Inference of Pair Bonds from Capture Data Based on Low Variation of the Sex Ratio Among Catches

Paired individuals of monogamous animal species often stay in close contact when moving in their home range and can be caught together if sampling is done by a method that does not disrupt their spatial association (e.g., capture of birds by mist nets). We propose a statistical procedure that uses the counts by sex of specimens in a number of samples to detect the presence of pairs and to roughly estimate the proportion of paired individuals in the population. This method appears to be efficient enough to be used with data sets of small to moderate size (10-40 samples of 2-10 specimens each). The use of the method is exemplified by an analysis of capture data from two European populations of bearded tit (Panurus biarmicus): a large set of 536 net samples (1,111 specimens) from Le Sambuc in Camargue, France, and a much smaller one of 46 samples (367 specimens) from Malcontenta, near Venice, Italy. A separate analysis of premolt juveniles in Malcontenta (15 samples, 53 individuals) detects a significant proportion of pairs among them. This finding confirms quantitatively the anecdotal observation of precocious pair formation in the bearded tit.