Table 2. Parameters selected for Discriminant function analysis (DFA) and results of DFA for grouped calls.

Category	Number of calls in DFA	Parameters selected for DFA	Wilks' Lambda	Canonical correlation	Eigenvalue
	unselected/selected
a)
Social context (N = 117)	29/88	Call duration	0.553	Function 1: 0.802	1.807
		Mean fundamental frequency	0.729
		Relative amplitude changes/time	0.631
		Formant 2	0.682
b)
Sex (N = 112)	28/84	Jitter (a)	0.892	Function 1: 0.590	0.533
		HNR	0.736
		Formant 3	0.874
		Duration of tonal parts	0.753
c)
Males (N = 82)	21/61	Maximum fundamental frequency	0.083	Function 1: 0.965	13.383
		Formant dispersion	0.164	Function 2: 0.912	4.925
Females (N = 28)	7/21	Mean fundamental frequency	0.054	Function 1: 0.988	39.517
		Formant 2	0.027	Function 2: 0.965	13.484

N =  number of calls per category, (a) indicate that parameter was not used to calculate discriminant functions.