*2.5. Statistical Analysis*

The correlation between the concentrations of GC in the di fferent sampling tissues (i.e., plasma, feces and hair) was tested using Pearson's correlation. To analyse the physiological stress reactions, general linear models (LM) were used to test the e ffects of sex, age-class and month (independent variables) on the di fferent GC levels (cortisol in plasma and hair, FCMs; dependent variables). Additionally, to evaluate the influence of the physical condition and the immunological status of the animals on GC levels, BMF and percentage of lymphocytes (the most abundant white blood cells (WBC)) were included in this analysis as independent variables. The concentrations of GC or its metabolites were transformed using a log transformation. BMF and WBC were logit transformed to achieve an approximation of a normal distribution and to reduce heterogeneity [46]. Since no significant interaction e ffects between the independent variables were found, only the main e ffects were used in the final models. The year had no significant e ffect on the concentrations of cortisol in plasma (F(1,48) = 1.989; *p* = 0.165) or hair (F(1,74) = 0.022; *p* = 0.882), nor on the concentrations of FCMs (F(1,72) = 2.691; *p* = 0.105). Therefore, data of both years were pooled.

To evaluate the physical and immunological conditions of red deer, we tested the e ffects of sex, age-class and month (independent variables) on BMF and WBC (dependent variables) using general linear models (LM).

The results are expressed as mean ± SE and 95% confidence intervals (CI). All the statistical tests were considered significant when *p* < 0.05. The statistical analyses were performed using IBM.SPSS ®, version 22 (IBM Corporation, New York, NY, USA).
