Odorants tend to smell more intense with increasing vapor phase concentration, but most odorants continue to elicit the same quality of odor perception across a wide range of concentrations. The exceptions to this rule are notable, however. About 8% of the odorants surveyed in an archive of flavor and fragrance chemicals were reported to change in their odor quality with concentration (Arctander, 1994). The differences in odor can be striking: for example, beta-ionone is reported to smell like violets at low concentrations and like cedar at higher concentrations (Dravnieks, 1985). 

We studied the patterns of 2-DG uptake evoked by five aliphatic compounds differing in functional groups (Johnson and Leon, 2000a). Three of the compounds (valeric acid, 1-pentanol, and methyl valerate) were not reported by humans to change in perceived odor quality with stimulus concentration, whereas the other two compounds (2-hexanone and pentanal) were reported to evoke distinct odor perceptions at high and low concentrations. When we expressed our data in units of z scores to illustrate the pattern of uptake independently of the amount of uptake, we found that the evoked activity patterns were constant with concentration for the three odorants that displayed perceptual constancy, and that the activity patterns differed with concentration for the two odorants that differed in perceived odor at high and low concentrations (Johnson and Leon, 2000a). This finding represents yet another correlation between activity patterns as measured by 2-DG uptake and perceived odor, assuming that rats perceive the odors of these compounds in a manner similar to humans. These data also suggest that odor quality perception may be coded in terms of the relative pattern of activity in particular neurons compared to the overall level of activity across the entire olfactory bulb. It may be the case that the perception of odor changes when concentration is increased for some odorants because increasing the concentration of the target odorant also increases odorant contaminants that affect both neural pattern and odor perception. In the case of pentanal, the acid contaminant in pentanal, pentanoic acid, could reasonably account for the observed changes in neural response and odor perception (Johnson et al., 2004).  

The z-score activity pattern evoked by 2-hexanone changed with odorant concentration, whereas the z-score pattern evoked by valeric acid was constant across different concentrations.

When we expressed our data in units allowing us to see the amount of 2-DG uptake in addition to the pattern of uptake, increasing stimulus concentration was associated with increased 2-DG uptake in most cases (Johnson and Leon, 2000a). The general increase in activity gave the impression that individual activity foci grew in size, recruiting additional glomeruli at any given threshold value of activity. This pattern is similar to what has been found using other methods of monitoring glomerular activity (Guthrie and Gall, 1995; Rubin and Katz, 1999; Meister and Bonhoeffer, 2001; Wachowiak and Cohen, 2001). The major exception to this rule of increasing activity with increasing odorant concentration involved high concentrations of valeric acid, where 2-DG uptake decreased in all response modules at the highest concentration tested (Johnson and Leon, 2000a). Rats withhold their inspiration at high concentrations of valeric acid, which is an irritant detected by the trigeminal nerve (Alarie, 1973). Decreased sniffing of valeric acid at high vapor concentrations could explain the phenomenon of decreased 2-DG uptake at these high stimulus concentrations. 

When expressed in terms of glomerular layer 2-DG uptake divided by subependymal zone 2-DG uptake, activity was seen to increase with concentration up to 12 parts per million of valeric acid, followed by a decline that may reflect decreased sniffing of the irritating odorant.