Abstract
Floodplains are characterised by high biodiversity and productivity, providing habitat and refugia for a range of native species. These landscape features are also crucial for many aspects of human survival and quality of life. Inundation by flooding is the predominant factor that controls the observed ecological patterns and processes, and biological productivity on the floodplain. Semi-arid floodplains experience a highly variable flooding regime which has structured the vegetation communities.This thesis addresses some of the current gaps in our understanding of floodplain plants and their responses to water, on the semi-arid floodplain of the lower Lachlan River, Australia. It characterises spatial and temporal differences in seed bank characteristics relative to hydrological gradients, investigates the life-history responses of two floodplain shrub species (tangled lignum (Duma florulenta (Meisn.), and nitre goosefoot (Chenopodium nitrariaceum (F.Muell.)) to watering, and investigates gene-flow and population genetic structure in tangled lignum and river cooba (Acacia stenophylla (A. Cunn. ex Benth.)). This study is framed within the context of the changes in river floodplain connectivity caused by water resource development.
I identified and compared the changes in floodplain-river connectivity characteristics that have occurred as a consequence of water resource development on four floodplain habitat types which occur on the lower Lachlan River, using a hydrological modelling approach. The results show that the floodplain of the lower Lachlan River connects to the river half as often under current flow conditions compared to flow conditions without water resource development. While all habitat types have been impacted by water resource development, the impact reduces away from the main channel of the river, with temporary floodplain lakes experiencing the largest change in connection regime and intermittent black box swamps which occur on the margins of floodplains experiencing the least.
A seedling emergence experiment was conducted using soil collected from a total of 25 locations from three typical floodplain habitat types (temporary floodplain lakes, intermittent river red gum swamps, and intermittent black box swamps) on the lower Lachlan River. The potted soil was subjected to four experimental treatments (rainfall, waterlogged soil, submergence, and submergence followed by recession). Areas on the floodplain that connect to the river irregularly had a higher seed bank density and species richness compared to areas on the floodplain which connect more frequently. Sites collected in 2016 had significantly greater seed bank density and species richness than sites collected in 2017, following large-scale flooding. Comparatively lower numbers of terrestrial species were present at sites collected in 2017 than sites collected in 2016 compared to aquatic and semi-aquatic species.
I examined the effects of hydrological regime on germination of tangled lignum and nitre goosefoot, and the buoyancy of the seeds of the two species, and hence the ability of the seeds to disperse via water. Seeds of tangled lignum germinated best on soaked soil and on soil inundated for 20 days. Seeds of nitre goosefoot germinated best on soil inundated for five days and on soaked soil. The majority of tangled lignum seeds floated for at least seven days. The majority of nitre goosefoot seeds sank within seven days. Tangled lignum requires about 20 days of flooding, and wet soils following flood recession for optimal germination. Nitre goosefoot requires a few days of flooding, and wet soils following flood recession for optimal germination.
I investigated the effects of simulated rainfall and depth and duration of experimental flooding on mortality, leaf production, biomass and seed production of nitre goosefoot. The plants were subjected to 14 hydrological treatments: dry (no water applied), rainfall (simulating rainfall conditions at Hillston, New South Wales (NSW)) and 12 combinations of three water depths (10 cm, 50 cm, 75 cm) with four durations of inundation (5 days, 10 days, 20 days, 40 days). Nitre goosefoot plants survived flooding, providing plants were not totally submerged, leaf production increased during flooding and after drawdown, and leaf production, biomass and seeding were highest under shallow flooding for approximately one month. The results of this study allow the hydrological responses of nitre goosefoot to be considered in environmental watering programs.
I investigated and compared the population structure and dispersal patterns of tangled lignum and river cooba, across a large inland floodplain using a landscape genetics approach. Both species are widespread throughout flood prone areas of arid and semi-arid Australia but have differing distributions on rivers and floodplains. Leaves were collected from 144 tangled lignum plants across 10 sites and 84 river cooba plants across six sites, on the floodplain of the lower and mid Lachlan River Catchment, and lower Murrumbidgee River Catchment, NSW. DNA was extracted and genotyped using DArTseq platforms. Genetic diversity was compared with floodplain-river connection frequency, and genetic distance (FST) compared with river distance, geographic distance and floodplain-river connection frequency between sites. It was found that genetic connectivity increased with increasing floodplain-river connection frequency in tangled lignum but not in river cooba. In tangled lignum, sites that experience more frequent flooding had higher levels of genetic diversity and were more genetically homogenous. There was also an isolation by distance effect where increasing geographic distance correlated with increasing genetic differentiation in tangled lignum, but not in river cooba. These data demonstrate how genetic patterns can highlight influential mechanisms over species distribution and persistence on floodplains.
Semi-arid floodplains such as the floodplain of the lower Lachlan River are temporally and spatially dynamic and complex. The organisation of vegetation on the floodplain is strongly related to the hydrological regime and the requirements and tolerances of each species. The dependency of floodplain vegetation on flooding means hydrological changes are likely to affect floodplain vegetation by reducing opportunities for life history processes such as growth, dispersal, seed production, and seed germination. The results suggest changes to the flow regime will influence the distribution and abundance of species, genetic diversity and genetic structure, and the distribution of populations and communities across floodplains. To maintain the diversity of vegetation and habitat types on the floodplain, environmental flows must consider the life-history strategies of the plants.
| Date of Award | 2019 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Fiona Dyer (Supervisor) & Dianne Gleeson (Supervisor) |