It appears, at least to the outside world, that scientific processes are strict and universal. It is easy to assume that the marine biological data used to inform decisions about the marine environment, are fit for purpose. We would expect that ensuring their accuracy is a simple and routine process. But processes such as identification of marine life, are driven by research that is constantly changing. This makes processes varied and not universal. Yet, these scientific data need to be of a certain standard and consistency to inform decisions on impacts on marine habitats and to understand their conservation value.

Why are marine data always changing?

The scientific processes, such as those behind marine life identification, are driven by new research. As we explore, reassess and check, the facts are updated. There have been changes in expectations over time. Laboratories undertaking the work often follow different practices, impacting the results.

The resources used for identification are a good example of the constant change. Once upon a time, laboratories used a small number of all-inclusive books to identify British marine life. Over the years, many new species have been described, both from UK and neighbouring waters. Several well-known species have since proved to be ‘complexes’ (groups of similar related species that can be difficult to differentiate).

Every year, new finds from British waters are discovered for species previously unknown here. This could be either because they were overlooked or because they have recently arrived, either as non-natives (due to human activity) or through fluctuations in distribution due to habitat or climate change. The result is that biologists must be constantly alert to new literature and identification resources.

There have always been differences in data outputs between laboratories, and between years. These require standardisation through ‘data truncation’. This is where some species need to be excluded from data analyses or included with reduced identification accuracy in order to improve consistency between different data sources. The greater the need for truncation, the more information is lost through analysis of data.

There may be differences in quality or in policy. Policy rules might include decisions to record or exclude certain species (whether to count or record as present), which animals are considered to be identifiable to species level and when to note that a given animal is considered juvenile. There may be similar variation in records of litter and other material. Data quality can be reduced by identification mistakes or by species or individuals being missed during sample analysis.

Standardising marine biological data

One significant step towards providing standardised, quality assured data is the NE Atlantic Marine Biological Analytical Quality Control (NMBAQC) scheme. This provides a focus for quality assurance (QA) and data consistency. The Scheme provides guidance on standardising data and reducing the need for truncation. It also provides standard audits for the quality control (QC) of data. Agreed pass/fail criteria are applied after re-analysis of samples by a second laboratory and improvements are stipulated where necessary. The NMBAQC Scheme also brings together research and resources. It provides expert training through specialist workshops and identification exercises.

APEM’s role

APEM have administered components of the NMBAQC Scheme for five years. Many of our staff have been involved since it began in the 1990s. We are pleased to have been recently awarded the contract for the next four years . We organise workshops and exercises, carry out audits, and take a lead in the development of protocols and literature research. We have seen data consistency improve greatly over the years. Yet, we recognise much remains to be done, particularly the extension of the ethos of quality and consistency for those currently outside the direct influence of the scheme.

If you have any queries, please contact Tim Worsfold, technical specialist.
Alternatively you can email us here. Or call 0161 442 8938.