Table 2.

Competition does not change slopes. Neither competition nor species identity had a statistically significant effect (p > 0.1) on the slope of log(variance) as a function of log(mean), nor were there statistically significant effects on the slope of interactions of log(mean) with competition or interactions of log(mean) with species identity, by ANCOVA.

	d.f.	sum-squared	mean-squared	F-value	Pr(>F)
ANCOVA table to test for the effect of competition on regressions slopes. Response: log(variance)
log(mean)	1	78.626	78.626	706.1704	<2e−16***
competition	1	0.186	0.186	1.6662	0.2081
species	1	0.131	0.131	1.1747	0.2884
log(mean) × competition	1	0.096	0.096	0.8630	0.3614
log(mean) × species	1	0.296	0.296	2.6541	0.1153
residuals	26	2.895	0.111

***p < 0.001.