Figure 5.

Massed activity provides odor dependent and limited concentration dependent information. The main effects of odor identity (A,B) and stimulus loading (C,D) on spike counts (A,C) and peak spike frequency (B,D). For this analysis spatial and temporal information are collapsed. By comparison to blank, all odors produced a significantly higher peak spike frequency (B; p < 0.01; Kruskal-Wallis) but not significantly different spike counts (A) though the trend was consistent with frequency. Note that there are no significant between-odor differences in either count or frequency. In response to different concentrations of odor stimulation, neat odor presentation resulted in a significantly greater overall spike count (C; p < 0.01) but not peak spike frequency. Conversely, blank stimuli resulted in significantly lower peak spike frequency than did odor at any stimulus loading (D; p < 0.01). (E,F) The two-way interactions of odor and stimulus loading on spike counts and peak spike frequency. Note in both measures that neat odor did produce some odor dependent effects, these appear to relate to odor volatility. (G) Mean response onset latency relative to stimulus onset (in ms) as a function of stimulus loading (blank responses dropped). Inset letters indicate statistical differences in latency (one-tailed paired t-test; p < 0.05). In all panels, error bars represent standard error.