Highland Breeding Bird Data 2002-2003 - non-sensitive breeding species

Quality Control

The data are of high quality in terms of identification and verification. However, see notes on Geographical Coverage (above) re resolution of records, and note that the dataset is currently biased towards less common species. Observer effort is spread thinly across a large recording area, such that many areas are poorly recorded. Breeding records do not imply that breeding definitely took place, as this category includes possible and probable breeding records.

Method steps

1. Data were not collected systematically. Records were submitted to the Highland Bird Recorder and came from a variety of sources, both amateur and professional. All records are validated by the recorder or by the appropriate committee for rarities (Highland records committee for local rarities, Scottish Bird Records Committee for Scottish rarities and British Birds Rarities Committee for national rarities). Records were mostly submitted in either hard copy or via the Highland Bird Recording spreadsheet. The latter is managed by Dave Butterfield (RSPB) and where possible DB categorised records according to their status (breeding, non-breeding, migrant, wintering etc). Some records came from repeat counts of local sites or reserves, others are completely random. Records came from local observers and from visiting birdwatchers. Because many observers submit their records with a view to their possible inclusion in a bird report, this results in an over-representation of records of the "rarer" and "more interesting" species. Consequently common species are often under-recorded. Records of common species with poor geo-referencing are generally excluded from the data set.

All records come from the Highland Recording area (see http://www.thesoc.myzen.co.uk/soc-recorders.htm) which comprises the old counties of Sutherland, Ross-shire and Inverness-shire. The latter includes the Isle of Skye, and the islands of Rhum, Eigg, Muck and Canna. Records are concentrated in areas of human population, and as Highland is both the largest recording area in Scotland and has the lowest human population density, absence of records does not infer that a species is absent! Only a small proportion of submitted records came with an accurate grid reference. Records without their own grid reference are allocated a centroid 1 km grid reference based on site name. As many of the sites are large, this centroid reference may introduce artificial precision when presented on the NBN Gateway. To reduce this effect records are presented on NBN at tetrad scale. However, it should be borne in mind that the true resolution scale for each record can only be inferred by reference to the site name attributes in the raw data. Where records are plotted at 1km or 100m precision, these are at true resolution , unless the record is considered sensitive, in which case a coarser precision may have been used.