Discussion

This experiment was performed to observe the types of growing media and their effects on carbohydrate and protein productions of the Wisconsin fast plant. In this experiment, Brassica rapa was studied using three treatments and a control group, each of which had four fast plants. The control group consisted of fast plants growing in normal potting soil, and the three treatments included plants that were grown hydroponically, with Miracle Gro®, or with Meijer ä Brand Malathion Insect Control Concentrate. All of the tested plants went through carbohydrate and photosynthesis tests, as well as a statistical analysis based on plant height. We omitted the Bradford (Protein) Assay because of a shortage of biomass; the majority of the plants’ protein is stored in the seeds produced by the plant, and the Brassica rapa plants that were tested were not mature enough to produce seeds in time that we had to perform this lab, and leaves tend to have very low amounts protein present (Campbell, 2002). An originally planned pigment analysis using paper chromatography was omitted due to the fact that we did not have enough plant material to do both the Hill Reaction and the pigment analysis, and it was decided that a quantitative test like the Hill Reaction would be more valuable when looking for positive and negative effects of the treatments.

Comparing the treatments with the control group, it was hypothesized that the differing growing media would yield plants with different growth, carbohydrate composition, and photosynthetic activity. It was predicted that the Miracle Gro® treatment would have a more positive net growth and productivity because Miracle Gro® contains nitrogen, which is said to increase plant productivity with the help of nitrogen fixation (Freeman, 2002). The pesticide treatment was predicted to have a more negative effect of growth and productivity on fast plants due to pesticide residue that can enter the plant through the stomata (if the pesticide was in vapor form) or through the roots and xylem (if the pesticide was in the soil). The residue from past studies contributed to the contamination of soil that plants grow in (Waliszewski et al., 2004). Because of a large failure rate of hydroponically-grown plants (even by planting experts) and the requirement of high degree management skills with hydroponic plants, it was predicted that the hydroponic treatment in this experiment would yield less growth and productivity than the control (Anonymous-2, 2005).

Before analyzing the results of the various assays, it should be mentioned that the hydroponically-grown plants did not survive past germination, so no tests could be conducted on this group, and they will not be discussed in the conclusions of each assay. While unexpected, this result does reaffirm that growing is a difficult process, one often met with failure (Jensen, 2005). The most plausible explanation is that the sprouting seeds did not get enough water to survive. Our planting medium of perlite and vermiculite held water very poorly. This was understood in advance, and was the reason that a nutrient reservoir was included in the system with a string for a wick that brought water and nutrients into the growth medium. However, our string did not start to wick the nutrient solution to the growth medium until approximately 5-6 days had passed. Until then very little water stayed in the growth medium, when watered was added it would run through the growth medium into the nutrient reservoir. The string may have had some kind of protective coating on it that did not wear off until the string had been soaked for some time. In future endeavors into hydroponics, the wick should probably be soaked for several days before setting up the hydroponic system, and the growth medium used should be better at holding water than vermiculite and perlite.

In the first set of experiments, five different carbohydrate tests were performed on the leaves of the experimental Brassicarapa. The first was Benedict’s Test, which tested to see if there were free-forming aldehyde or ketone groups. It was predicted that our leaves would test positive due to the fact that sugars like glucose and fructose are products of photosynthesis (Freeman, 2002), and these sugars contain free aldehyde and ketone groups. The prediction was incorrect, since all of our plants tested negative. This could have resulted from the Benedict’s reagent not being sensitive enough to react with the low amount of sugar in the plants. Also, carbon-5 of fructose and carbon-1 of glucose can link together to form sucrose (which stores the glucose in that form), which hides the free aldehyde and ketone groups from the test (Krha, et al., 2005). If the plant stored all of its sugars in the form sucrose, then this would explain the negative result.

The next test that was conducted was Barfoed’s Test, which distinguished monosaccharides from di-/poly-saccharides. For all treatments, because of the predicted large amount of glucose produced by photosynthesis present in the leaves, it was predicted that the leaves would test positive for monosaccharides, but because it would test positive for monosaccharides, the presence of di-/poly-saccharides could not be determined based on the timing of the test, even though sucrose is present in the leaves during photosynthesis (Freeman, 2002). It turned out that there was no reaction from Barfoed’s solution for each of the plants. However, there was also no reaction from Barfoed’s solution for the positive control sugars. We tested the provided Barfoed’s solution on galactose, fructose and glucose (all monosaccharides), which should have tested positive. Each of the monosaccharides tested negative when reacted with the Barfoed’s reagent. Attempts were made on two separate dates, but each time no red precipitate was formed when the reagent was reacted with various monosaccharides. This means that the solution that was used was incorrectly made, contaminated, or too old to use. So the test results from Barfoed’s test could not be used as evidence to test the hypothesis. However, since Barfoed’s Reagent is identical to Benedict’s except for pH and the time reacted, it is reasonable to assume that a negative result would also be obtained with a working Barfoed’s solution.

Selivanoff’s Test, which distinguishes aldoses from ketoses, was performed next. For all treatments, it was predicted that the Brassica rapa leaves would test positive for ketoses. However, we also predicted that there would be aldoses in the leaves. Both fructose (a ketose) and glucose (an aldose) are made in photosynthesis, eventually forming sucrose (Freeman, 2002). Because of the timing of Selivanoff’s test, though, if the leaves tested positive for ketoses, there would be no way to prove the presence of aldoses through this test. This was not an issue for our tests, however, because results showed that all of the plants tested positive for aldoses. In theory, the leaves should have tested positive for ketoses if sucrose was present. The reason may be that the sugars in the plant may in such a low concentration that it takes quite a bit of time before enough sugar has reacted with the reagent to produce a visible result.

Bial’s test was the next carbohydrate test conducted, which determines the presence of furanose rings. For all treatments, because glucose (a pyranose) and fructose (a furanose) are both produced in photosynthesis (Freeman, 2002), we predicted that we would see a color change since there is fructose being produced. However, there was no color change in our results, which is an indication that pyranose sugars were present instead of furanose sugars. Again, the most plausible explanation is that there simply are not enough sugars present in plant leaves to react with the reagent, or that it is in a form that is not detectable by this test, such as a disaccharide, starch, or cellulose.

The last of the five carbohydrate tests performed was the Iodine test, which tests for the presence of starch. It is a well-known fact that plants store glucose in the form of starch (Freeman, 2002), so it was predicted that the plants would test positive for starch. However, our results showed the opposite: all of the plants tested negative for starch. A possible reason is that while plants do store sugar as starch, they usually don’t store it has starch in the leaves, which is what the test was performed on. Cellulose makes up a large portion of a leaf ( Campbell, 2002), and this most likely is not detected by Iodine.

An absorption spectrum was used to test which wavelengths of light were absorbed by the Brassica rapa leaves, it was predicted that the leaves of each treatment would absorb all the colors of the color spectrum at relatively high absorbencies except for green and yellow wavelengths of light. This was predicted because the colors of objects that people see is the color that is reflected off of the objects’ surfaces and not absorbed (Anonymous-5, 2005). Another reason is that plants tend to absorb 400-450 nm wavelengths (blue light) and 650-700 nm (red light) more than other wavelengths light (Freeman, 2002). The data supported our prediction, all of the treatments had their highest absorption peak between 400 and 430 nm, and the second highest peak occurred at approximately 675 nm. However, the absorption peaks for the pesticide treatment were much less defined than the control and Miracle Gro® treatments. This result may be another indication that the pesticide treated plants have a lowered amount of chlorophyll in their leaves. With lowered amounts of chlorophyll, the plant’s ability to absorb red and blue wavelengths would be diminished, which is what seems to be the case here, which is evidence supporting our hypothesis that the lowered nitrogen to phosphorus ratio induced by the pesticide would negatively affect chlorophyll production. Little difference is observed in the definition of the absorption peaks between the Miracle Gro® group and control group, implying that both these groups absorb light similarly.

The Action Spectrum was used to test which treatment of Brassica rapa produces the most photosynthesis. We predicted that the Miracle Gro® treatment would produce the most photosynthesis, and therefore have the lowest absorbance reading. It is known that Miracle Gro® increases growth in a shorter time period, but growth is a result from the photosynthesis that takes place within each plant. The more photosynthesis that a plant accomplishes means that the plants will go through more growth (Freeman, 2002). It is also known that a high nitrogen to phosphorus ratio increases the amount of chlorophyll production (Carpenter et al, 1999). We predicted that the pesticide treated plants would have the least photosynthesis, since the pesticide we used is phosphate-based, which would lower the nitrogen to phosphorus ratio.

We were therefore quite surprised to find that the absorbance taken after the reaction was run was lowest in the pesticide plants, indicating that they had done the most photosynthesis. The Miracle Gro® plants had a lower absorbance than the control group, which was expected, but it was no where near as low as the reading from the pesticide treatment. We doubted that the pesticides were undergoing more photosynthesis; it didn’t make sense with any of the research we had done. An examination of the cuvette containing the reacted solution from the pesticide treatment provided a possible answer. The solution was much lighter in comparison to the reacted solutions from other plant groups. Since we were careful to use the same amount of plant material for each test, we knew that it couldn’t be that there was simply less material there. We concluded that there must be less chlorophyll present in the leaves of the plants treated with pesticide. This would explain the lower absorbance with the pesticide treatment; it was because there was physically less pigment floating in solution, and so more light was transmitted. If this is the case, then it is support for the prediction that the increased phosphorus from the pesticide would cause less chlorophyll to be produced. It also brings up a dilemma with the Hill reaction. One cannot reliably use it to compare photosynthesis of different plants by following the instructions verbatim from the lab manual. Something must be added to make sure that the difference in photosynthesis observed is not due to less material present in the solution. Perhaps if the plant solutions were all made to have the same transmittance before running the Hill reaction, then that would eliminate that variable.

For our independent lab, a quantitative sugar test on the different treatments and control group was conducted. During this lab, the intensity of light was measured on the sugar samples from Benedict’s Test. The theory behind it was that a smaller absorbed wavelength signifies a larger presence of sugars in plants. Benedict’s test turns red in the presence of free aldehyde and ketone groups, and the more groups that are present, the redder the solution will turn (Anonymous-4, 2005). A redder solution will absorb bluer colors of light, which have more energy and smaller wavelengths (Anonymous-5, 2005). It was predicted that the treatment with Miracle Gro® would have the smallest absorbed wavelength because more sugars were predicted to be present in the treatment than the amount of sugars present in the control group. Since Miracle Gro® was predicted to have the most photosynthesis, and therefore the most sugars, it was predicted to have a reacted reagent that absorbed light in the shortest wavelengths. The control’s reagent would be the next shortest wavelength, and the pesticide’s reagent was predicted to absorb at the longest wavelength. However, since our plants did not react with Benedict’s solution under normal conditions, modifications had to be made to the procedure. We decided to try and add amylase when grinding the plants for the Benedict’s test, since amylase is known to break down polysaccharides into smaller chains and eventually down to monosaccharides (Freeman, 2002), so it was hoped that this would free more carbonyl groups to react with the Benedict’s solution. As hoped, a color change was observed when the amylase treated plant solutions were reacted with Benedict’s reaction. When the absorption spectra of the reacted solutions were measured, no clearly defined peaks were observed among any of the treatments, or the positive controls, so unfortunately peaks could not be used to quantify the sugar test. However, when these spectra are compared with the spectrum of a Benedict’s solution run with distilled water, certain trends become apparent. The distilled water spectrum has very low absorbance at shorter wavelengths and very high absorbance at longer wavelengths. Comparing this to the reagent reacted with 1% fructose, which has a roughly level absorption spectrum. This means that the solution absorbs more blue light and less red light than the negative control. The 0.2% solution has a curve that is intermediate between the two extremes. Therefore, our assay is usable as a quantifying sugar test by analyzing the how much the absorbance curve has shifted away from the curve of the negative control curve. However, it is not very sensitive, since the actual plant treatments all look more or less the same in regards to their spectra.

For a period of twenty days in the experiment, we measured the height of each plant at various times. We predicted that the Miracle Gro® plants would grow taller than the control group, which would grow taller than the pesticide-treated plants. This is because the extra nitrogen in the Miracle Gro® would cause the bacteria in the soil to accomplish more nitrogen fixation, which would cause the plants to grow taller (Freeman, 2002). It is also because pesticides can contaminate the soil and plants as mentioned in the beginning of the discussion, resulting in shorter growth (Waliszewski et al., 2004). It turned out that our prediction was correct, since the average growth was 7.3 cm for the control group, 9.6 cm for the Miracle Gro® treatment, and 6.0 cm for the pesticide treatment. We decided to do an ANOVA statistical analysis of all the plant heights to see if there was any statistical significant difference between them. The p-value calculated was 0.115, a value greater than 0.05, which means that there is no statistical significance in height difference between each of the different height treatments (Anonymous-1, 2005). A reasonable objection to our conclusion can be drawn from the fact that our statistical analysis omitted the plants that did not grow, therefore skewing the results which could have dramatically changed the p-value.

Throughout the experiment, there have been many places where errors could have occurred. One place was in measuring the height of each plant, in which some plants were growing upward while others were leaning, which could have affected the actual height measurement. Another place in which error could have occurred was the watering of the plants, in which the water was not measured out before it was added to the plants. If the water was distributed evenly, then our data might have been different from the data that we collected here. Also, in using the pipettes, a little bit of the liquid that was sucked into the pipette could have leaked out of the tip, resulting in a different amount of solution being used for each sugar test. This could have changed results from the sugar tests if this problem was fixed. Also, when measuring absorptions for the independent assay, the red precipitate would begin to settle, causing lower and lower readings, so the cuvettes had to be shaken periodically. This could have led to false low readings while taking the absorbance spectra.

It was concluded that our main hypothesis was supported by our data, in which the varied growing media would affect carbohydrate and pigment production. Our hypothesis that the increased nitrogen to phosphorus ratio provided by the Miracle Gro® could not be supported, due to the fact that there were no significant differences between the Miracle Gro® and control group in any of the tests. However, the pesticide group had large differences in the results of the absorbance and Hill reactions, so we support our hypothesis that the poor nitrogen to phosphorus ratio had a negative effect on sugar and chlorophyll production. No conclusions could be drawn in regards to the hydroponic group, as they never grew large enough to run tests on.

Figure/Table

Table 4. The statistical results of the ANOVA test performed on the heights of all the treated plants. The p-value of the ANOVA test was 0.115, showing that there is no statistical significance between the height of each plant.

ANOVA: Results

The results of the ANOVA statistical test performed at 19:51 on 26-FEB-2005

The probability of this result, assuming the null hypothesis, is 0.115

Miracle Gro® treatment: Number of items= 4
8.30 8.50 10.2 11.3

Mean = 9.57
95% confidence interval for Mean: 7.231 thru 11.92
Standard Deviation = 1.43
Hi = 11.3 Low = 8.30
Median = 9.35
Average Absolute Deviation from Median = 1.17

Pesticide treatment: Number of items= 3
3.60 7.00 7.30

Mean = 5.97
95% confidence interval for Mean: 3.260 thru 8.674
Standard Deviation = 2.06
Hi = 7.30 Low = 3.60
Median = 7.00
Average Absolute Deviation from Median = 1.23

Control Group: Number of items= 3
4.50 7.80 9.50

Spring 2005 Team Gandalf
Last updated on March 3, 2005
Web master: tomaszyc@msu.edu