The primary objective when collecting a soil sample for laboratory analysis is that its composition be representative of the conditions that exist in the field. Review your objective and you can get that answer

I would sample at 1 cm intervals. You can then use this information to assess the entire soil column, including the individual horizons. If you are limited by the number of samples you can take I would take a sample from each soil horizon. These would then be representative of the entire soil profile.

The best way is to take soil samples at 15 cm depth randomly at places across the paddy field. It should be borne in mind that the place of sample collection should be in V shape, the site of each collection point. Moreover, the edges of the field should be avoided since there are chances of contamination and mixed soils in these places. The interval, however does not matter.

It depends on your objective whether want the soil profile or just soil quality. Based on this you can choose either soil profile or surface sampling.

It depends on how you want to use the data. If you want to relate it to let's say soil animal data, then the upper 10 to 15 cm are enough. If you want to relate it to plant data, then you should take samples up to the depth of plant roots. If you are interested in soil property variation then you need several samples taken randomly or from a regular grid from the center of the field to avoid edge effects. If you just need an overall overview, why not take several samples and pool them.

I agree that it depends on your objective, but most standard is either to take integrated depth samples (i.e. 0-15, 15-30) or integrated samples from the soil horizons. The point samples you mention (i.e. just at 15 cm) is not a whole lots of use because they are not representative of anything except just that point. So unless you are going to follow Dan's advice and sample every single centimeter, which I have never heard of being done in agricultural settings, I would avoid the point sampling.

I think your objective of analyzing the soil profile should determine your sampling procedure. Generally, i think sampling by layers give a general overview of what that soil layer contains and its concentrations. So it generates more reasonable data than sampling at a particular depth. The interpretation of the data collected in this manner is more concrete in my opinion .

Even if you want to take only surface soil sample, taking it from 0-20 cm depth instead of 0-15 cm is more advisable; it will improve the representativeness of your sample. in fact, this strategy is widely used these days, and is evident from the recent soil science literature.

While I agree with much of the above, especially those that suggest that your research question should inform your sampling strategy, there is one additional point I can add. Consider the depth of any tillage operations when deciding on the depths of integrated samples. If ploughing is carried out to a depth of 23cm, for example, then an integrated sample of 0-23cm may be appropriate.

There is a real challenge for sampling in a plot with several soil series. Thus, we need to perhaps making sub-plotting. Surely, sampling for different depth is necessary. However, the key thing here is the prime objective and specific objectives of the task to accomplish.

Hi Mantan,

if it´s not to late for suggestions there is one thing that springs to my mind: If (a) your paddy field is waterlogged due to the cultivation of lowland rice and (b) there is a plough pan below the puddled layer, the latter should be rather homogenic in terms of physical parameters. Chemical parameters will most likely be influenced by oxygen diffusion through the irrigation water into soil and by distance to roots.

As a conclusion, one should also pay attention to rooted vs. non-rooted areas in addition to the choice of different sampling depths.

I have done some unpublished researchers on paddy fields ' soil layers and I found that the hard pan (paddling layer) had enormous effect on the overall results of my research. So I suggest that you must know the exact physical and chemical properties of this layer to have a better understanding on your outputs.