Divergent Evolution of Male Aggressive Behaviour: Another Reproductive Isolation Barrier in Extremophile Poeciliid Fishes?

Table 2

Results from (a) MANCOVA and (b) ANCOVA models analysing attributes of dyadic male aggressive interactions in Experiment  2 (lab-reared males). 𝐹-ratios were approximated using Wilk’s 𝜆. Partial variance was estimated using Wilk’s partial 𝜂². Significant effects are in bold typeface.

	df	𝐹	𝑃	Partial variance explained [%]
(a) MANCOVA (number of aggressive behaviours)
Light (absent/present)	3,62	8.97	<0.001	0.30
Sulphide (absent/present)	3,62	2.45	0.072	0.10
Light × sulphide	3,62	3.77	0.015	0.15
Male body size difference	3,62	3.35	0.025	0.14
Mean pair body size	3,62	2.81	0.044	0.12
(b) ANCOVA (fight duration)
Light (absent/present)	1	8.44	0.005	0.12
Sulphide (absent/present)	1	7.07	0.010	0.10
Light × sulphide	1	2.48	0.120	0.04
Male body size difference	1	0.25	0.622	<0.01
Mean pair body size	1	0.71	0.403	0.01
Error	64