The water quality impact of a herd of 246 dairy cows crossing a stream ford was documented. Two cow crossings produced plumes of turbid water associated with very high concentrations of faecal indicator bacteria ( Escherichia coli ) and high suspended solids (SS) and total nitrogen (TN). On the first crossing, towards the milking shed, the cows were tightly‐bunched and produced a sharp spike of contamination ( E. coli peaking at 50 000 cfu/100 ml). After milking, the cows wandered back across the stream as individuals or small groups, and contaminants were less elevated, albeit for a longer period. Light attenuation, measured continuously by beam transmissometer, correlated closely with E. coli , SS, and TN, permitting the total yield of these contaminants to be estimated. Contaminant yields for the two crossings were very similar, suggesting that time taken and whether or not cows are herded may not greatly influence water quality impact. The cows defecated c. 50 times more per metre of stream crossing than elsewhere on the raceway. This study has shown that cattle accessing stream channels can cause appreciable direct water contamination, suggesting that excluding cattle from streams will have major water quality benefits.

Streams and rivers provide important services to humans, and therefore, their ecological integrity should be a societal goal. Although ecological integrity encompasses structural and functional integrity, stream bioassessment rarely considers ecosystem functioning. Organic matter decomposition and ecosystem metabolism are prime candidate indicators of stream functional integrity, and here we review each of these functions, the methods used for their determination, and their strengths and limitations for bioassessment. We also provide a systematic review of studies that have addressed organic matter decomposition (88 studies) and ecosystem metabolism (50 studies) for stream bioassessment since the year 2000. Most studies were conducted in temperate regions. Bioassessment based on organic matter decomposition mostly used leaf litter in coarse-mesh bags, but fine-mesh bags were also common, and cotton strips and wood were frequent in New Zealand. Ecosystem metabolism was most often based on the open-channel method and used a single-station approach. Organic matter decomposition and ecosystem metabolism performed well at detecting environmental change (≈75% studies), with performances varying between 50 and 100% depending on the type of environmental change; both functions were sensitive to restoration practices in 100% of the studies examined. Finally, we provide examples where functional tools are used to complement the assessments of stream ecological integrity. With this review, we hope to facilitate the widespread incorporation of ecosystem processes into bioassessment programs with the broader aim of more effectively managing stream and river ecosystems.