Hi Deepthi,

I think it is a somewhat broad question, and its answer is gonna depend on what are you planing to do. In general, keep in mind your research question, and then find the way to answer it. You can also randomly place a few quadrats in your transects and count the species within the quadrats (i.e 1x1 m) to save some time.

I hope this helps!

Alternatively there is:

Survey manual for tropical marine resources

by English, Susan A; Baker, V. J; Wilkinson, Clive R; ASEAN-Australia Marine Science Project; Australian Institute of Marine Science

1997, 2nd ed., ISBN 0642259534, x, 390 p.

I would say that if your questions concern all species within the transects then yes, you should count them all.

As for the method itself (whole transect, quadrats, smaller transects, etc.) you should weight statistical power and cost. Considering you want to count every individual of every species within the transect, how much time/money does it take to sample 1 transect? How many transects do you need to detect a difference between sites/treatments/etc., considering the variability of abundance estimates from one transect to another and the magnitude of the expected differences? You have to think about these questions with the help of previous studies or a little pilot survey.

Power is a function of sample size and variance. With the same time and money you may be able to sample 6 small quadrats or 2 large transects. 6 small quadrats would be a larger sample size, but would probably be much more variable (especially for species with patchy distributions). On the other hand, opting for a few larger areas you will be sacrificing sample size for within-sample stability. A third option is to sample a lot of big transects, if time and money are not a problem :)

Beside people have mentioned the different way (quadrat, intersection, point). And calculation and analyze method by Miguel Pais and Manuel Weinkauf. I want to rise another attention on replication and length of transect you mentioned.

Since corals grow patchy on the real world, the place you drop the transect belt might affect the result a lot. You might have nearly 80% Acropora on one patch reef then a softcoral or rubble in another 10 meter.

Therefore, if you want to survey the species composition or coverage from one spot, you might consider spread the transect in different depth then have long enough transects to do that.

Back to what Mauricio Urbina has said, it all depends on what you want to ask/answer.

If you could make it more clear, people could help more.

The common one I saw here (Australia) is 50m transect tape with 2-4 meter width and use camera + CPCe (http://www.nova.edu/ocean/cpce/) randomization point sampling to collect the relative composition.

Thank you all for the information.My main focus is on Corals of 'intertidal zone' and yes as Mr.Weinkauf guessed i am interested in species richness as well as relative species abundance,and am planning on laying a 10m transect line and 1mx1m quadrat,so i would like to know if this can be carried out using these dimensions? Also am focusing only on corals and no other species.

I appreciate Colin Wen's suggestion: use camera + CPCe. Do not restrict to only corals when there are many others around.Tomorrow you may want to know more - you may need to learn more. I worked for almost 18 years in similar work area. The data you generate with a little more effort might help some one else. The records tell you the compatibility and otherwise of different groups of fauna. It will tell you after a couple of years what is happening in the experimental area. May be you can share what is not relevant to you with others interested.

What method are you using? the Point intercept method (PIT) uses random points in the transect, you can calculate percentage cover from this. if you can identify down to species level that would be good, You know of coral watch? that could greatly help you with identification.

One of friend Dr. Chandresh Dave is presently working with Gulf of Kachchh Marine National Park, Gujarat, India. You can contact him for further details and study protocols. His email ID is: "Chandresh Dave" ,