils is widely acknowledged (Cowling, 1969, Charley and West, 1975, Schlesinger and Pilmanis, 1998, Whitford, 2002). It is generally considered that fine-textured soils tend to have higher labile C and N pools than coarse-textured soils, and often show a much greater flush of N mineralisation due to microbial activity following inundation or rainfall (Austin et al., 2004, McIntyre et al., 2009). Microbial respiration rates can thus provide

a useful proxy for the fertility of soils (David Tongway pers. comm.). Inundation trials under two temperature regimes reflecting summer and winter inundation events, monitoring resulting levels of microbial respiration, could provide interesting insights into the productivity of the range of floodplain soils sampled.

8.3.2.3 Palaeoecological research

8.3.2.3.1 Palynology and sedimentology

Throughout the study area, particularly along channels and in the vicinity of semi- permanent waterholes and lakes there are frequent and often dense scatters of aboriginal artefacts. Occasionally within the scatters are two different forms of large stone artefacts (Plate 7-3a, b).