Territory mapping in kilometre squares

Thanks to the 1993–1996 atlas and the common breeding bird monitoring scheme, many volunteers were already acquainted with the simplified territory mapping method. Applying the same method to the 2013–2016 atlas was therefore a natural choice. The method involves surveying a kilometre square three times during the breeding season following a predetermined route. Only two visits per breeding season are required in kilometre squares above the tree line. Territories are then mapped based on the observations recorded during the visits. This method has the advantage of not having to survey the atlas square in full; rather, a representative sample of selected kilometre squares allows us to derive information for the whole area. Simulations run prior to fieldwork showed that a sample of five kilometre squares per atlas square was sufficient given the modelling methods that had been developed since the last atlas. Only three to four kilometre squares were selected in high-altitude atlas squares and those containing large areas of open water. In the end, 2318 kilometre squares were surveyed for the 2013–2016 atlas compared to 2943 in 1993–1996. Thus, 5 % of the total area within the atlas perimeter were mapped in 2013–2016.

Selecting the kilometre squares

Three conditions had to be met when selecting the five kilometre squares per atlas square:

1. All kilometre squares surveyed for the common breeding bird monitoring scheme (Monitoring Häufige Brutvögel MHB) and «Biodiversity Monitoring Switzerland (BDM)» were included in the atlas sample. The annual surveys of the 267 MHB-squares continued through 2013–2016. One fifth of the 264 BDM-squares are surveyed every year, which means that each square is surveyed every five years. To allow for the inclusion of all BDM-squares in the atlas calculations, we incorporated data from 56 BDM-squares surveyed in 2012.
2. Of the kilometre squares surveyed in 1993–1996, as many as possible were retained.
3. The five kilometre squares were representative of the atlas square in terms of habitat and altitude. The following habitat types and their respective area were considered when making the final selection: forest, open landscape, settlements, wetlands including lakes, length of roads and pathways.

Accessibility was also taken into account, especially at high altitudes. Kilometre squares that were poorly accessible and considered too dangerous to survey were omitted. While kilometre squares above 2300 m were underrepresented in 1993–1996 and none above 2600 m were included, the altitudinal distribution of the kilometre squares in 2013–2016 was more balanced. The highest kilometre square reached an altitude of 3100 m.

Ultimately, a total of 2318 kilometre squares were surveyed, including 267 MHB-squares, 264 BDM-squares, 1600 kilometre squares surveyed in 1993–1996, and 187 newly selected kilometre squares. The survey visits (except MHB-squares) were distributed over the four field seasons as follows: 505 kilometre squares in 2013, 606 in 2014, 569 in 2015 and 315 in 2016.

Survey procedure

A simplified territory mapping survey was to be conducted in each kilometre square in one of the four field seasons 2013–2016. Only the MHB-squares were surveyed annually. The survey year was fixed in the case of BDM-squares, but for all other kilometre squares, observers were free to choose the year in which to conduct the surveys.

To make sure surveys were comparable across kilometre squares, we made the following recommendations: In lowland squares, the first visit should take place between 15 and 30 April; 2013 was an exception, when fieldwork was allowed to start on the weekend of 13 and 14 April. The second visit followed in the first half of May. The third and last visit was to be carried out between mid-May and mid-July. At higher altitudes, we asked that visits be scheduled between the end of April and the end of June, in high-mountain areas between late May and mid-July, depending on snow conditions.

We chose to start the field season on 15 April in accordance with the 1993–1996 atlas and the MHB. By this date, many residents and short-distance migrants have established their territories, at least at lower altitudes. Only a few species whose singing activity peaks early in the year (e.g. woodpeckers, Eurasian Nuthatch and tits) were at risk of being underestimated as a result. Phenological data for several species were analysed prior to the atlas surveys to assess whether the timing of the field surveys had to be advanced. The results showed that according to the available data, the activity peaks had not changed substantially since 1993–1996. The reference date of 15 April was therefore retained.

In view of the territory mapping that would follow, records of simultaneous observations were particularly important (e.g. two singing males observed at the same time). We therefore regularly reminded our volunteers to pay particular attention to simultaneous observations.

The routes for all kilometre squares were predetermined by the atlas team, an approach that had proven effective in the MHB. However, if deviation from the defined route was unavoidable, the adjusted route had to be drawn onto the field maps and followed during subsequent visits. The routes were chosen so as to cover as much of the kilometre square as possible and were generally 3–6 km long. Observers were asked to choose a different starting point or direction on each visit, if possible. Areas that were inaccessible, e.g. due to cliffs, were surveyed as well as possible using binoculars. During the 1993–1996 surveys, by contrast, observers were free to choose a different route for each visit.

In keeping with the MHB surveys, we aimed to record the total population of each species. Also, we required all observations of individuals of all species to be recorded. This approach differed from 1993–1996, when we had defined an upper limit for territories of widespread species; once this limit was reached, the species no longer had to be reported in that kilometre square. In the case of Common Chaffinch, for example, recording could stop once more than ten territories had been noted. In addition, the list of species to be recorded in 1993–1996 was more or less limited to those classed as widespread.

Processing the survey results

The field maps used for territory mapping were redesigned and tested during the MHB surveys in 2012. The new A3-size maps were fitted with four QR codes (Quick Response Codes) for automatic georeferencing. Following the survey visits, observers sent the field maps by post to the Swiss Ornithological Institute, where they were scanned and uploaded to «Terrimap online» for further processing.

Digitising all species records marked on the field maps and mapping the territories was the observers’ responsibility. Only observations that met the species-specific inclusion criteria were used to define territories. The «Terrimap online» system would not allow users to map a territory based exclusively on observations made before the reference date or on insufficient Atlas Codes. On the other hand, observations that met the criteria could be used to define a territory even if they were recorded during a single visit. Territories that extended beyond the square's border were counted as long as one observation was located within the kilometre square. No territories were defined for birds that were obviously passage migrants, altitudinal migrants or wandering individuals. Foraging birds such as hunting raptors were included as long as there were potential nest sites nearby.