



| It is very important
not to over-interpret any information that you may have collected. Unless you
are absolutely sure that you have counted every single individual or every
single species (depending on the question you are asking) in your
study area, you will not and cannot know how many there are. Remember this.
It is very important. If you have collected numerous, representative samples
from different sites within your area, you will have a good estimate of how
many there are but you will still not know the true number. Information
about numbers is usually presented as a measure of density (number per unit
area) or the percentage cover (the amount of space they occupy). There will
always be some error in how well your estimate measures the true abundance
or cover, no matter how well you sample. This is called sampling error. Any
difference between two estimates of abundance, for example the number of gum
trees in two bushland parks, might therefore be due to sampling error (that
is, how well your samples measure the true number of trees) or to the fact
that one park really does have less trees than the other. Unless differences
really are clear-cut, e.g. there are only a few trees in one park but
apparently hundreds in the other, statistical tests are necessary to tell
whether different estimates of abundance are likely to show real differences
between different places (or different times). These tests take into account
the number of samples used to estimate the abundances, the variability among
them and the difference between the different sets of samples, e.g.
the different areas. Similarly, you might see what appear to be strong associations between species or between species and patches of habitat. There may seem to be more blue wrens in areas where there are few ravens; more thistles in patches where there are lots of blackberry bushes than where there are few blackberry bushes. Sea anemones may appear to be more common in rock pools than in deep crevices. Dandelions may appear more common next to roadways than in forests. Again unless these relationships are extremely obvious, you need statistical tests to indicate whether they represent real associations. Otherwise you are just guessing and anyone's guess is as good as anyone else's. Guessing is not science and not ecology. So there are problems with interpreting whether any patterns in abundances of different animals and plants are indeed real, or whether they appear because of the ways that you have measured them. More detail about measuring and interpreting these patterns is given in Appendix 3. The next question that people are most interested in is Why? Why are animals found here and not there? Here the scientific basis of many apparently valid "ecological" studies founders - even if the patterns themselves are correct. There are a number of ways of approaching such questions, but only one is really useful so let's deal with the others first. You can guess. That is what some people do, but it is dishonest and wastes all the effort you put into collecting the information in the first place. Second, you can ask someone else and they can guess? This doesn't help much. You can ask "an expert" and they can tell you the answer - a common approach. This is often the same as guessing unless they really do know what they are talking about. An honest "expert" may tell you that people do not know the answer. Do not dismiss such advice - there is a lot more that is not known about animals and plants in Australia than what is known. Finally, he or she may be able to tell about some field experiments that have been done to try to answer the same or similar questions elsewhere. These are really the only way to investigate interactions between different species and their habitats and, thereby, even to begin to answer such complex questions about why something is found where it is. If these experiments have been done a number times in different places, you may feel justified in thinking that the same processes might apply to your study site. Unfortunately, such field experiments are not common, except for a few habitats such as intertidal rocky shores in New South Wales. These have been extensively studied using field experiments and there is a wealth of data available about them in the scientific literature. This cannot be said for most other habitats. Unfortunately, these studies frequently show that similar patterns can be caused by different processes from place to place. Therefore an experiment done once in one place may not necessarily tell you anything much about your own study site. This is not the place to go into much detail about the need for field experiments to interpret patterns of different species in nature. A brief outline about the way in which science in general and ecology in particular is done to answer such questions is given in Appendix 3. Finally, a last note about the advantages of networking.
Networking will not, however, work unless you
GOOD LUCK! |
© Centre for Research on Ecological Impacts of Coastal Cities
email: rgraham@pacific.net.au