As in earlier this website studies, Voronoi tessellation was used to transform the discretely sampled surface into a continuous map using the assumption that each point on the map has the response characteristics of the nearest recording site (Kilgard and Merzenich, 1998). Because regions with above average sampling density have smaller tessellations, they do not bias estimates of the cortical response. A1 sites were identified on the basis of latency and topography. The percent of the cortical area of A1 responding to each tone was estimated as the sum of the

areas of all tessellations from sites in A1 with receptive fields that included the tone, divided by the total area of the field. For the time course study in which animals were mapped after NBS-tone pairing alone, we measured the percentage of A1 cortex that responded to the frequency that was paired with NBS, a 19 kHz, 60 Selleck CB-839 dB SPL tone. For all behaviorally trained animals, we reported changes in the representation of behaviorally relevant tones by reporting the ratio of the percent of cortex that responded to a 2 kHz, 60 dB SPL tone divided by the percentage of cortex that responded to a 19 kHz, 60 dB SPL tone. In behaviorally trained animals, we commonly observe both a shift in tuning toward behaviorally relevant tones and a decrease in receptive field sizes (Figure S1). The net effect of

this plasticity is to cause the cortical response to behaviorally irrelevant tones

to decrease whereas the response to behaviorally relevant tones is only slightly increased or unchanged (Figure S1). Therefore a ratio measure provides a reliable indicator of the relative frequency organization of low versus high tones in A1. Discrimination performance was measured using the signal detection Thymidine kinase theory measure d′ during all stages of training ( Abdi, 2010 and Klein, 2001). Statistical comparisons between three or more groups were done using repeated-measure ANOVA. Tones <0.38 octaves above the target stimulus were excluded from the repeated-measure ANOVA because these sounds were not reliably discriminated from the target stimulus and therefore were not expected to change significantly after NBS-tone pairing. Statistical comparisons between only two groups and single tone frequencies relative to zero were done using t tests, and t tests were used for all statistical comparisons of physiological measures between two groups. Unless otherwise noted, p-values reported are for two-tailed t tests. This work was supported by the James S. McDonnell Foundation. The authors would like to thank the many undergraduate volunteers who contributed to data collection and would like to thank Nick Reed, Crystal Engineer, Navzer Engineer, Kamalini Ranasinghe, Jai Shetake, and Benjamin Porter for comments on the manuscript.