Our Results

            Because of time constraints, we were only able to let some sets of plates sit for two days, so we had two sets of comparable data—the mold counts after two days from Quadrats 2, 3, and 4, and the mold counts after three days from Quadrats 1, 2, and 4.  Table 1 shows the number of molds per cubic centimeter of soil counted after being plated for two days in Quadrats 2,3, and 4.  Although it can clearly be seen that the average mold population per cubic centimeter increases as the exposure to Ultraviolet Radiation decreases from Clingwrap to Hard Plastic.  However, when we performed t-tests on the comparable data and discovered absolutely no statistical significance between any of the coverings for any of the data sets.  This means that the null hypothesis is confirmed and the slight increases are just random chance.  For the average number of molds per cubic centimeter of soil in Quadrats 1, 2, and 4 counted after being plated for three days (see Table 2), the mold populations varied randomly.  Additional t-testing showed that there was no statistical significance between the mold populations under the covers either. 

            We then performed a chi-squared test between mold populations in Site 4 during the course of our experiment[1] and the mold populations previous to our experiment from the general biota survey that we had performed.  Achieving a chi-squared value of approximately 252,000,000 in comparison to the expected chi-squared value of 23.685, it was blatantly clear that something in the environment besides the amount of ultraviolet radiation that the soil was receiving changed during the course of our experiment. 

            We then t-tested the overall changes between the days that the samples were taken.  Between Day Two and Day Five, there was t-test value of 2.05 in comparison to the t-alpha value for 90% surety of 1.729 and a t-alpha value for 95% surety of 2.093.  This means that we are at least 90% sure and almost 95% sure that there was a statistical difference in the overall populations between Day Two and Day Five.  Between Day Five and Day Six, there was a t-test value of 2.514 in comparison to the t-alpha value for 95% surety of 2.145*, which means that there is over a 95% surety that there is statistical significance in the overall mold populations between Day Five and Day Six. 

Conclusion:

The statistical significance of the t-test values between overall mold population densities on each day of the experiment gives a clear picture of why there was no statistical significance in population densities between the covers despite the fact that they allowed through different level of ultraviolet radiation.  For the week of the biota survey, the temperature was blisteringly hot with an incredibly high ultraviolet index, averaging 8.23 (NOAA, 2002) and therefore causing the low mold population density.  There was also only an average of 23.33% cloud cover (National Weather Service, 2002).  However, from the time we started our individual experiment, the temperature decreased and the clouds rolled in, substantially decreasing the amount of ultraviolet radiation.  The ultraviolet index was only 6.5 (NOAA, 2002) and there was 80% cloud cover (National Weather Service, 2002) during the first two days of our experiment.  This caused a universal bloom in the mold population density.  The decreased influence of ultraviolet light allowed other factors like nutrients to become important and to begin influencing the mold population density creating seemingly random variations in the mold population density.  Not even the clingwrap samples were getting enough ultraviolet light to decrease their mold population density by any significant amount.  However, as the week progressed, the sun reemerged, with 46.67% cloud cover (National Weather Service, 2002) and the ultraviolet index increased again, to an average of 7.2 (NOAA, 2002).  The reinstatement of ultraviolet light as the dominant influencing factor is evidentially supported by the beginnings of a visible skewed distribution, with clingwrap having the lowest population and hard plastic having the highest population (See graphs 3 and 4).  The positive control similarly supported our hypothesis.  Starting with a high average mold population of 45,000,000 because the dirt was almost completely covered by ivy, by the time we took the second sample, it had significantly dropped with an average of 10,875,000 after being plated for two days and 36,416,667 after being plated for three days.  Despite the cloud cover at that time, there was still significantly more exposure to ultraviolet radiation with bare dirt than with dirt covered completely by English Ivy. 

            This line of reasoning led us to the conclusion that we have strong evidence supporting the hypothesis that the increased influence of ultraviolet light on exposed soil causes a subsequent decrease in the mold population density under that soil.  Although, because of time constraints, we were unable to take additional samples which would have increased the difference between the covers to the extent that there would be statistical difference between the mold population density under each cover.  If the ultraviolet index had been dramatically higher or lower, we believe that the actual experimental results would have been statistically significant between each of the covers.  This would have allowed us to claim conclusive proof for our hypothesis.  This being the case, for future research we would advise running the experiment trying to choose a time when the weather forecasted is either an extremely high ultraviolet index or at least fairly constant amount of cloud cover, and running the experiment for a longer time period.

Certain errors, other than the variations in the ultraviolet index, occurred that might have slightly altered our results.  For the first day on which we took samples, we were unable to perfectly repeat our experiment.  Taking our samples on different days from different quadrats provided replication, but the first day we did not repeat by taking more than one sample from each covering per day.  The second day we tried to correct that error—we took multiple samples from under each covering.  When plating Quadrat 1, Canvas, 10^-1 and 10^-2, the easygel solution was accidentally placed in normal petri dishes instead of easygel dishes, obviously causing it not to harden.  When the mistake was realized, we poured the easygel from the sterile petri dishes into easygel plates.  In addition, the saran wrap came off one of the squares (Quadrat 4, Saran).  We were able to take a sample (# 1) from under the area that was still covered, but our other sample (# 2) came from the area that would have been covered if the saran wrap had not torn.

BACK TO CONTENTS                                                         BACK TO DESCRIPTION OF RESEARCH

FORWARD TO PROCEDURE

Table 1:  Mold Population per Cm³ of Soil

Table 2:  Mold Population per Cm³ of Soil

Graph 1:

Graph 2:

Graph 3:

Graph 4:

Graph 5:

Graph 6: