Table 2. Threshold of the discrimination performance for ripple detection.

The threshold is the ripple height where bats exceeded a 0.7 probability of a correct choice. For each bat, 1000 simulated discrimination thresholds were generated with a binomial generalized linear model. The lower 0.025 and upper 0.975 percentiles of those data give lower and upper values of the 95% confidence intervals. Noise treatments have been abbreviated here (as compared to the text) to reduce visual clutter (smooth non-overlapping noise = “Non-overlap”; smooth-overlapping noise = “Smooth-overlap”; sparse-overlapping noise = “Sparse-overlap”).

Noise	p	Threshold	Bat	Lower 95% CI	Upper 95% CI
Silent (Control)	0.7	7.05	A	5.57	8.37
Silent (Control)	0.7	5.39	B	4.42	6.41
Silent (Control)	0.7	7.58	C	5.47	9.81
Silent (Control)	0.7	11.42	D	8.02	15.14
Non-overlap	0.7	5.83	A	4.05	7.49
Non-overlap	0.7	3.49	B	2.44	4.63
Non-overlap	0.7	11.62	C	6.16	25.07
Non-overlap	0.7	12.14	D	9.37	14.98
Smooth-overlap	0.7	6.91	A	4.59	9.15
Smooth-overlap	0.7	6.60	B	4.81	8.38
Smooth-overlap	0.7	22.92	C	11.77	702.42
Smooth-overlap	0.7	621.33	D	25.87	>100000
Sparse-overlap	0.7	7.63	A	4.97	10.81
Sparse-overlap	0.7	4.54	B	2.85	5.95
Sparse-overlap	0.7	28.40	C	10.32	>100000
Sparse-overlap	0.7	4395.17	D	18.24	>100000