Male reproductive traits of full-sibs of different age classes in three-spined sticklebacks (Gasterosteus aculeatus)

SpringerPlus

Table 1 Listing of all conducted linear mixed effect models “lme”

Dependent variables	Explanatory variable male age		Explanatory variable body size		AIC male age	AIC body size	Sample size (N)
	χ²	p-value	χ²	p-value			Young	Old
Body size [cm]	47.842	<0.001	/	/	61.707	/	60	101
Body mass [g]	39.942	<0.001	/	/	−90.424	/	60	101
Body condition	0.448	0.503	/	/	−315.952	/	60	101
Red area [%]	0.192	0.662	5.245	0.022	607.296	602.243	59	101
Carotenoid chroma (A)	5.312	0.021	4.447	0.035	−59.083	-58.218	59	99
UV chroma (A)	11.091	<0.001	0.855	0.355	−538.148	−527.912	59	99
Red/orange hue (A) [nm]	3.985	0.046	1.269	0.260	43316.630	43319.350	59	99
Total brightness (A)	8.210	0.004	6.903	0.009	−1240.811	−1239.504	59	99
Total brightness (R)	16.374	<0.001	1.595	0.207	262.555	277.334	59	101
Absolute testis mass [g]	54.517	<0.001	76.359	<0.001	−452.293	−474.135	60	96
Relative testis mass (GSI)	20.887	<0.001	3.304	0.069	45.989	63.572	60	96
Testis melanisation (L*_total)	64.591	<0.001	19.391	<0.001	1955.718	2008.869	60	96
Sperm number	4.256	0.039	11.441	<0.001	223.120	207.984	59	99
Head length [μm]	73.952	<0.001	15.977	<0.001	−277.861	−219.886	60	22
Tail length [μm]	11.816	<0.001	6.773	0.009	−163.597	−158.555	60	22
Head width [μm]	53.729	<0.001	9.646	0.002	−303.683	−259.599	60	22
Mid-piece width [μm]	4.268	0.039	0.676	0.411	−271.444	−267.852	60	22
Mid-piece volume [μm³]	9.003	0.003	3.934	0.047	−298.985	−293.916	60	22
Mid-piece length [μm]	2.537	0.111	2.369	0.124	−225.356	−225.187	60	22
Head to tail length ratio	0.052	0.819	0.036	0.849	−540.046	-540.030	60	22

(A) colour measurements were made directly below the eye (breeding coloration).
(R) measurements were taken at the dorsum below the first dorsal spine.
(GSI) gonadosomatic index after de Vlaming et al. (1982).
(L*_total) testis brightness (see Mehlis et al. 2012 for details).
In all models family identity was included as random factor and never removed to control for possible ancestral influences. The model with the lowest AIC (Akaike's information criterion) value represents the best approximating model.