I completed a survey and gained 137 sightings of terrapins, 87 in still water and 47 in flowing water. I originally thought I could do Chi goodness of fit but a stats book I have says I cant do that as it is testing the habitat rather than the species. I don't have any other way of testing it other than absence data (61 still and 187 flowing). Can I do something with this data or do I need to give up and find something else?
Hi! The way I understand it right now is you only have a frequency/descriptive data of terrapins in still water (87) and in flowing water (47) [87 + 47 = 134 not 137, please check]. With these specified data alone that deal with a single variable (location of terrapin sighting), I don't think you cannot perform statistical analysis of association (by Chi-square, for instance), let alone "confirm that terrapins prefer still or flowing water." Right now, you have no other variable to associate with location of terrapin sighting (e.g., biological sex, age, adult versus young, time spent in body of water type, etc.).
Please clarify as well on what you mean by the 61 and 187 numbers that you mentioned. What other data have you obtained from each sighting, other than the location where you found them? As it appears right now, unfortunately, I don't think that the data you presented will be insufficient any question with overlap to the one you are asking. Perhaps, you can go back to drawing board with your research supervisor/mentor/adviser to discuss/refine your research question, hypotheses and study design.
Your hypothesis is that flow velocity is an important factor determining the abundance of terrapins. Unfortunately, you are unable to measure flow velocity because it is assumed that still waters have a flow velocity of zero.
First, let's start with some basic questions. Did you obtain your counts from a single sighting or you have a table from different sightings which you have summed up? Also, did you measure any environmental variable that differs between lentic and lotic systems to find out the key components determining the distribution of terrapins? If yes,
Then, you're left with comparing abundance of a species in two different locations experiencing different physicochemical parameters (if you can find a way to measure some water physicochemical parameters related to your work - maybe DO, TDS, transparency, depth, width, discharge or parameters that differ between still waters and lotic systems). This can then be analysed by using stats which relate environmental variables with abundance of a species.
Thanks
@Aedrian Abrilla Thank you for responding, yes I must have made a mistake there 134 total. Unfortunately, I couldn't get the people I asked to give me velocity or similar as much of the sighting data is historical. The other data set I have is absence data with a different set of people who have never seen a terrapin I asked them to put down their local waterway or body, these are the other numbers. Unfortunately, I have gone too far now to go back so I will have to analyse some other data I collected on the people who gave me sightings. The only other data I have on the terrapin sightings is if it was fresh or brackish water, however, I only had 3 brackish sightings. This was what I originally wanted to do which was compare the groups and see where there were more sightings.
Abiodun Matthew Adedapo My study was a survey from which I got data on the people surveying and sightings from some of them (116/252). Some of them added more sightings which is why my total for lotic and lentic water is 134. As I said above I also asked if the water was brackish or fresh, however, my data was unbalanced with only 3 brackish sightings. I also got absence data from the remaining 136 people. It is due at the end of the week so I will have to look at something else.
No, you can definitely use Chi Squared to tests expected numbers of terrapins in different water conditions
I would use occupancy modeling which you can read about here: https://www.nps.gov/olym/learn/nature/upload/OccupancyModelFactSheet.pdf.
Julie Grinstead Thank you so much this looks like just what I am looking for I will give this a go and let you know how it goes!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics