Avian density inflation in continental fragment islands is consistent with species impoverishment compared to nearby mainland

<title>Abstract</title> <p> We used data from two major continental fragment islands (Mallorca and Menorca) and from the nearby mainland (Catalonia) to test the avian density compensation hypothesis (i.e. abnormally high density of species inhabiting islands with species-poor biotas with a large data set. Our data sets covered 22 years in Catalonia, 10 years in Mallorca and 19 years in Menorca. We calculated island-mainland density ratios and their hierarchically bootstrapped confidence intervals for 87 <italic>species x habitat</italic> crossings, and then fitted logistic regression models to analyse whether island/mainland ratios were &gt; 1 as a function of diet, habitat and taxonomic Order. Additionally, we compared densities of the three most common breeding raptors on the islands. Average avian densities were 2.8 times higher in Mallorca and 2.2 times in Menorca. From all <italic>species x habitat</italic> crossings analysed density ratios were &gt; 1 in 73% of cases for Mallorca and 83% for Menorca. Some island top ten <italic>species x habitat</italic> crossings showed densities more than 20 times higher than those on the mainland. Raptors density ratios were 20–60 times higher in the islands. Most cases of higher island density were well explained by competitive release within guilds. Since under high densities birds are known to become more tolerant with conspecifics, we suggest a new explanation for island raptor sedentarism. When considering all diurnal breeding raptors as a test, the mainland/island density ratio was &gt; 1, which rules out the possibility that large island densities could be the cause of island species impoverishment rather than the consequence. </p>