Tuesday, 13 September 2016

Why data are important

There was an item on the BBC's 'Countryfile' on Sunday 11 September in which the findings of the latest 'State of Nature' report were discussed. I've yet to see the report, but in the past the HRS data have been used extensively in analyses of the status of Britain's wildlife.

The State of Nature Report presents disturbing but not unexpected findings – a substantial decline in wildlife across the British Isles, with a number of localised success stories for single species. On the whole, the message is very alarming (or should be). Sadly, the main agricultural contribution to the item laid the blame squarely on a rise in predator numbers but their assertions can be rapidly dismissed when one looks at the data for hoverflies:

Between 40 & 60% of hoverfly species are declining in abundance, whilst fewer than 20% are gaining in numbers/range. The declines are primarily amongst species with complicated habitat requirements, whilst gains are largely amongst species capable of dwelling in the urban environment and species introduced as a result of poor biosecurity.

If, as is asserted , the problem is predators, then by all rights with bird numbers declining hoverfly numbers should be rising! The opposite obtains: bird numbers are declining and so too are the numbers of hoverfly species that occupy the specialist niches. The conclusion is pretty obvious – something else is affecting the abundance of wildlife!

Making these connections is reliant upon sound data: the bigger the dataset is, the more robust the resulting analyses will be. Regular recording from a 'patch' or garden is a very important way of generating this robust data, but the casual contributions of all participants in the Facebook page help too.

The HRS was used for a decade or more as the test data for developing the models that inform the State of Nature report. How can we know this? Stuart Ball who is the HRS data guru was JNCC's Chief Analyst (until March this year) and did a considerable amount of the work developing current models. He always turned to the HRS data for validating models because he knew the data, understood the biology of the animals and was therefore able to sense-test the outputs.

So, when we ask for data, we are asking participants to help develop one of the datasets that might just help to change the most entrenched views about the health of Britain's wildlife.

No comments:

Post a Comment