Latvian Ornithological Society Global alliance of conservation organisations working together for the world's birds and people

Results of the study in 2003–2005

Results of the study of distribution population size and habitat choice of Hoopoes in Latvia (2003–2005)

The first study of Hoopoes Upupa epops in Latvia was finished in June, 2005. It was started in 2003 in order to estimate the population size of Hoopoes in Latvia and to clarify their distribution and habitat choice. To achieve this goal all Hoopoe observations in 2000–2004 were gathered and compiled. Basic data sources were the second Latvian Breeding Bird Atlas and results of inquiry carried out in mass media.


To clarify current distribution of Hoopoes in Latvia the methods of Latvian Breeding Bird Atlas were used: all observations were registered in 2785 5x5 km squares in LKS-92 coordinate system. Observations from other sources were also attributed to this square grid. Only the data received until March 30, 2005 were used in the study.

The first Latvian Breeding Bird Atlas was compiled during years 1980–1984. The data for this period were gathered in 10x10 squares in UTM system. In order judge on the changes in the distribution of Hoopoes, data of the current study were transferred to UTM system.

In years 2000–2004 Hoopoe has been recorded in 122 5x5 km squares (Fig. 1). In 90 squares its breeding is possible, in 21 – probable and in 11 squares – confirmed. Results show that the distribution of Hoopoes in Latvia is uneven: in the district of Riga they occur in higher density than elsewhere.

Figure 1. The distribution of Hoopoe in Latvia in 2000–2004 (5x5 km squares in LKS-92 coordinate system). The colour from the lightest to darkest represents possible, probable and confirmed breeding respectively.

When comparing the results of current research with those of the first Latvian Breeding Bird Atlas it can be seen that in 42 squares Hoopoe has been recorded only in 1980–1984, in 65 – only in 2000–2004 but in 18 squares it has been recorded in both periods (Fig. 2). However, it is unlikely that this comparison shows real increase in the distribution of Hoopoes. The differences in the results are more likely to be attributed to changes in survey coverage, which was considerably poorer in 1980–1984. On the other hand the results show no shrinking of the breeding range of Hoopoes.

Figure 2. Comparison of Hoopoe distribution between the first (1980–1984) and the second (2000–2004) Latvian Breeding Bird Atlas in 10x10 km UTM squares.


The lower limit of population estimate was calculated taking into account all Hoopoe observations in years 2000–2004. Observation points that were closer than 450 m to each other were counted as a single territory unless it was certain that the birds belonged to different pairs. The upper limit of population estimate was calculated by taking into account the survey coverage of Latvia during Atlas project. 64.24% of atlas squares were covered satisfactorily and therefore the lower limit of population estimate was divided by 0.6424, assuming that the overall coverage matches the survey coverage of Hoopoes.

During years 2000–2004 the population size of Hoopoes in Latvia was 160–250 breeding pairs. 66–85 pairs (26–41% of the population of Latvia) of Hoopoes breed in the district of Riga. The overall breeding density of Hoopoes in Latvia is 0.25–0.39 pairs/100 km2, but in the district of Riga – 1.86–2.40 pairs/100 km2.

Although comparison of population estimates obtained by different methods is risky, the results of this study suggest that there is no reason to assume that Hoopoe population has declined in the last decade since the previous estimate: 100–300 breeding pairs.

Habitat choice

Only Hoopoe observation points with the precision of coordinates at least 50 m were used to judge on habitat choice (n=129). Habitats were analysed in two levels: in hypothetical territories of Hoopoes and at the exact observation points. The information on habitats was acquired from satellite map of Latvia.

To estimate the habitat composition in the territories of Hoopoes, circular buffers with radius of 450 m were drawn around the Hoopoe observation points. If observation points were nearer than 450 m to each other, they were attributed to one territory. In such cases a polygon was drawn using all the points of one territory. The 450 m buffer was drawn around the central point of this polygon. In such way 98 hypothetical Hoopoe territories were acquired.

All these territories were intersected with the habitat layer in GIS program ArcView 3.1. After that the areas of habitats within the territories were calculated. These results were compared to those acquired from 98 randomly selected circular polygons with radius of 450 m in all Latvia using Mann-Whitney test.

As said before, habitat composition was calculated also at the exact points of Hoopoe observation. In this case 50 m buffer was drawn around each of the 129 points. This distance was chosen because it matches the error in the measured coordinates.

By analysing the results some significant relationships can be seen. These clearly result from the habitat features required by Hoopoes – bare or short-vegetation land and features in which to breed. Hoopoes are stronger connected to the habitats that provide both of these features: built-up areas and seasonal villages. It must be noted that seasonal villages are a suitable habitat for one of the most important food items of Hoopoe: mole-cricket Gryllotalpa gryllotalpa. Hoopoes were significantly associated also with habitats that are valuable as feeding sites: pits and peateries. This study gives only general information on habitat choice of Hoopoes in Latvia and further studies are needed to clarify this question.

Figure 3. Breeding habitat of Hoopoe Upupa epops – seasonal village near Tīraine (district of Riga).


Results of this study are compiled in bachelor’s thesis „Distribution, abundance and habitat choice of Hoopoes Upupa epops in Latvia” by Viesturs Ķerus. The thesis is in Latvian but it has English summary and figure legends.