Table 3.

Comparison of various measured interaction features for different outcome types, also showing result of Kruskal-Wallis rank tests to check for differences in the given variable (χ² statistic and p-value).

Measured feature		Interaction outcome						χ2	p-value
		Copula		Passive rejection		Active rejection
Total number of outcomes observed		41		77		142		—	—
Median (IQR) contact time (s)		21.45 (16.41, 26.24)		5.33 (2.57, 11.55)		0.84 (0.33, 1.56)		131.63	2.61 × 10 −29
Mean (±sd) number kicks		5.63 ± 4.96		5.43 ± 4.60		5.06 ± 4.32		0.55	7.61 × 10−1
Median time to first kick (s)		0.17		0.13		0.11		7.49	2.37 × 10 −2
Median duration of kick period (s)		1.14		0.83		0.68		6.35	4.17 × 10 −2
Median kick rate (kicks/s)		4.74		6.86		7.41		8.35	1.54 × 10 −2
Number exhibiting flight cessation		39		74		14		—	—
Median (IQR) time to cessation (s)		2.74 (1.79, 5.39)		1.47 (1.05, 2.23)		1.05 (0.56, 2.03)		24.92	3.87 × 10 −6
Median time (s) |% terminated	Move to front	0.30	—	0.49	10.3%	0.34	54.9%	3.93	1.40 × 10−1
	Secure tarsal hold	1.12	—	0.97	18.2%	1.17	39.4%	—	—
	Ventrally align	2.28	—	3.19	64.9%	2.26	3.5%	—	—
	Genital contact	4.58	—	—	6.5%	—	2.1%	—	—

In each case, normality of measures was tested using Shapiro-Wilk tests. When calculating contact times, only those interactions that concluded before the end of the video recording period (n = 234) were included. The median kick period and kick rate tests include only those interactions that involved more than one kick (i.e. kick period ≥ 0, n = 215). Flight cessation is classified as the point at which paired flight ended (i.e. when either individual stopped). For rejections, the median time to each pre-copulatory behaviour identified in Fig. 1 is given for those interactions in which the stage was completed. Also presented is the proportion of all rejections that occurred between each step. Kruskal-Wallis tests were not performed on time measurements for males securing female tarsi and paired ventral alignment due to the small sample sizes involved.