Sooraj -

This is not a subject matter field in which I've engaged, but it would seem you would need to decide first on a reasonable quadrat size, and then divide the geographic area of interest into those quadrats.  If each quadrat were, say, 0.25% of the total area, then you have a population of quadrats size of N = 400.  The sample size, n, of quadrats should not be more than you have resources to handle regarding nonsampling error, but not too few so that you have too much sampling error.  Further, for the various species you wish to study, some would ideally need larger sample sizes than others, so some compromise may be considered.  You may need to prioritize what species you wish to know better, and base the sample size of quadrats on needs for them.

You may find a textbook that more exactly addresses your subject matter, but there are various survey statistics books which may help you, such as Cochran, W.G(1977), Sampling Techniques, 3rd ed., John Wiley & Sons. You might look up simple random sampling, area sampling/cluster sampling, multistage sampling, and stratified random sampling.  Sample size needs are based largely on standard deviations, leading to standard errors. 

If you could do a small pilot study first, that could help work out sample size considerations, and logistic issues, and perhaps other issues before your main study.

Perhaps someone with experience in this subject matter can give you a few example approaches.

Best wishes with your project.

Cheers - Jim

Hi Sooraj,

As James has mentioned that scale is very important for your study, I think that along with this, you should consider what kind of forest ecosystem you are considering. Like in the Himalaya, 5m.*5m. quadrat and 1m.*1m. quadrat is taken for tropical rain forest ecosystem. Also, consider the diversity index, if there is a similar species, then large scale will be appropriate. So, the type of species, diversity of species, the ecosystem type, purpose/objective of the study are important  things to be considered. I hope these things might help you.

Good luck!

I do have only project experience. For a forest disease project we decided to have a refular grid with 25 m distance between each sample point, see http://ww.w.isprs.org/proceedings/XXVII/congress/part4/29_XXVII-part4.pdf

Good luck!

Franz-Josef -

The link you gave did not function, as it had an extra period that confused my phone.  :-) 

Everyone can try this: 

 http://www.isprs.org/proceedings/XXVII/congress/part4/29_XXVII-part4.pdf

In tracking it down, it appears to be this conference paper:  

GIS Based Integration and Evaluation of Multisource Datasets for Forest Damage Investigations

Behr, F.J., Saurer, H.O., Bähr, H.P.

Page(s) 29-37

Conference Paper (Scanned PDF, 682 KB) 

Can you tell us all a little more about it?  I mean, a little summary of how it relates to the current question? 

Thanks - Jim 

Hi Sooraj Sir,

Species Area Curve will help in deciding the minimum number of quadrats to conduct forest vegetation survey..

Please follow the link

http://www.sci.sdsu.edu/SERG/techniques/mfps.html

Might be helpful

Good Luck!!

Hello Sooraj Sir,

The link provided by Zarreen seems pretty useful. But depending upon the study objectives and study area, I would suggest that the easiest way is to see past studies on the same line. The vegetation entity (herbs, shrubs or trees) also has got to do a lot with the sampling protocol.

Good Luck