Three river conceptual models make differing predictions about the major source of primary production in lowland rivers, acknowledging the importance of primary productivity in the ecology and management of lowland rivers. Patterns of primary production in lowland rivers are still an area of considerable uncertainty. The objective of this study was to examine the major sources and transformations of organic matter in an Australian lowland river and compare them to the predictions of existing models. The broad approach adopted was to quantify the contribution from the major ecosystem components and compare these with estimates of system metabolism determined using open water measures of diel oxygen change. Three 4-km river reaches were selected to represent the extent of variation found along the free-flowing lowland sections of the Murray River, one of Australia's largest and most regulated rivers. Annual open water gross primary production (GPP) estimates for the Murray R. during this study ranged from 221 to 376 gC m À2 y À1 and were similar to other large rivers. Examination of the net contribution of organic matter to the channel indicates that primary productivity in the Murray R. is derived from a combination of phytoplankton, riparian vegetation and macrophytes, but that the major source varies both spatially and temporally. The present study confirms that the River Continuum Concept (RCC), the Flood Pulse Concept (FPC) and Riverine Productivity Model (RPM) all have some application to Australian lowland rivers, but that synthesis of the models will be difficult until we can incorporate the extent, causes and consequences of primary production variability. This study also highlights the importance of the microbial loop and macrophytes in the ecology of the Murray R.

Bacterial production is important in aquatic carbon cycles because it represents a key component whereby dissolved and particulate carbon can be recycled back into food webs. Despite its acknowledged importance, few studies have examined bacterial production in lowland rivers. Since studies have suggested bacterial production is closely related to some carbon pools, we anticipated this to be the case in the Murray River, but that the timing and type of carbon inputs in the Murray River may lead to bacterial dynamics that differ from studies from other sites. Bacterial abundance and production were measured at three contrasting sites of the lowland Murray River, southeastern Australia, over an 18-month period. Bacterial abundance varied across the three sites on the Murray River and was correlated with chlorophyll a concentrations but not with temperature, nutrients, particulate organic carbon and dissolved organic carbon concentrations. Bacterial production also varied across the sites. Lowest production was at the site most immediately downstream of a large reservoir, with production generally ranging from 0.88 to 8.00 µg C L −1 h −1 . Bacterial production in a reach within a large forest ranged from 4.00 to 17.38 µg C L −1 h −1 . Production at the reach furthest downstream ranged from 1.04 to 23.50 µg C L −1 h −1 . Bacterial production in the Murray River was generally greater than in the European River Spree, reaches of the Meuse and Rhine without immediate impacts from major urban centres and the Amazon River, but was similar to the concentration measured in the Mississippi and Hudson Rivers. Bacterial production was closely correlated with chlorophyll a concentration and total phosphorus, but not with temperature, dissolved organic carbon, particulate organic carbon or inorganic nitrogen. Despite the differences in production and respiration measured at different sites across the Murray River, bacterial growth efficiency was very similar at the three sites. Bacterial populations in the Murray River appear to be influenced by reach-specific conditions rather than broad-scale drivers such as temperature, carbon and nutrient concentrations.

Benthic respiration is an important measure of decomposition processes occurring in streams, but our understanding of benthic respiration in lowland rivers is not well developed, particularly the factors that affect benthic respiration. In our study we measured benthic respiration at three sites in three contrasting lowland rivers in southeastern Australia. On most sampling occasions, rates of oxygen consumption in benthic chambers were linear. However, oxygen consumption rates fitted exponential decay curves during periods of highest microbial activity. Benthic respiration was closely correlated with water temperature, but not with sediment carbon content, sediment particle size, water column nutrients or water column dissolved organic carbon concentrations. Average carbon turnover periods were between 1.7 and 6 years for the three rivers, but were as low as 0.1 year immediately following an event that gave rise to mobilization of in-stream dissolved organic carbon, sufficient to produce coloured water. The latter occurred in the Ovens River as a consequence of a rain event during a period of base-flow. Flow regime as such did not have a major impact on benthic community respiration. Induced changes in respiration, by altering flows, would only occur by altering the quality and timing of carbon inputs, since temperature and carbon quality, rather than quantity, appear more important in determining lowland river benthic respiration.