All other variables will need to be controlled or their effects minimised, in order to ensure that change of the dependent variable is predominantly caused only by change in the independent variable. A control is unsuitable for an ecological investigation however controlling the other variables should make the accuracy sufficient enough to draw a sound conclusion.
The climate will be fairly constant as all the results will be taken within a two-day period very little change will be caused by plants dying/growing. Factors which could change across the land include the pH and fertility, (organic matter and mineral content), of the soil. However, for the purposes of this investigation, I have assumed that they are constant, but this could be the subject of further examination.
If this investigation is successful, further investigations could be conducted into the effects of other abiotic factors on the distribution of these species across such an ecosystem. For example, how changes in the pH, organic matter or mineral content of soil may affect their distribution and abundance.
Equipment:
- A ½ m² quadrat - To mark out the set area in which to take the individual measurements.
- Pencil - To take down the results.
- Data tables - To record the results.
- Calculator - To calculate the running total, running mean and to perform the statistical tests needed to analyse the results.
- Clipboard/file, keep results safe and organised.
Outline Method:
- Use a random numbers* table to decide location of quadrat to take in that area.
-
Use a ½ m2 quadrat, count the abundance and variety of grass in each square of the quadrat (there are 50 in each quadrat).
- Work out a running total and a running mean as results are taken.
- Take 15 samples after initial sample has been carried out which will not be counted but should improve accuracy.
- Try to carry out investigation in both areas on the same day leaving another day in case of problems with the results.
- Use the t-test to deduce whether there are sufficient grounds to be able to reject the null hypothesis or whether it must be accepted.
*Using random numbers will avoid bias and produce a more representative set of results with which to base the conclusion and discussion on.
Where I intend to carry out my study:
Sheltered Woodland Barrow House Derwentwater
Unsheltered managed Grassland Barrow House Derwentwater
Grid Ref: NY268200
Risk assessment:
There are several major safety issues when conducting this investigation. Firstly there is the possibility of buried barbed wire and other dangerous objects; therefore care must be taken whilst moving around on the grassland. Secondly there is the wildlife to contend with, it is best to avoid any possible habitats when able, to protect me and the wildlife from harm. Even though it will be September when the investigation will be carried out, weather reports should be noted in case of sun, wear sunscreen, or heavy rain, bring waterproof coat. Due to the investigation being undertaken in an unknown area to me I should be careful of wandering off and getting lost. The last major safety issue is ‘hostile’ vegetation. These are any plants that may have sharp fronds or be poisonous or an irritant. Care MUST be taken at all times.
Environmental Impact:
Care for the environment in which the investigation is being carried out is also important. The impact of the investigation on the ecosystem should be minimised by avoiding trampling on easily avoidable plants, such as orchids and by not picking anything for identification purposes unless absolutely necessary. Any rabbit warrens should be left alone to avoid disturbing them and a general respect for the environment kept in mind at all times.
Method:
In order to assist with the planning of this investigation, preliminary tests were conducted at Barrow house, Derwentwater woodland and grassland to evaluate methods and investigative techniques for use in the final procedure.
Prior to collecting my data I will have already organised the random sampling on a calculator using RAN# function that produces a random decimal number between 0 and 1 which I will then multiply by 20(the maximum number of steps I will take in one direction) then randomly picking a direction (north, south, east, west) using a random selecting computer program
Randomly select a starting point on the unsheltered land; count steps in direction and place (do not throw for safety reasons) quadrat on the ground, count the variety of plant species in it and note down in results table.
Repeat this test 14 more times in the unsheltered land.
Carry out the same procedure on sheltered land once finished go over all results from both areas and see if there are outliers or if any of the tests could be take again.
Start the statistical analysis on the results
The ‘T’-Test is used to test for a significant difference between the two means of my quantitive data set from the sheltered and exposed shore. I will make use of the student ‘T’-test because it is the best method to deduce whether my results are sufficient to be able to reject the Null Hypothesis or whether it must be accepted.
Preliminary Tests
I carried out my method as a trial before starting my official investigation, I did this to check how the method would work out and whether I would get viable results that I could base strong conclusions on. Instead of carrying out 15 tests at each site I chose to carry out 5 samples at each site, although I did not use the random sampling technique I had prepared for the final investigation because I did not think that the sampling technique needed to be tested in the preliminary tests.
I have included the tables of results that I got for the preliminary tests, once I saw that they followed my hypothesis and there were no anomalous results, so I then moved onto taking my final results and not changing the method because it worked well in the preliminary investigation.
When comparing the preliminary results to the final results you can see that they have a similar trend when it comes to the variety of plant species in grassland and Woodland and the percentage of grass in the grassland is similar.
Results:
Figure 1
Mean: 3.3
Diversity index= 750822/245304=3.06
Standard deviation: 1.46
Grassland
Figure 2
Mean: 9.1
Diversity index=2003640/1560716=1.28
Standard deviation: 1.13
Figure 3
The t test
T= 9.1 – 3.3 5.8
√1.23 + 2.09 = 0.221 = 26.24 (See figure 4 for full calculations)
15 15
The critical value is 2.048
Analysis
As can be seen from the tables and graphs there is a clear difference in the number of species between the unsheltered grassland and sheltered woodland areas. In the woodland area the variety of plant species is significantly higher than in the managed grassland. When the t-test is carried out we can see that the attained value for t of 26.24 is significantly greater than the critical value of 2.048. This means the null hypothesis must be rejected and the alternative hypotheses accepted, and say that there is a considerable difference between the variety of plants in grassland and woodland.
Figure 1 and 2 both show the exact number of each species found in the Grassland and woodland, because moss and grass could not be counted therefore I gave percentages instead, as I thought there was much higher percentage of grass in the grassland than there was in the woodland, also there were more plants in the woodland than in the grassland.
All number of species were double checked on the amount to make sure that the data was as reliable as possible.
Figure 3 is a graph to allow the comparison of both sets of data and easy identification of which area had a higher variety of plant species. The only anomalous reading was site 7 with 7 species in the grassland in that it was significantly larger from the average reading, 3, which is why it is impossible to say that plant species variation will always be lower in managed grassland, but there were some factors that may have caused this anomaly; the main reason being because I was collecting data from random sampling therefore I was following certain directions and number of steps to take and for sample site 7 it was on the edge of the grassland, next to some woodland, consequently some of the species of plants may have encroached on my quadrat, I should have taken another extra sample to compensate for this outlier but I had not realised it during the time.
Discussion and evaluation
The results obtained in this investigation support the hypothesis that there is a higher variety of plant species in the unmanaged woodland than managed grassland.
Several factors would have affected the accuracy of my results. Firstly although the two sites were very similar, in was impossible to ensure they were at exactly the same height as the hill of the woodland sloped gently down towards the grassland area. Even the small difference in height could have affected the results.
The limitations that I faced in my investigations were;
Equipment limitations- The equipment I used to collect my data was not of high technology, the quadrat I used was a bit bent out of shape leaving me to guess the percentage covers.
Human error- To identify the species of plants I took pictures and identified them when I got back but I may have been mistaken about some of the species. Also I took I5 samples in each area but as shown with the anomaly at site 7 in the grassland, maybe 15 wasn’t enough if I had taken maybe 25 then I would have been able to discard the anomaly but I felt that 14 was too little a number despite this there seems to be a definite pattern about which area had the most biodiversity.
On the whole the variables were kept fairly constant and errors from them severely reduced. However, there are many improvements that could be made if the investigation was to be repeated. Firstly, more results should be taken, as although this investigation provided adequate results, there was not much room for error. A larger sample size would provide more accurate and reliable results and encourage my hypothesis.
There are many possible ways to extend the investigation. I could take further measurements such as pH levels, wind speed, light levels and humidity. I could carry out my experiment to see if the results are the same in different areas such as near the sea or in towns and cities where there will be more pollution to see if any of these factors contribute to the variations in my results.
Also the grassland was managed by sheep eating it, in other areas where it is not sheep feeding on the grass, the results may be different because the sheep may prefer eating some of the different plant species and that why there was little or none of them in the centre and why I found them at the edge near the woodland where the sheep may not feed.
Bibliography
1. Comparative Plant Ecology, J. Grime, J.G.Hodgson and R.Hunt; Chapman Hall, London, 1988.
2. Biology 1, Mary Jones & Jennifer Gregory, OCR Cambridge University Press 2000.
3. Biology 2, Mary Jones, Richard Fosbery & Dennis Taylor, OCR Cambridge University Press 2001.
4. Plant Species
Merriam-Webster 2009
15/9/2010
This is a reliable source; it is thought to be one of the most trusted sources of information in the world and used as a link on many official plant identification websites.
5. Kent, M. Advanced biology. Oxford University Press, 2000.
6. Connecticut Botanical Society
2002 Janet Novak
15/9/2010
This site is useful and reliable, I used it to back up and check what was on other websites, increasing its reliability.
7. Salters Nuffield Advanced Biology: A2 Student Book by Angela Hall and Nick Owens
8. Botany.com
Hillclimb Media 1996 - 2008
1/10/2010