If I sample animal species diversity using a transect sampling design (so the traps were set in a sequence, forming a straight line towards a direction), could I say that this transect is a representative sample of a nested sampling?
Technically, I can analyze the sampling result of a transect as if it is of a nested patches, but as it is a line, I do not know if I can assume it represent a patch. Does anyone knows if there is a statistical assessment on this kind of analysis?
A nested design comprises subplots within a larger plot. A transect is not a nested design unless it is a wide strip with actual subplots. It sounds as if your design comprises simply a transect with traps.
See: Using standardized sampling designs from population ecology to assess biodiversity patterns of therophyte vegetation across scales, Journal of Biogeography (2004) 31, 363–377.
I hope this helps :-)
A couple issues need to be clear in your mind: 1. Is the 'transect' your sampling unit ir the traps
You need to be clear on what is your sampling unit, the 'transect' or the traps. Also be clear on whether your intetest is on alpha or beta diversity. Finally, endure that your sampling design avoids the problem of pseudoreplication.
I agree with Agustine, it is necessary to review the replics within the transect, please review Hurlbert, 1984, the Pseudoreplication
Depending upon the number of your transects laid, and with respect to your total area covered, your dataset can be considered as representative or not. Thus, you should assess this yourself.
As to whether your transect sampling can be presumed to be a nested design, I fully agree with @Andew's view that the description of your transect sampling is not a nested design. It is not clear in your question why you want to consider nested design/approach in your analysis. Diversity metrics such as alpha or beta or modeling options can be considered for your statistical analysis if that can answer your research question(s). Autocorrelation is also important to check in your data. All the best in your analysis.
Dear Sabhrina,
In a nested design the levels of one factor are nested within or hierarchically subsumed under levels of another factor in some way… In theory, you can organize transect in this way but, frankly, I never heard of transects used in nested design. Usually, nested designs are based on randomly distributed plots within distinct areas. If you have a special reason to consider your sampling desing nested, you should explain this explicitly.
Thank you for all your answers. The sampling units is the trapping points, not the transect. I was wondering if analyzing it as nested sampling is doable because I suddenly find this paper. Unless I got it wrong, transect can be seen as nested if I consider spatially explicit model in my analysis.
In addition, sampling animals is technically point sampling (you cannot really sample moving animals in plots like sampling plants, can you?) so I was wondering if the moving range of animal per sampling points quite big, it is somehow nested.
I suggest you look at the autocorrelation between samples for the most abundant taxa in your transects (if you have enough samples in one transect, that is!). It tells you, whether there are regular patterns in the distribution of your taxa.
Dear Sabhrina,
It seems ok if you can group your sampling sites witjh traps in some way, traps in plantastion vs. wild forest, open area vs forested area, close to coast vs. far from it, etc., species-rich vs species poor (I refer to plants pr fruit treeis bats visit), etc.
I found the answer to my question in several researches with nested analysis. The transect itself is not a measure for nested analysis but the species I sample could imply nested structure like what Irene and Zaal said.
Article Additive Partitioning of Species Diversity across Multiple S...
Good to know that you have found an answer to your question that was asked..
- Badges
- Science topic
- Similar topics
- Education
- Higher Education
- Conferences