The lack of a role of body asymmetry in shaping tick load in European robins (Erithacus rubecula) during autumn migration

Description

Data to reproduce the analysis performed in The lack of a role of body asymmetry in shaping tick load in European robins (Erithacus rubecula) during autumn migration currently in final review at Ibis. Abstract: Fluctuating asymmetry (FA) of the body, reflecting developmental instability, has been linked to fitness traits in birds. However, its potential role in shaping susceptibility to ectoparasitism remains largely unexplored. We investigated the relationship between FA in tarsus length and tick parasitism in European robins Erithacus rubecula during autumn migration on the Baltic Sea coast in northern Poland. We examined 198 robins for tick infestations. FA was quantified as the difference between lengths of the left and right tarsus. The prevalence of tick-infested birds was 49.5%, with ticks predominantly identified as Ixodes ricinus (98.5%). Negative binomial models revealed that migration period was the sole predictor of tick abundance, with early migrants carrying significantly more ticks than late migrants. We found no association between tarsus FA and either total tick load or body side-specific tick presence. Although our results are correlative, they suggest that moderate asymmetry in robin leg morphology may not influence susceptibility to ticks. However, this susceptibility might depend on the context, e.g., the condition, sex and age of birds, or their stopover length, which suggests directions for further research. Our results highlight the importance of seasonal factors in host-parasite dynamics.

Steps to reproduce

robin.txt - full dataset used for the analysis robin_info.txt - general information to variable names used in robin.txt robin_script.R - R-script for analysis associated with the above mentioned manuscript. Data analysis Before modelling, birds with incomplete measurements were removed, retaining only those for which all focal variables (morphological and tick-related) were available (Table 1). The variable AsymTarsus was obtained by subtracting LeftTarsus from RightTarsus; positive values indicated a longer right leg and negative values indicated a longer left leg. Explanatory variables considered in the model were AsymTarsus, age, and the period of robin migration. This period refers to the timing within the autumn season, divided into the peak migration period (early period: 26 September–13 October 2022) and the late migration period (24 October–5 November 2022), based on the many-year migration dynamics of robins in the sampling area (Busse & Halastra 1981, Nowakowski et al. 2005). All statistical analyses were carried out in R (ver. 4.1.3). A Poisson GLM was initially fitted to examine whether total tick abundance (TotalTicks), treated as response variable, was related to the asymmetry index AsymTarsus and covariates age and period. Deviance residuals and overdispersion tests (using the dispersion test function from the AER package) were used to assess model fit. In case of overdispersion, a negative binomial GLM (glm.nb from the MASS package) (Venables & Ripley 2002) replaced the Poisson approach. Model selection was conducted by sequentially removing terms with drop1(..., test = "Chisq") and retaining predictors whose removal significantly increased model fit. Likelihood ratio tests were used to assess the significance of each predictor. We also tested models including interaction terms (e.g. AsymTarsus × Period, AsymTarsus × Age), but none of the interactions were statistically significant and, hence, are not presented in the text. An additional binomial GLM with a logit link function was used to analyse the relationship between side-specific tarsus length (explanatory variable) and tick presence/absence on the corresponding side (dependent variable, e.g. TicksRightOccur: yes=1, no=0). The testing procedure was the same as in the total tick load model. Robins cannot be reliably sexed in the field, particularly given that individuals may originate from different breeding populations, including not only birds from Scandinavia, where sex-specific morphometric features have been established (Polakowski & Jankowiak 2012), but also from other breeding grounds (Remisiewicz 2002). Thus, we were unable to include sex as a reliable covariate, which limited the detection of sex-specific effects (Dudek et al. 2016).

Institutions

• University of Szczecin

  West Pomerania, Szczecin
• University of Siedlce

  Mazovia, Siedlce

Categories

Licence