Discussion

The results showed a wide range of variability, both between replicates from the same puree, as well as between other purees made by separate groups from the same sample of fruit.  For example, one groups replicates for the 10/24 light autumn berries yielded a range of values from 107.77 mg/kg to 171 mg/kg with a standard deviation of 33.39.  Similarly, one group’s average lycopene content for watermelon puree was 4.75, while the other two groups showed comparable average lycopene contents of 42.97 mg/kg and 44.43 mg/kg for tissue from the same melon. There are a number of factors thought to be responsible for these widely variable results.  First, the sorting of the berries by the landowner was performed after the fruit had been frozen; thus, the effectiveness of sorting the berries based upon color may not have been adequate for grouping berries after freezing.  In future trials, if berry color is to be used as a distinguishing factor, samples should be sorted prior to freezing.  In addition, freezing of the fruit before extracting and quantifying lycopene may have affected the experimental lycopene content.  Fish and Davis (2003) reported that in watermelon, there was a significant difference between lycopene contents of fresh and previously frozen tissue. This suggests that results may be more accurate if fresh tissue were used instead of frozen tissue.  They also found that lycopene in samples which were stored at -20˚C was less stable than in samples stored at -80˚C, the standard freezing temperature for scientific materials.  Since our berries were stored at -20˚C, the actual lycopene content may have originally been much higher than our results showed.  Additionally, it has been reported that lycopene deteriorates faster in frozen watermelon stored as chunks compared to watermelon stored as puree (Fish and Davis, 2003).  Since our berries were initially frozen whole, it seems reasonable to believe that lycopene content in the whole fruits declined similarly to that in the chunks of watermelon.  Setting aside any losses in lycopene that occurred as a result of sorting and storage of the berries, there remain other issues with our experiment that could explain our inconsistent results.  Although care was taken to achieve a homogenous puree, some vials still contained chunks of material in the bottom of the vial, indicating tissue was not completely ground. 

            Despite the variability in our numbers, some positive conclusions can be drawn.  First, two groups obtained results for the watermelon comparable to those published by (Fish, et al., 2002).  This indicated that our extraction and quantification method was reliable, at least for watermelon.  For tomato paste, the results were not as consistent, with two groups quantifying very low lycopene contents, while one group had much higher values.  These data could indicate our extraction method for the tomato paste was not as effective, or possibly that the method does not work for tomato paste.  Consequently, it seems best to compare the autumn berry results using the watermelon results as a control rather than the tomato paste. 

            The autumn berry samples showed a range of lycopene contents, both from those picked on the same day and those picked on different days.  The highest lycopene content was found for the light berries picked on 10/24/03.  However, the dark berries picked that same day consistently showed a very small lycopene content.  Interestingly, the berries picked on 9/21/03 showed a lycopene content in between the two previous samples.  Thus, it seems that in this particular trial, the light berries picked later in the season had the most lycopene.