Leks in ground-displaying birds: hotspots or safe places?

We studied the effects of female distribution, topography, and human infrastructures on lek-site selection in a ground-displaying bird, the great bustard Otis tarda. Our aim was to investigate the relative importance of maximizing visibility to females (hotspot hypothesis) versus minimizing predation risk (predation avoidance hypothesis). Using a geographic information system, a very high-resolution digital elevation model, and an extensive survey database, we compared the 350 lek centers known in Spain with randomly generated sites. Males preferred lek sites that increase their visibility to surrounding females, reduce the distance to them, enhance the probability of detecting predators, and increase the distance to human disturbance sources. Logistic regression analyses confirmed these patterns. A final model identified five relevant variables indicative of higher short-range visibility, more centered location with respect to females, and farther distance to human disturbance sources in lek centers compared with random points. The viewshed gain of lek centers relative to random points was maximal at the center and disappeared gradually at randomly generated replicas located progressively farther away from the center, but still within a buffer equivalent to the home range of the male flock during the display period. Visibility maximization may be the reason why lek centers are extremely fixed within and between consecutive breeding seasons. Our results are compatible with both hotspot and predation avoidance hypotheses. We suggest that both models are important, at least for species displaying in open habitats. Lek-site selection probably represents a combined response to female attraction and predator avoidance selective pressures.