A completely accurate biomass sample would involve collecting 100 percent of the organic matter in the 50cm2 sampling area. However, some of this organic matter that should have been collected for the biomass was inevitably left behind due to the difficulty involved in fully removing grass and other rooted organisms from the ground. This would have resulted in a systematic error due to the fact that this would decrease the biomass of the producer trophic level that was calculated using the biomass collected from the 50cm2 area. The amount of variation that this would cause is unknown since we cannot determine just how much biomass was not collected in retrospect. In order to improve upon this source of error, more time should be spent carefully collecting all of the biomass in order to retrieve as close as possible to 100 percent of the biomass in the area.
Bugs were also exceedingly difficult to catch using bug nets in the forest due to the amount of bushes, the lack of room to swing the net and the number of plants with thorns that were in close proximity, limiting the ability to swing the net without damaging the net. This systematic error causes the numbers of bugs that were caught to be lower than the actual amount of bugs present in the sampling area. This causes the ecological pyramids to be less representative of the ecosystem due to the fact that they contain fewer primary and secondary consumers. The amount of variation that is caused by this source of error is unknown because the number of bugs that were not caught in the sampling area cannot be determined. This source of error is unable to be significantly reduced since it is caused by the nature of the forest itself, meaning that this source of error would only be less of a factor if bugs were being caught in another ecosystem altogether. The best way to improve the chances of catching more bugs, however, is to use a larger net with smaller holes than many of the nets that were being used.
In order to determine the number of grasses for the numbers pyramid, a sample of thirty blades of grass were counted and their mass was then calculated. The total biomass of the full sample of grass was then divided by the mass of thirty blades of grass in order to give an approximate value for the number of grasses in the 50cm2 area. However, the primary fault in this method of counting is that the masses of each individual blade of grass have a very large degree of variation. Because the thirty randomly selected grasses could be either heavier or lighter than the average mass of all of the grass, the error is random. Variability of mass of grass used when calculating numbers for grasses. In order to reduce the impact of the variation, more grass should be individually counted. 100 grasses would provide a great deal more reliability while remaining to be not overtly time consuming.
When calculating the biomass of trees in the 50cm2 sampling area, certain assumptions were made in order to make this possible. The main assumption that was made in order to calculate the tree’s biomass is that the tree is a cylinder shape without any branches or leaves. This assumption is made in order to make it possible to calculate the biomass of the tree using the tree’s circumference, height and density. However, because the branches and leaves are not included in the biomass of the tree, the calculated biomass is less than what the mass should be, causing a systematic error. Furthermore, the source of error causes the amount of energy in the energy pyramid to be lower than it should be for the producer trophic level. The amount of biomass and energy that is lost when the branches and leaves are not counted is undeterminable since the mass of the branches and leaves varies from one tree to another. The only way to improve on this source of error would be to measure the dimensions of the branches and calculate their mass, however this would be very inefficient.
Some groups were unaware that tree masses were supposed to be calculated, and/or unaware of how to calculate the mass of trees in the transect. This systematic error causes the biomass of producers and the energy of the producers to decrease on the respective pyramids. It is not possible to determine the true extent of the decrease in biomass and energy unless the biomasses and energy of the trees is calculated now. However this is likely no longer possible unless the groups that did not take biomass or energy for the trees know which trees were included in their sample area. The only way to fix this source of error entirely is to ensure that all groups are aware of how to calculate the mass of a tree before going to Nose Hill.
Some groups were not aware that the numbers of grasses were supposed to be counted, or they were unsure of how to go about estimating the total number of grasses. This is a systematic error that immensely altered the numbers for the producers trophic level. Because the grasses often accounted for thousands of producers, in comparison to numbers of below fifty when grasses aren’t included, the percentage of error would be very high, likely around 100 percent. Because of this extremely significant error in certain group’s numbers for producers were removed from the calculations of averages used from which the pyramids were made. This reduced the impact of the error, however it also reduced the number of replicates used for the numbers pyramid, reducing the accuracy of the average. The best way to fix this error is to make it clear that all grasses must be counted in the biomass sample and to tell all people how to go about estimating the number.
While at Nose Hill soil samples were taken and brought back for testing. The pH, nitrate, phosphate, calcium and ammonia levels were all tested and recorded. However different groups tested their soil samples at varying times. Some groups tested the soil soon after it was removed from Nose Hill, creating accurate results as the soil would not have had too much time where the levels of pH, nitrate, phosphate, calcium and ammonia could have changed. However other groups tested their soil samples two weeks or even a month after the samples were taken. This systematic error caused great variation between the samples. It is impossible to calculate how great of an error it is because we have not measured how much the levels change over time, and we did not record when each group tested their samples. To prevent this from happening, the soil samples could have been tested within one or two days from when they were taken, this would reduce the variation and make the results more accurate.
To create a soil profile a soil corer was used. This tool however was not very affective at obtaining a good sample for drawing the soil profile. This was due mostly to the conditions of the soil, which were dry, hard and full of roots. These conditions made it hard to extract the soil and when the soil had been extracted from the ground using the corer the sample did not remain fixed making it hard to create an accurate representation of the layers of soil. This was a random error as there were many variables such as how many roots were in the spot where the corer was used, and how hard and dry the ground was in a particular spot. It is not possible to calculate the percentage of error that the corer caused because there is no way to determine how different our soil profiles are compared to how the soil looks in the ground. To increase accuracy we could have used a sharper corer that was easier to twist, this would allow us to extract a sample more easily.
© S. Karim, S. Kushwaha
Some errors that we may have encountered include the chemical tests for both water and soil. The sooner you do the test after collecting your samples the more accurate the results. The longer you what the amount of nitrate, ammonia and calcium could have changed, but testing times varied form group to group therefore creating results that are different from each other. To eliminate this error in the future groups could test both water and soil as soon as possible to get the most accurate result.
(c) Poonam R. Brennan S. 2009
Sources of Error
The most significant source of error in this application of Hardy-Weinberg equilibrium is that we can not confirm whether the dark-brown head or light red-brown head is an inherited phenotype. Given the possibility that variation in head colour is due to environmental conditions such as exposure to sunlight, Hardy-Weinberg equilibrium is irrelevant in that there is no genotype responsible for it.
Another significant source of error in the Hardy-Weinberg application is the limited size and area by which our wood ant sample was collected. Firstly, a sample of 12 wood ants may have been too small to accurately represent the distribution of dark-brown and light red-brown heads amongst the entire population of wood ants in the Nose Hill forest. The small size of the sample makes it more likely that the frequencies of each collected genotype are coincidental, where the higher number of dark-brown-headed wood ants may have been simply due to chance. Secondly, wood ants were gathered from a very limited area within the Nose Hill forest, which may not be representative of the entire wood ant population due to variation in distributions. We assumed that the distribution and frequencies of the dark-brown and red-brown-headed wood ants were consistent throughout all areas of the forest which may not necessarily be true. These errors have a significant and random effect on our Hardy-Weinberg analysis as we cannot determine whether our sample has allele frequencies higher or lower than that of the overall forest population. We can reduce the impact of this source of error by either collecting more wood ants or pooling samples with other groups, over a larger area of the forest microecosystem.
©Isaac Lin, uploaded October 28, 2009 at 8:16 PM
Sources of Error
The data contained in Figure I and the process in which we gathered and measured plant biomass contains sources of error which render this analysis and inference relatively unreliable. Firstly, no replicates of gathering plant biomass from transects of a certain soil temperature were carried out, thus the biomass data we did gather is not necessarily representative of average biomasses in
A second source of error exists in that the biomass extracted from each transect may have been foreign material, or that biomass that originally belonged in the transect was removed either through biotic or abiotic factors. Certain factors such as wind, human movement or animal movement may have moved some biomass, mainly leaf detritus and branches, in and out of the transect. This source of error would have had a random effect on the collected data as the amount of foreign or native material that is moved in or out of the transect by the external environment cannot be controlled or quantified. Furthermore, this source of error significantly impacts the data gathering as due to lack of replicates and averaging, any loss or excess of biomass immediately affects the analyzing of biomass as a function of soil temperature. The impact of this source of error can be reduced in future experiments by altering the procedure to not include detritus material since some of it may be foreign. Thus, any detritus leaves or branches should be removed from the transect with a rake before removing producers still fixed to the soil.
Another identified source of error involves detritus leaf material left from the Quaking Aspen trees in certain transects. Although special consideration was taken to not include the Quaking Aspen’s biomass as part of the measurements, some leaf material and branches from the tree may have been included in the gathered biomass. This would be considered foreign addition of biomass from a plant we chose to exclude and thus, it has a systematic source of error in that transects with Quaking Aspens will have some excess biomass from the tree’s fallen leaves and branches. This source of error varies is significance depending on whether the excess biomass from the Quaking Aspens consisted of branches or leaves. If only leaves, their numbers and individual biomasses are relatively small and can be considered to be negligible amongst the transect’s total biomass. If branches are present however, they have large individual biomasses and significantly impact the transect’s total biomass. Again, the impact of this source of error can be reduced by removing detritus leaves and branches from all the transects as to avoid the issue of foreign material being included as part of the biomass.
©Isaac Lin, uploaded October 28, 2009 at 8:49 PM
A source of error exemplified during the process of making the numbers pyramid is how each plant and animal was counted. It was extremely difficult to gather an accurate data for the number of producers, primary consumers, and secondary consumers because of a wide range of implications. One was that they were hard to identify in the murky waters, so it could’ve been easy to miscount a secondary consumer for a primary consumers. It would also be confusing for some organism because they would fit in both categories of primary and secondary consumers. Confusion of pond organisms caused a huge impact on the precision of the data. Not only were they difficult to identify and categorized, many of the organisms hid in the soil and lived at the bottom of the pond or attached to pond plants. They also moved around at quite an alarming speed that it was difficult to keep track of them and most of the time, even see them. Producers proved to be an easy count. However, the plants were tangled amongst each other and were counted in such a way that “one” plant was a huge clump of algae that would’ve definitely fed more than any one of the aquatic creatures alone. Thus is the problem with number pyramids that they do not account for size. The miscounting of the organisms at each level of the numbers pyramids is a random error and caused a major effect on the outcome of the results. The only way to improve and reduce this error would be if we “took out” the entire transect along with the soil and water together, without disturbing the natural environment of the transect, which would be nearly impossible to do.
©Mei Chen, uploaded October 30, 2009 at 6:03 PM
Sources of Error
1. The chemical test kits used to determine chemical concentrations in the pond gave highly imprecise results. Different concentrations were indicated by different shades of a colour, and concentrations were given in increments of 1.0 mg/L. Thus, for most of the chemical tests, concentrations could only be reported to the nearest 1.0 mg/L, compromising the precision of the data. This error was more pronounced in the calcium test, which gave results which were rounded to the nearest 10.0 mg/L. Thus, a sample with a concentration of 12.0 mg/L would be indicated by the test as having a concentration of 20.0 mg/L for a low 60% accuracy. This is a random error because we cannot determine whether we overestimated or underestimated chemical concentrations. It can be reduced by performing more replicates so that any erroneous data will even out given more trials.
2. The concentrations of the chemicals were determined by adding indicators to a sample and comparing the consequent colour to a colour chart. It was often hard to determine which colour on the chart was the most similar to the colour of the sample. Thus, the concentrations of chemicals may have been over- or under-estimated due to the ambiguity in the colour comparison. This is a random error because we do not know whether or how the experimenters interpreted and compared the colour of the sample to the colours on the chart. It can be reduced by performing more replicates so that any erroneous data will be evened out given more trials.
3. Pond samples for chemical tests were all taken from the shallow area at the edge of the pool, which was the most accessible location. Thus the samples taken are only representative of a very limited area of the pond. For example, substances might have sunk down to the deepest area of the pond, near its middle. These substances cannot then be accounted for in chemical tests. This is a random error because we do not know quantifiably how our biased sampling technique will account for or discount for the chemical composition in the unsampled areas of the pond. We can reduce the error by preparing more equipment which would enable us to take samples from the middle of the pond (for example, getting a kayak).
4. Similar to above, pond organisms were all taken from the shallow, accessible areas of the pond. Thus, we may have missed a few organisms which live deeper and towards the middle of the pond. As above, it is a random error which can be reduced by preparing equipment which would allow us to use the middle of the pond as a sampling area as well as the sides.
©Justine Zhang, uploaded October 31, 2009 at 9:33 AM