I think, that the number of piezometers should permit obtaining sufficient number of records for statistical analyses. I think 42 piezometers, one per plot is good idea, because it allows 7 replications for each compartment (1 floodable plain, 2 intermedary 3 dry forest). May be 3 replications would be minimum for statistics, but 7 or even more (it depends if it is possible) is much better. I think it would be also recommendable to describe and sample soils from each sites. At least from augerings, although digging soil pits would be much better to make photos and good description. It is important that soils are very variable, and one one field or even plot you can find very different soils. At least in Poland we can find both sandy and clayey soils in the same field,  within a distance of about 20-50m, sometimes even in the same, one soil pit.

The depth of piezometers should depend on the compartment, probably on floodable plains the piezometers should be shallower, than in highest places. I think you should study places in the end of the dry period, when the ground water level is expected to be the deepest. And I think you should add 20-50cm to this depth for piezometer installation. I suppose, that the depth could be of about 1-2m?

On the slopes, it depends, if the soil is not underlaid by hard rock or more clayey layers. In this case the depth of piezometers might be limited by the depth of the rock. On the slope there is probably no ground water, however, it is possible subsuperficial flow of water on the clayey layer or rock. So, studying of this phenomenon would be very interesting

On highest and sandy areas the ground water level may be at greater depths (several meters?) so may be you should not make piezometers reaching this depth, but I suppose the should be not shallower than 2 meters, or better - 4 meters. But, if you have possibilty to make deeper piezometers - it would be better.

As you know this area, you have knowledge to make a good decision.

Best regards!

My previous answer disappeared from the Researchgate, so I try to repeat it more shortly.

1) The number of piezometers seems to be adequate, although if more replications, better. It would be necessary to describe and sample soil from each piezometer - better from soil pit, but if it is impossible, from augering. Soil may vary greatly even at short distances (for example - aluvial soils);

2) The depth of piezometers depends - in compartment 1) (floodable plain) even ca 1m may be sufficient, if the water table is at small depth in the end of dry season. In compartment 2) (steep slope) presence of typical ground water is rather not expected, but subsuperficial flow of water is possible, if there is a solid rock or clay layer at shallow depths. The piezometer reaching such rock or layer would be useful - 1 or 2 m. In compartment 3) (highest place, poor sandy soil) the presence of ground water is not expected, excluding situation of soild rock or clay layers and it would rather would be so called perched water. In such positions even deep piezometers of about 3-4 or more meters might be not sufficient to investigate water depth.

Hi Michal Stępień!

I´ve just receive both messages you sent me.

Thanks alot for the answers. Very explicative and easy to understand. I really appreciate your ideas and certainly is going to be very useful for me!. I was already thinking in install 7 piezometers per compartment, totalising 21 piezometers. It sounds good! I really believe it could be executable and satisfactory for the statistic analises.

In this case, i could install the piezometers in the mean point of the plots: 1 to 7 (floodable plain); 15 to 21 (steep slope); and 28 to 35 (dry forest). I guess it could be a good idea.

Best regards!

Yes, but it is important to observe soil in each site- if it is different within a compartment it may create the necessity of increasing the number of piezometers. I think a close collaboration with soil scientist is very important, although I think, that the moisture conditions are the most important factor determining plant communities.

It is difficult to answer this question without first knowing the hydro-geological circumstances that prevail in the area that you are interested in. If you have the opportunity, it would be better to first establish the watertable aquifer and likely groundwater depth using at least 10 metre deep piezometers. Then use this information to design a more targeted approach approach. Let your understanding of the groundwater system, and its response to your treatments  guide you ahead of any  faith in statistics.     

Hi Phil,

Thank you for the answer!

I would like to give you a better idea about my study area (restinga), in geological terms.

The biome known in Brazil as restinga (coastal woodland) is an ecosystem typical of the Brazilian Atlantic coast positioned over and between sandbanks, which is the result of soils originating from the accumulation of crystalline sediments and marine transgression dating from the Quaternary. These transgressions events gave the main traits of restinga, which is characterized as a low plain with soft undulations.

So, in geological scale, the restinga is a recent environment of just few thousand years, whose deepest soils are sandy. My research environment is composed by: 1) a low area (between sandbanks) floodable organic soil with the highest plant community; 2) steep slope; and 3) the highest place which is composed by sandy soils.

No doubt it would be very interesting to know the hydrological profile of the study area, but I have doubts about the cost-benefit of this study, considering I do not really know how far the knowledge of the deeper water in the soil would be predictive for understanding the dynamics of plant community, considering that the analyzed community has very shallow roots, I believe that the more upper soil layers have a greater influence on the dynamics of the ecological community.

Hi Arash Malekian! 

Entropy analysis you mean are those non-parametric analysis, like PCA, RDA,... isn't? Thanks!

If the question is about the influence of shallow groundwater on the transpiration and  water balance of the vegetation you will need to know something about the watertable depth ahead of designing your assessment. I suggest you establish a few piezometers or bores to a depth of 5 to 10 meters, and then consider the results in terms of guiding the ultimate design of your assessment.  

Very good discussion and surely will help the researcher.

I agree with Phil Dyson comment " the influence of shallow groundwater on the transpiration and  water balance of the vegetation you will need to know something about the water table depth ahead of designing your assessment. I suggest you establish a few piezometers or bores to a depth of 5 to 10 meters, and then consider the results in terms of guiding the ultimate design of your assessment".  

Jehova,

Did you try to measure water table variations with laser depth measurements?

Laser sensors have the ability to accurately measure the level of surfaces ranging from water to oil and all kinds of bulk solids and powdered materials. These sensors are non-contact, are easily installed and configured, and can offer significant advantages for certain applications over other sensing technologies.

Laser level is particularly useful for measuring water levels in constrained spaces such as deep wells.

Typical laser systems offer a measuring range of up to up to 45 m (147ft.). Laser level measuring products are available with a heated lens option that enables measurement in the presence of condensation.

Moreover, you can set-up a monitoring system, where data are sent to a central data hub, which offers big data analysis for example to monitor water volumes in an aquifer.

I have used this technique to monitor daily transpration of trees in a forest by measuring the variation in water level in a set of water wells. Worked pretty well under specific boundary conditions.

Cheers,

Frank

I guess that you start with a first sight at the aquifer parameters, depth, dynamics and homogeneity in the piezometric map. Then to identify the main sectors with some identical behaviour s well as homogeniety of the geologic layers nature. Then for its sector you have to identify the one or plus representative piezommeters.. A minimum of piezometric history would be very helpful. In any case for each chosen well try to have at least between 3 to 5m of water thickness. A significative history must consider either the dry or the rainy years..

Good luck

I think the data collection and analysis are based on your objectives. i.e. if you want to understand the influence of shallow groundwater on the transpiration and  water balance of the vegetation and/or influence on soil erosion like gully, you have consider how mach affect the shallow ground water table on that point. To know the influence of shallow ground water on gully erosion the depth of pizometere should be deeper than the gully depth and the same for other parameters.

Valuable contribution....

I agree with Omar Al-Khashman. the depth should be below the water table.

Interesting discussion. I think spatial distribution of piezometers should be tally to capture the WT variation, so that you have to fine tune the setup after monitoring existing ones for a certain period. sorry.

Good discussion

First of all you need to consider what you mean by "water table variation", e.g. spatial variability of depth to the top of the water table at fixed time points, temporal flucuation of the depth at one or more points (over some fixed time period) are two possibilities. Note that a piezometer will not give a reading for an entire plot nor an average value for the plot so one per plot will not produce very good information unless there is very little spatial variability within the plot.

There is an old paper (I think it came from the Kansas Geological Survey in the late 70's or early 80's) about how many and the spatial placement of wells to monitor the Ogallala aquifer. That is probably somewhat similar to your problem. It still requires at least some preliminary monitoring data to try and optimize the number and locations of the monitoring locations.

Hi everyone anyone have the complete answer for those questions?

First of all, consider the scale of your study area. With 42 plots of 5 to 25 meters divided over 7 lines:

1) I assume your study area is on the order of about 100 m (42*15m/7) along the transect (correct me if I am wrong).

2) Assuming horizontal groundwater head gradients of around 1 to 10%, one might expect to see variations of groundwater head around 1 to 10 m from the floodplain to the edge of your transects.

3) Assuming the groundwater is near the surface in the floodplain, you might expect the water table to be 1 to 10 m higher 100 m away.

4) Assuming your "steep slope" is around 20%, your land surface would be around 20 m higher than the floodplain.

So I would start looking for groundwater around 10-19 m below ground surface approximately 100 m from the floodplain.

If you are limited to 1.5 m piezometers, you may be limited to near the floodplain. I would start at an elevation of about 1 to 3 meters above the floodplain, put in a 1.5 m piezometer, and see if you get groundwater. Then estimate the gradient, project upslope, and see if you can hit groundwater higher.

However, you might be more interested in soil moisture near rooting depth. I believe you could install soil moisture probes within the upper 1 m fairly easily, but you will need quite a few to capture the variability. Or maybe the best approach is piezometers near the floodplain and soil moisture probes in the root zone up higher.

I agree Wes Kitlasten for your nice answer.