Biogeochemical Selectiveness of Cedars Over Metamorphic Rocks of the Escambray Complex, Sancti Spíritus, Cuba/Vegetation

Vegetation
Cedar (Thuja L.) is the dominant species in the survey area. Other shrubs and common grass (gramineae) are also found. Shrubs are usually located near drainages, while the rest of the area is covered by grass.

Sample collection
The most frequent type of tree in the area is a variety of cedar. We used a sampling grid of 500 x 100 meters in an area of 1.6 square kilometres, which has a pyrite-copper outcrop near to its centre. This grid gives us a total of 32 sampling points. At each of these sampling points, we took a sample from leaves, flowers, and from the end of young branches. As a guide, a soil sample was also taken, to compare the results.

Sample preparation and analysis
The initial weight of the tree's samples was 200 grams. After ignition in an open oven, we obtained near 10 grams of ashes analyzed for Pb, Ag, Mo, Cu and Zn by AA. Soil samples were taken at a depth of 0,2 meters, nearly the sampled tree. Each soil sample weighted around 200 grams. After been Sun dried, they were pulverized and sifted at -200 Mesh, and then analyzed for the same set of elements, using also AA.

Sampling took place only very early in the morning. By midday, the samples were sent to the laboratory and processed during the same day. The entire sector was sampled in two days.

Data Interpretation and Geomathematical Modelling.
All the original data, the "corrected" ones, and their statistical analysis are presented in tables 1, 2, and 3, respectively. To facilitate the reading process, all the tables are located in a separated chapter.

The process of "correction" of the data included:

a- The substitution of the "zero" (0) values, by the half of the limit of sensibility of the used equipment, to avoid mathematical problems during the use of logarithms. Therefore, the limits in ppm. for each element are the following:


 * {|class="wikitable"


 * Pb = 0.1		||	Mo = 3		||	Zn = 3
 * Ag = 0.1		||	Cu = 2 ||
 * }
 * }

b- The detection of "hurricane" data and their substitution by the mean of the sample, calculated without excluding those hurricane data. We considered as hurricane, all data that fulfil the condition stated in equation (1).

'''Equation 1. Determination of hurricane values.'''

$$X_h > Mx \pm 3\times4\sigma \,$$ Due to the similarity between the S.G.L and the L.G.S formations, our first step was to determinate their geochemical analogy as well. To do so we use the Student (2) and Fisher (3) criteria of analogy (Kazhdan, 1979), establishing the critical value in both cases as two.

$$t = \frac{{}^a\bar{X} - \bar{Y}^a}{\sqrt{\frac{S_x^2}{n_x} + \frac{S_y^2}{n_y}}}$$ Where:
 * $$\bar{X}$$ - Media of the X sample
 * $$\bar{Y}$$ - Media of the Y sample
 * $$S_x^2$$ - Deviation of the X sample
 * $$S_y^2$$ - Deviation of the Y sample
 * $$n_x$$ - Size of the X sample
 * $$n_y$$ - Size of the Y sample

$$F=\frac{S^2_1}{S^2_2} \quad for \ S^2_1 > S^2_2 \quad$$

Tables 4 and 5 shows the initial data for these calculations. For obvious reasons this test was done only for soil samples. As table 6 shows, there are geochemical differences between the two compared formations, so they should be treated independently.

Let us present now the geochemical characteristic of each geological formation so we may reach to some generalized results.

The S.G.L. formation.
Next a set of tables will be presented from which it will be easy to observe which elements are concentrated in each part of the plant (tables 7 - 9), and also a comparison of the biogeochemical with the soil sampling (see equation 4). Clark values for plants were taken from Yagodin (1986), while same values for soil samples were obtained from Röster and Lange (1972).

$$W = \frac{\bar{X}_{l,f,b}}{\bar{X}_{soil}}$$ To determinate the geochemical selectiveness of each sample, we determinated the range of their media values, and substitute them with roman numerals. I will correspond to the highest value of the media and III - to the lowest as it is shown in table 10.

Although all the sampled parts of the tree were good concentrators of the elements, only the leaves in this formation presented better concentrations that the soil samples, and that only for the lead and silver. The geochemical specialization of this formation is the following:
 * {| class="wikitable"

!LEAVES: !FLOWERS: !BRANCHES:
 * Pb, Ag, Cu, Mo, (Zn)
 * Mo, (Ag), [Pb, Cu, Zn]
 * Zn, (Pb, Ag, Mo, Cu)
 * }

It is easy to see that the leaves present the best concentration values for this formation.

The L.G.S. formation.
The geochemical characteristics of each part of the trees in this formation are shown on tables 11 - 13. As it can be observed from table 14, the geochemical specialization of the L.G.S. formation is the following:
 * {| class="wikitable"

!LEAVES: !FLOWERS: !BRANCHES:
 * Pb, Ag, (Mo, Cu, Zn)
 * Mo, Cu, Zn, (Ag), [Pb]
 * (Pb, Ag), [Mo, Cu, Zn]
 * }

These results are very much alike to those from the S.G.L. formation, but with better concentration values for the flowers. The H.A.R. formation.

The geochemical characteristics of each part of the trees in this formation are shown on tables 15 - 17. As it can be observed in table 18, the geochemical specialization of the H.A.R. formation is the following:


 * {| class="wikitable"

!LEAVES: !FLOWERS: !BRANCHES: These results are similar to those obtained earlier. Clearly the best elements detected by cedars in this area are lead and silver.
 * Pb, Ag, Mo, Cu, Zn
 * (Ag, Mo, Cu, Zn), [Pb]
 * (Pb), [Ag, Mo, Cu, Zn]
 * }

The O.S. formation.
See the geochemical characteristics of this formation on tables 19 - 21. As table 22 shows, the geochemical specialization of the O.S. formation is the following:


 * {| class="wikitable"

!LEAVES: !FLOWERS: !BRANCHES:
 * Pb, Ag, Zn, (Cu), [Mo]
 * Mo, Zn, Cu, Ag
 * [Zn, Cu, Ag]
 * }

Since these results are very similar for all the geological formations in the area, it is possible to compare them by showing which part of the plant is the best concentrator, as it can be observed in table 23.

From table 23, we can conclude that the best concentrators in the area are the leaves and in a second place the flowers. The general effectiveness of the studied parts is shown above:


 * {| class="wikitable"

!LEAVES: !FLOWERS: !BRANCHES:
 * Pb, Ag, (Mo, Cu, Zn)
 * Mo, (Zn), [Cu, Ag]
 * (Zn), [Ag, Cu]
 * }

From Fig. 3, it can be noticed that cedars have a complete different set of concentrated elements than those from soil samples. In fact, cedars chiefly concentrates those elements related to the ore, while soil samples are composed mainly by molybdenum (Fig. 4).

Figure 3. Biochemical concentrations. The first two diagram pies show the different sets of elements that are concentrated on cedar (a), and on the soil samples (b). The third pie (c) shows that cedars concentrate better the elements related directly to the ore, while the soil samples concentrate mainly the molybdenum, which is not related to this type of ore.

Figure 4. Geochemical behavior of elements. Diagram (a) shows the relationship between the concentration of elements on soil and cedar samples. Diagram (b) shows the same relationship after all the data were transformed to a 100% range. Diagram (c) clearly shows the dominating role of molybdenum in soil samples, while all the ore-related elements are concentrated in the cedar samples.

The next question that we should answer is which method provides better results, the biogeochemical or the soil sampling, and what part of the cedar is the best to be sampled, which is our next subject.

Effectiveness of the Method.
To determine the most informative method, we are going to calculate the degree of contrast of their anomalies (Valls, 1993), using for that the equations (5) and (6).

$$\varepsilon = \sqrt[3]{\frac{X_{3+}}{\bar{X}}}$$

$$\tau = \frac{1}{\log{\varepsilon}} \times \log \left( \frac{X_{max}}{\bar{X}} \right)$$

The better method would not be the one with the greatest medium value, but the one with the higher contrast of its anomalies (τ). The initial data for these calculations are presented in table 24. The results of this procedure are display in table 25.

From table 25 it is clear that biogeochemical samples provide better results then soil samples. Another inference that follows from this table is that branches are the best part of these cedars to be sampled.