Geospatial information and 3D modelling of spatially-complex cliff ledges used by breeding seabirds

Credit: Elspeth Kenny and Jamie Thompson

Nest site selection is an important component of reproductive success of many birds, amphibians and reptiles. This behaviour is defined as the placement of eggs by a female in a non-random site within a particular habitat. Characteristics of the nest site determine how well the egg is protected against predators and extreme weather, and how well the egg is physically supported – holding the egg in place and preventing it from rolling away.

Many breeding colonies of common guillemots (Uria aalge) use narrow, sloped cliff ledges to lay their eggs and do not construct a nest. The stability of their eggs against rolling down a slope or off the cliff ledge is likely dependant on microhabitat surface features and the shape of the egg. Egg shape of common guillemots is thought to show strong repeatability within individuals, and thus, egg shape may be considered by females when selecting an appropriate nest site.

Many eggs are tucked next to a wall or in a depression to prevent them from rolling away. Credit: Jamie Thompson.

Research project

In this study, we evaluated whether variation in egg shape influenced nest microhabitat choice of female common guillemots on Skomer Island in Wales. We anticipated that birds that typically lay rounder eggs, will prefer surface features that promote stability (reduce movement) of the egg. Surfaces with higher stability will have higher surface complexity, reduced slope, depressions, and presence of physical barriers like walls and pebbles. 

Over 2000 photographs were taken of two cliff-ledges during the peak of the breeding season, after eggs were laid. We used markers to identify where eggs were laid. Large scale bars were randomly placed on the ledge to scale the 3D models during post-processing.

Breeding cliff ledges were photographed to create high-resolution 3D models. This photo shows labelled markers placed next to each egg, before the egg was temporarily removed to image the surface. Credit: Jamie Thompson
Large grid markers were placed randomly on the site so I could get an accurate scale during post-processing.

All photographs were imported into photogrammetry software (Agisoft photoscan) to generate high-resolution 3D models of each cliff-ledge. To quantify spatial properties of the rocky nest surface, a digital elevation model (DEM) measuring 10 x 10 cm per nest site was created from the dense point cloud. The DEM represents the model surface (topography) as a regular grid of interpolated elevation data.

Example of digital elevation model of surface created with Agisoft Photoscan. The blue boxes are separate photographs that are stitched together. The green object in the centre is an easter egg from Germany, where the first trials were conducted.

Five spatial parameters were to describe the surface tomography of bird nest sites: slope (steepness of terrain), aspect (direction the slope faces), curvature (slope of the slope, or derivative of the surface), roughness and relative depth. Slope and aspect were used to identify the direction and extent to which an egg was at greatest risk of rolling from the nest site. Curvature was used to identify convex surfaces, like ridges, that are more exposed but allow water and bird faeces to drain to other surfaces, and concave surfaces, like depressions, that are generally more sheltered but receive drainage from other areas. Rough surfaces exert greater friction, therefore, will slow down the progress of a rolling egg once it is already in motion. Spatial parameters were obtained using ArcGIS and SAGA GIS.

This manuscript is currently being prepared, so please stay tuned.


Attard MRG, Thompson J, Mallison H and Birkhead T (In prep) Geospatial information and 3D modelling of spatially-complex cliff ledges used by Common guillemots (Uria aalge)