Page 93 - Freshwater-Biology-and-Ecology-Handbook

Freshwater Biological Association Training

RIVPACS sampling is based on catch-per-unit effort and is appropriate for semi-quantitative analysis.
Although suitable for assessing environmental quality and community composition, it is not suitable for
quantitative assessment of population sizes.

• Because this approach requires samplers to identify all The net must be made of soft-woven multifilament
the potential invertebrate habitats at any site, they must polyester. This is much easier to repair than hard
have a good understanding of invertebrate ecology. monofilament nylon nets, which are also difficult to empty.
They also need to understand the practicalities of Because the size of the mesh is critical, the condition of
laboratory analysis and the RIVPACS model for which the nets is important and damaged nets must not be used.
the samples are collected. The level of knowledge Always take spare nets with you when collecting samples.
required for RIVPACS sample collection should not be
underestimated because all other analysis depends on • RIVPACS sampling is quicker and therefore cheaper than
the quality of sample collection and its adherence to the many other methods. One of its main advantages is that its
strict protocols.
limitations are well known and documented. There is good
quantitative information about its precision, with estimates
• For RIVPACS, the sample is standardised by the time of sampling variation based on extensive replicate
spent actively sampling. In shallow streams this is 3 sampling tests. There are also robust quality assurance
minutes, but in deep rivers it is 3 minutes in the main procedures for laboratory analysis and estimates of its
channel and 1 minute sweeping the margins. An additional errors. This means that, although apparently imprecise
minute is always spent searching for and collecting sampling methods are used, statistically significant
individual specimens of attached or surface-dwelling differences between samples from those caused by
taxa. The 3 minutes for the main sample are allocated to random error can be distinguished.
different habitats in proportion to their cover. Where it is
not possible to see different habitats, in deep or turbid
water, samples are collected along transects across the • Smaller, 1-minute samples are fine for detecting gross
river, covering shallower areas nearer the banks and the impacts. However, the benefits of the time saved in
deeper mid channel. laboratory analysis must be balanced against their
incompatibility with standard samples and their quantified
estimates of precision, incompatibility of measures
• RIVPACS sampling is based on a 1 mm mesh net. The net of taxonomic richness or other measures of diversity
is a critical component for standardisation that is common based on them, the inability to calculate indices that
to all the sampling equipment. It collects invertebrates take abundances into account (Chapter 3), and their
of a sufficient size for effective laboratory sorting and incompatibility with RIVPACS and therefore the official
identification and is less prone to blockage than finer WFD status classification.
meshes, which helps its efficiency in the field. Most other
European national methods use a 500 μm mesh.

Freshwater Biology and Ecology Handbook | 93
–