Factors influencing the abundance and distribution of feral pigeons (Columba livia) in urban environments in Malta

Abstract

Feral pigeons (Columba livia) have long been regarded as a pest in urban environments. A key first step for effective pest control is to evaluate the population size. Apart from studying the abundance and distribution of the bird in selected urban environments on Malta, the present research aimed to investigate which, and how, urban environmental characteristics influence feral pigeon numbers. The tool used for this was Distance Sampling, a frequently used methodology for estimating animal density and/or abundance. Three towns from each district of mainland Malta were chosen. (Southern Harbour: Paola, Kalkara, Fgura; Northern Harbour: Msida, Swieqi, St Julians; South Eastern: Birzebbugia, Marsaskala, Marsaxlokk; Western: Iklin, Mtarfa, Rabat; Northern: St Paul’s Bay, Mgarr, Mellieha. Ten transects per town were then allocated along roads categorised in five urban environment types: (i) Local centres, (ii) Residential areas, (iii) Residential Priority Areas, (iv) Public open spaces within town, (v) Public open spaces at periphery. After a pilot study to select the transects, the actual study was implemented between May and August 2020. Fieldwork was carried out during weekdays only, from 8:00 till 12:00, to ensure consistency. Each transect was walked in one direction taking note of the frequency of occurrences of various urban environmental parameters on one side of the road and was then walked back on the other side of the road. The walking session was then repeated this time recording any feral pigeons present whilst measuring their perpendicular distance to the transect line using a laser rangefinder. The data were compiled into a matrix of feral pigeon abundance within the districts, localities and transect types. To test whether there was a correlation between the abundance of feral pigeons and urban environment characteristics, Generalized Linear models (GLM) were used, specifically the Negative Binomial Regression after first using the Spearman’s Rank-Order Correlation to check for multicollinearity. The abundance and density of the feral pigeon population in the urban setting was subsequently estimated using Distance Sampling Analysis. Significant differences in pigeon abundance was found between transect types paired up with either ‘Mainroad’ and ‘Park’, suggesting that they share features that attract pigeons. The Spearman’s Rank-Order Correlation indicated that there were various urban environmental variables that resulted as significant, however, ‘Old Buildings’ and ‘Modern Buildings’ were the two that had the highest correlation coefficient (0.507 and -0.420 respectively). Of a number of models tested, Model 6, which included only the predictor variable ‘Modern Buildings’, resulted as the best fitting, hence showing that ‘Modern buildings’ is an urban environmental characteristic that significantly and negatively affected the feral pigeon abundance, that is, feral pigeon abundance is low where modern buildings are predominant. The Distance Sampling Analysis showed that, the total abundance of feral pigeon clusters was estimated to be 293.89±70.87, each cluster having two or three pigeons, and with the ‘Mainroad’ transect type having the highest numbers of clusters, followed by ‘Park’, ‘Sidestreet’, ‘Agricultural area’ and ‘Suburb’ in this order. The highest estimate of cluster density of feral pigeons was for ‘Park’, followed by ‘Mainroad’, ‘Agricultural area’, ‘Sidestreet’ and ‘Suburb’, in this order. The present study demonstrates that urban characteristics do indeed influence the abundance and distribution of feral pigeons even if the birds travel long distances and have home ranges that include multiple habitat types. The present results can serve as a stepping stone for further research on feral pigeon populations and urban ecology in general.