Display site selection in a ground dwelling bird: the importance of viewshed

Abstract

We studied the effects of visibility, female and male distribution, microhabitat and distance to human infrastructure on display site selection in a ground-dwelling bird, the Canarian houbara bustard. Using a very high-resolution digital elevation model based on LIDAR technology, and a complete census of the breeding population, we compared 98 display sites with randomly generated sites through generalized linear models. Univariate analyses showed that males displayed at locations that increased their visibility, both at short and long distances. Interestingly, although numbers of females and males around sites did not differ between display and random locations, from display locations males could see more females and males at both distance ranges. The absence of vegetation and stones was also critical as it allowed males to perform display runs on a ground free of obstacles. The amount of trophic resources did not correlate with the selection of the display site itself, though an appropriate vegetation cover seemed to be important at a wider habitat scale. Finally, display sites were farther away than random sites from sources of human disturbance, such as urban nuclei, buildings and tracks. Logistic regression analyses confirmed the importance of viewshed, low stone and vegetation cover, and distance to urban centres and tracks, and model averaging identified short-range visibility and females visible in the long range as the most important visibility variables. These results are compatible with the sexual advertisement and predator avoidance hypotheses. We provide recommendations to ensure an appropriate management of the breeding habitat of this endangered subspecies.

Keywords: Canarian houbara bustard; Chlamydotis undulata fuertaventurae; display site; microhabitat; visibility.

Figure 1

Map showing the distribution of display sites (red dots) and random sites (blue dots) within the distribution range of Chamydotis undulata fuertaventurae (yellow-shaded area). Urban centres and roads are shown in black. The distribution range was calculated as a combination of the kernel 95 and a buffer around each bustard location during the breeding population census (see Section 2 for details).

Figure 2

Details of 10 display sites of our sample. The band of ground where the male performs its circular displays can be clearly identified. In some images, the male’s footprints are visible. Photos: Carlos Palacín, Alberto Ucero and Juan Carlos Alonso.

Figure 3

Maps of two sectors of the study area showing two Canarian houbara display sites and two random sites, overlaid on top of the DEM. The circles around display and random sites represent the two buffers (467 m and 3000 m, map A), or only the 467 m buffer (map B). Within the buffers, the area visible from the display site (green) and random site (red) are shown. Black dots are females and white dots are males, as recorded during the breeding population census. In example A, it can be appreciated that many more females and males are found within both buffers around the display site compared to the random site. Example B illustrates how a displaying male can see more females in the short-range buffer than it would see from a random site. The viewshed extends well off to the south, where at least 3 female locations exist.