Stream Mesocosm Experiment Data on Benthic Communities, 1996-1997

Eleven standardized biological variables from two stream mesocosm experiments conducted in the summer-autumn of 1996 and 1997. The data was collected to train artificial neural networks for diagnosing nutrient and fine sediment impacts in Tasmanian rivers. Results are published in Magierowski et al. (2015) in PLoS ONE.

Use Cases

• Training neural network models for environmental diagnosis based on standardized biological response variables.
• Analyzing the correlation between specific taxa abundances and experimental treatment levels of nutrients and sediments.
• Comparing the response of algal biomass (chlorophyll a) and macroinvertebrate community metrics to controlled stressors.
• Benchmarking ecological models using data from replicated mesocosm experiments.

Strengths

• Data is derived from two replicated mesocosm experiments conducted in 1996 and 1997.
• Includes 11 specific biological variables, such as chlorophyll a and abundances of nine common Tasmanian invertebrate taxa.
• Variables were standardized against control samples, potentially reducing experimental noise.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• The authors note some experimental treatment effects were ignored to simplify neural network training, limiting direct ecological conclusions.

Provenance

Source

Terrestrial Ecosystem Research Network's Data Discovery via data_gov_au.

Collection Method

Results from controlled stream mesocosm experiments.

Time Range

Summer-autumn of 1996 and 1997.

Freshness

Last updated 2026-06-10 20:03:50.963390; freshness should be verified.

Geography

Experiments likely conducted in Tasmania, Australia, based on the described taxa.