The minimum number of randomly selected squares we need to survey has been guided by statistical analysis performed by the British Trust for Ornithology (BTO).  Minimum sample sizes were calculated to enable measurement of 25% change over 10 years nationally. This corresponds with IUCN guidelines for detecting shifts in conservation category to amber. Initial occupancies of one-km squares were estimated at 15% and 30%.  Power analyses have suggested that where a species is initially present in 30% of squares (i.e. its occupancy rate), 400 sites should be sufficient to detect such a change (with 80% statistical power) At a lower initial occupancy of 15% of squares, the sample size figure exceeds 1000 sites to detect such a change. Whether this is achievable depends on volunteer densities in different areas across the UK.

Stratification has been performed according to Vice Counties, spreading the minimum of 400 sites across Britain so that each Vice County has the requisite number of sites proportional to its size. The minimum number per Vice County (varies from 1 to 8 depending on size) can be exceeded, and if there is sufficient volunteer capacity in some areas, we will allocate additional random squares. Once the initial quota of squares is allocated in each Vice County, additional squares can be allocated to volunteers on the basis of proximity to their homes.

We have excluded any squares with greater than 50% urban or suburban cover, water, estuary and closed forest. No other stratification will be performed at this stage. It might be desirable for us to stratify to ensure various land cover types are proportionately represented, but this would require significant additional GIS resources and complexity that are unavailable at present. We will review the situation in due course when we have more experience of volunteer take-up. The Vice County stratification should ensure a representative spread of land covers anyway. In comparison, the BTO’s Breeding Bird Survey does not stratify samples according to land cover, only according to volunteer density.

Each square will be surveyed following a fixed protocol that will be posted here on the NARRS website shortly. A combination of visual searching and refugia checking will be encouraged. The latter will include any pre-existing refugia that the surveyors finds, but the surveyor will also be encouraged to lay artificial refugia where appropriate, subject to safety and permission. The surveyor will be able to choose the number, locations, and materials (e.g. metal, rubber, felt, Onduline) of refugia used. Training, licensing, health and safety, and animal welfare will be of paramount importance. It will be essential to keep good records of the methods used and other parameters, so that weightings can be applied to results later.

A survey pack will be downloadable from this website, and an online recording form will be posted soon to enable electronic submission of results. Instant feedback will be provided, and the data will be analysed in due course. Hard copy forms will be available as downloads too, but we will encourage online submission.

In the event of no reptiles being encountered by a surveyor after three survey visits, we will offer the option of another random square or a nature reserve or other ‘higher quality’ square. The Wildlife Trusts, Defence Estates, National Trust, Forestry Commission and all local authorities approached so far have agreed to participate in principle (subject to protocol, access arrangements and permissions). This will be arranged centrally by the HCT as much as possible, but volunteers will need to liaise and meet the relevant local arrangements too. Additional sites such as these will only be allocated once a random square has been surveyed.