Changes in soil moisture under various densities of mulga (Acacia aneura F. Muell.) were followed from December 1971 to October 1973. Stemflow was instrumental in storing water at depth in the soil, being particularly noticeable with medium falls of rain (~ 75 mm); heavy (~ 160 mm) falls tended to mask the effect. Throughfall increased linearly with aggregate rainfall, and percentage throughfall decreased with increasing tree density. Of incoming precipitation, 94% was partitioned as throughfall under a tree density of 40 trees/ha-1 compared with 86% at a density of 4000 trees/ha. No distinct soil moisture patterns due to throughfall were found. Infiltration rates of water into the soil decreased with increasing distance from trees, being 46, 22 and 17 mm/h-1 after 10 min at distances 0.25, 0.5 and 2 m from a tree bole respectively, thus demonstrating that stemflow waters are absorbed at close proximity to the tree bole.The results are discussed in terms of the ecological importance of stem flow and soil moisture patterning to the survival and growth of mulga and associated ground flora.

In this study of rainlfall characteristics, throughfall was high!y positively cor- related with precipitation. Stemflow and interception were positively correlated with precipitation. Rainfall intensity did not significantly affect the stemflow or interception components, though throughfall was shown to increase with increasing intensity for rainfall events up to 6 mm aggregate. A positive relation was shown between tree size and interception, while negative relationships were found between tree size, and throughfall and stemflow for various classes of precipitation events. Volume of stem- flow waters was positively related to tree size. The results are discussed in terms of soil infiltration characteristics, and the woodland community.

Run-off, bedload and sediment concentration data were collected over a five-year period from unbounded catchments in grazed and exclosed pastures in woodlands. Cover varied from 4% during drought conditions to almost 100% in exclosed areas after above-average rainfall. High bedload soil loss, sediment concentration and run-off percentages were associated with low cover (<30%). Run-off as a percentage of rainfall increased linearly with rainfall intensity; decreased linearly with cover; decreased slightly as soil moisture status declined; and reached a maximum at intermediate rainfall events. Interactions between these factors were observed. Run-off was up to 30% of rainfall in moderate rainfall events (30-40 mm) where maximum rainfall intensity over any 15 minute period (I15) exceeded 70 mm/h. When soil moisture status was high, mean run-off exceeded 30% for 40-80 mm rainfall events. For all rainfall event sizes, run-off exceeded 20% where I15 exceeded 60 mm/h. Cover had very little effect on run-off when rainfall intensity was low (I15<20 mm/h), soil water deficit was low (<10 mm) or when rainfall events were >75 mm or <10 mm. Bedload plus suspended sediment loads ranged from negligible to 1000 kg/ha/a, depending principally on cover. Soil movement from areas with >40-50% cover was very low. Pastures dominated by Bothriochloa pertusa (a stoloniferous, naturalised grass) had lower run-off and lower rates of soil movement than pastures dominated by Heteropogon contortus (a native tussocky perennial grass) when compared at the same level of cover. Differences between grazed and exclosed areas could be attributed solely to differences in cover.

Litter fall on two catchments, one with an almost continuous canopy of Eucalyptus spp., the other with a discontinuous eucalypt canopy was measured for 29 4-week periods. Annual litter accretion(2.5-3.7 tha~') was low by world standards, but similar to other Australian dry sclerophyll forests. Between 33 and 54% of the total litter fall occurred during summer, and only 6-13% in winter. Leaf was the major component of litter fall and usually contributed more than 50% of the total. Variability between the 4-weekly litter falls was high, but coefficients of variation for the total annual fall were low Half-life of total litter fall in the continuous forest (2.14 y) was double that in the discontinuous forest (0.98 y), and may have been due to differences in air temperature and surface soil moisture.The pattern of litter fall on each catchment was related to hydro-meteorological variables including air temperature, solar radiation, rainfall and soil moisture content.

Grazed and exclosed areas within north Queensland woodlands were studied over a six-year period which included a drought followed by an above average rainfall period. Despite drought conditions, woody plant density increased, with exclosed areas showing a greater increase than grazed areas. During the drought period, a pattern of herbaceous vegetation strips interspersed with bare areas developed. Standing dry matter and ground cover increased rapidly in response to above average rainfall following the drought. Also, the pattern of strips became less apparent during the more favourable growth conditions. The percentage contribution of native tussock grasses (principally Heteropogon contortus) to total pasture dry matter varied to a greater extent over the six-year period than the naturalised stoloniferous Bothriochloa pertusa.