Nest searching and nest-site measurements
Understanding habitat features necessary for cranes to choose a site to locate a nest in the marshes necessitated nests access after mapping of territorial pairs. This was carried out between December 2017 and October 2019, thus traversing three breeding seasons. Location of nests was discerned from cues such as nesting pair behaviour when an individual was going for a nest relief. A pair of 10x42 binoculars and a telescope were additionally used to scan marshes for incubating cranes. Collecting data on nesting waterfowl however involves potential disturbance to nesting birds, through direct (e.g., handling of eggs) or indirect (e.g., walking near birds on nests) effects (Austin & Buhl, 2008). Grey Crowned Crane nests are also difficult to locate from the ground (Morrison & Bothma, 1998). With these considerations, nests were not actively searched as this would have possibly resulted to trampling on vegetation in the breeding habitats, cause disturbance to nesting birds and expose clutches and chicks to egg poachers. Nests were thus visited once to record measurements.
Measurements of variables at an active nest-site (i.e. a nest having an egg and/or chick) and at the random nest-sites followed Dwyer & Tanner (1992) and Wu, Zhu, Zhang & Yang (2009). A random nest-site for each active nest was selected to compare characteristics of active nests (presence, scored as 1) with those of randomly-selected sites (unused areas or absent nests, scored as 0). The random nest-sites were placed at 15-25m in any direction of the actual nest to avoid territory overlap with neighbouring nests (this study established a minimum distance of 47 m), and to ensure this random location was within the marshes (for nests located very close to the shoreline). Random nest-sites were not necessarily sites that cranes might never, or have never used before, but were unused sites during that specific season. Each pair of nests (active and randomly-selected) were then coded and included as a random factor in the mixed-effect models to control for possible non-independence between pair of nests. Some of the measurements recorded were: nest dimensions, mean water depth and vegetation height (over four samples taken at 1.5 m from centre of the nest in each cardinal point), offshore distance of nest, distance to the nearest active nest, distance of nearest homestead, and grazing (from livestock or wildlife) intensity that was assessed around the nest and recorded as either 0 (no faecal remains, vegetation intact, no spoors), 1 (low: scattered faecal, slight grazing, scattered spoors), 2 (medium: faecal remains evident, vegetation grazing more evident) or 3 (high: faecal remains around the nest, dense spoors and intensive grazing).