Estimating the diet of urban birds: The problems of anthropogenic food and food digestibility

A variety of methods exist for assessing the diets of wild birds, each of which has its strengths and weaknesses. The objective of this study was to compare diet data collected in parallel by behavioural observations and faecal analysis from an urban population of Penelope superciliaris (rusty-margined guan) over a 12-month continuous period (April 2006 to March 2007). Visual observations (114 h) and faecal samples (N = 482) were analysed for food item content and material was identified to the finest level of taxonomic resolution possible. In total, 82 food items comprised the diet, including: fruits 67 items (81.7%), flowers 3 items (3.7%), insects 3 items (3.7%), seeds 2 items (2.4%), leaves I item (1.2%) and anthropogenic food 6 items (7.3%). However, the two methods produced significantly different results for the population's (our population consisted of at least 21 different individuals) mean monthly percentage of fruit (86.9% +/- 3.9 versus 47.2% +/- 6.8 for faecal analysis and visual observations, respectively) and artificial food (18.6% +/- 4.6 versus 0.1% +/- 0.1, for faecal analysis and visual observations, respectively) consumed (P < 0.001 in both cases). Importantly our analysis showed that faecal analysis when compared to behavioural observations overly simplified the population's diet both in complexity and temporally. One of the main problems we detected was with anthropogenic food that had a high digestibility. which meant that it was difficult to detect in faecal analysis as it appeared to be completely broken down and absorbed. These data demonstrate that digestibility should always be given due considerations in diet studies, especially in urban environments where anthropogenic foods can be a significant part of the diet (both in terms of proportion and energy) and tend to have high digestibilities. (C) 2008 Elsevier B.V. All rights reserved.