Previous Page Table of Contents Next Page

Evaluation of site fertility

Soils Department, Forest Research Institute, Helsinki, Finland (FAO/André Mayer Fellow)

MANY methods have been used for the evaluation of forest site fertility. The most important of these are based on certain characteristics of the stand, on the ground vegetation, on the physical or chemical properties of the soil, or on the nutrient content of the foliage. The disadvantage of the methods based on the characteristics of the stand is that they do not allow the evaluation of a treeless area. Moreover, the treatment of the stand may influence these characteristics. Evaluation based on ground vegetation is almost confined to Finland; elsewhere it has not been found to give a correct picture of site fertility. Since these methods are not wholly satisfactory, great hopes have been entertained of chemical or physical methods of evaluating the fertility of a forest site, either the soil or the foliage being analyzed.

A couple of decades ago soil analysis was much more widely employed than foliar analysis, but in recent years the latter has gained increasing recognition. (Foliar analysis is advocated, for instance, by Tamm 1956, Leyton 1958, Wittich 1958, Wehrmann 1959, and Strebel 1960, soil analysis by Aaltonen 1950, 1951, Viro 1951, 1955, and Wilde 1958). The purpose of the present study is to discuss which of these two methods, soil or foliar analysis, better describes the fertility of forest sites under various soil and climatic conditions.

The study was undertaken with the help of an FAO André Mayer Fellowship and this paper is a short résumé of the final report presented to FAO. The field work in the different countries was made possible by the courtesy of the local forest research organizations.

Material and methods

Area and samples

This study was restricted to Scots pine (Pinnus silvestris L.) and Norway spruce (Picea abies Link), since these species have quite an extensive distribution and great economic importance in Europe. The material was collected during the autumns of 1958 and 1959. In south Finland, Sweden, Germany, Poland and Austria, 10 pine and 10 spruce stands were studied, in France 10 pine stands, and in north Finland 30 pine stands. These sample plots are not necessarily representative of any country as a whole.

The location of the sample stands is shown for pine in Figure 1 and for spruce in Figure 2. The figures also indicate the natural geographical distribution of Scots pine and Norway spruce in Europe. As far as possible the plots were located in middle-aged stands and on level ground. The Finnish, Swedish, and Austrian stands were on morainic Boil, the Polish and French on alluvial soil, and the German stands on both.

As far as possible permanent sample plots belonging to the local forestry organizations were used. In hitherto uninvestigated stands the measurements were carried out by the current methods of the Finnish Forest Research Institute. The dominant height of the stand was measured, and the corresponding height at 60 years of age was determined from the local growth tables or with an increment borer in connection with foliage sampling.

The samples of mineral soil were taken from four pits, surface soil (0 to 30 centimeters) and subsoil (30 to 60 centimeters) being sampled separately, and the Samples of corresponding layers combined. In these pits the soil formation type was also determined and the horizons measured. The humus layer was sampled with a cylinder at 20 points on the diagonals of the plot. The content of stones was measured by the rod-testing method.

Needle samples were taken from five dominant trees at the highest level that could be reached by climbing, usually two to three meters from the top. The samples were taken from the south side and from unshaded branches. Current and two-year needles were sampled separately.

Laboratory work

All laboratory work was done at the soils laboratory of the Finnish Forest Research Institute, by the usual methods employed in this laboratory. The following analyses were carried out:

1. Mineral soil: bulk specific gravity, mechanical composition, exchangeable nutrients (NH4 - acetate), total nitrogen, humus content, basic mineral index (Tamm), water-holding capacity (Bouyoucos), and acidity in water and KCl.

2. Humus layer.- exchangeable and total nutrients, total nitrogen, acidity, and CO2 evolution.

3. Needles: total nutrients.

As usual, the soil analyses were made with the fine soil (> 2 millimeters). To ensure that the results of different countries were mutually comparable, the amount of nutrients per hectare was calculated and these figures were used in all discussions.

FIGURE 1. - Location of pine stands.

FIGURE 2. - Location of spruce stands.

Site index

Several characteristics of the stand, such as height, diameter, basal area and volume, have been used as indicators of site quality. In this study the site fertility is given by the average height at 60 years of age of the 100 thickest trees per hectare. It is here called the site index. Of the characteristics of the stand mentioned above, the dominant height is the least dependent on the treatment of the stand, provided that there has been no crown thinning.

The stands of the same species of tree in a country form a unit in the first discussion of the results. The stands of a unit were divided into two groups according to the site index. The five stands with the highest site indices form group I, the five with the smallest values group II. The average site index was equally high on brown soil and brown podsolic soil and these plots were combined in the discussion. The average site indices and ranges of variation were:





Pine, podsol soil



Pine, brown soil



Spruce, podsol soil



Spruce, brown soil



Results of the study

Results of soil analyses

It is impossible to present here the whole of the data obtained during the study. However, to provide the reader with some idea of the nutrient status, the average amounts of exchangeable nutrients, total nitrogen and organic carbon in the surface soil layer are given in Table 1. This Table shows very clearly that the level of nutrients varies greatly in different countries and different soil types. Almost without exception the quantities in a country were largest in the brown soils and smallest in the podsol soils. In the pine stands the most conspicuous features were the small amounts of calcium in the French soils and of magnesium in the Polish podsol soils, and the great amounts of phosphorus in the Polish soils. The nutrient levels of the Finnish and Swedish soils were surprisingly similar. In the spruce stands the most conspicuous features were the large amounts of lactate-soluble phosphorus in the Polish soils. These data do not allow a general comparison of pine and spruce stand soils but the nutritional status seems to vary greatly in different circumstances.

The distribution of the nutrients in the profile analyzed varied for different nutrients and soils. In the brown soils 50 percent or more of these nutrients were usually contained in the subsoil, in the brown podsolic soils this figure was somewhat lower and in podsol soils about 30 percent. A fact that is often overlooked is that usually only a small fraction of the organic material and nitrogen is present in the raw humus layer. Here the bulk of these elements was found to be in the surface soil, and even in the podsol soils the subsoil contained much more nitrogen and often organic matter, too, than the humus layer.


The C/N ratio was highest in the raw humus layer and lowest in the subsoil. In corresponding layers the ratio was usually highest in the podsol soils and lowest in the brown soils. The corresponding figures in the different countries did not usually vary to any great extent except for the Polish figures, which were exceptionally low, especially in podsol soils. The ratio for north Finnish soils was also very low. In the humus layer the ratio was higher in the spruce stand soils, in the mineral soils usually in the pine stand soils.

Soil nutrients and site fertility

The fertility of a site is determined by the climate and the soil. Since the other main factor affecting fertility, the climate, varies greatly within the research area, the sample plots in a country were selected within as small an area as possible, so that the climatic variations in a unit would be negligible. The local correlation between the site index and the properties of the soil was first examined graphically. After this the mean quantities of nutrients for the site index groups described earlier were calculated and the significance of the difference of means in a country was studied against the null hypothesis with the t-test. In more extensive areas the correlation was also studied by statistical analysis. Pine and spruce stands are discussed separately.

Examination of the pine stands. A significant difference in the quantities of nutrients between the index groups was most frequent with regard to nitrogen and manganese. As for the other nutrients, it was only in the amounts of phosphorus that a significant difference was encountered in two countries. As in earlier studies, a significant correlation was found between the fertility of pine stands and the amounts of calcium and magnesium in south Finland.

The site fertility showed less frequent correlation with the physical characteristics of the soil than with the chemical characteristics. In the French, Polish and north Finnish stands there was a correlation between the Bite fertility and the content of fine soil, in the south Finnish and German stands between the fertility and the basic mineral index. Correlation between the fertility and the water-holding capacity was found in the north Finnish stands only.

Examination of the spruce stands. No significant difference in the amounts of a nutrient was ever found in two countries. The most conspicuous feature is the highly significant negative correlation between site fertility and certain nutrients in Austria. This indicates that these nutrients are present in supraoptimal amounts in these calcareous soils. In south Finland a significant correlation between the site fertility and the amounts of calcium and magnesium was also demonstrable in spruce stands.

There was no correlation between any physical soil characteristic and the fertility of spruce stands.

The acidity of the soil and the total nutrients of the raw humus layer proved to be poor indicators of soil fertility in both pine and spruce stands.

The sample plots were intentionally chosen from different soil formation types. It is quite possible that this is a reason for some of the above results. It is to be regretted that more stands could not be studied in each country.

Needle analysis and site fertility

The average nutrient contents of needles are presented in Table 2. The contents were almost equal on brown and brown podsolic soils and they are presented combined. There was a tendency for the average content of most nutrients to be lower on podsol soils. The content varied considerably in different countries but this variation was rather irregular. The variations in the contents of current and two-year needles were in general similar. The nutrient content was calculated on both dry matter and ash. The latter figures did not provide any additional information for present purposes.


Soil type








Percentage of dry matter


Current needles

















Two-year needles


















Current needles

















Two-year needles

















The nutrient content of the needles varied much less than the quantity of nutrients of the soil. The former varied almost independently of the latter. Only for nitrogen was a highly significant (0.1 percent level) correlation found between the nitrogen content of the pine needles and the amount of nitrogen in the surface soil.

The correlation between the site fertility and the nutrient content of the needles was in general not significant. In pine stands a significant correlation is found in two countries only for the nitrogen and phosphorus of the current needles and the nitrogen and manganese of the two-year needles. In spruce stands a significant correlation occurred much less frequently than in pine stands and was never encountered for any nutrient in more than one country.


Effect of climatic factors

These studies did not reveal any distinct effect of climate on the nutrient status of the soil as a whole. The amounts of organic matter and nitrogen, however, were largest in central European soils, smaller in southern Scandinavia and smallest in north Finland. The effect of climate on the nutrient content of the needles proved to be very slight.

A significant correlation between the site fertility and the nutrient level was usually only found when all the stands of a unit were on the same soil type. Therefore, an examination was made of the possibility of combining stands of adjacent countries in the consideration of the results. It proved that in central European pine stands on podsol soil there was a highly significant correlation between the site index and the quantity of nitrogen in the surface soil. In brown and brown podsolic soils the correlation depended on the calcium level of the soil. When soils rich in calcium were excluded (maximum CaO included 1,700 kilograms, minimum excluded 5,200 kilograms per hectare) a highly significant correlation was found between the site index and the quantity of nitrogen of the surface soil both in pine and spruce stands. (On podsol soil there were three spruce stands only.) There was no significant correlation between the site index and the other nutrients in this extensive area.

The nutrient content of the needles showed some correlation with site fertility in the more extensive area. However, as is shown, the nitrogen content of the needles follows the nitrogen level of the soil and so this correlation is only apparent.

These stands are fairly representative of the natural range of pine and spruce in central Europe. The above correlations indicate that the climate throughout this area is equally favorable for the growth of these conifers. The climates of the Finnish and Swedish research areas seem to differ so much in this respect from one another and from the central European climate that these areas must be discussed separately.

Evaluation of site fertility

The material of this study is not sufficient to allow final conclusions as to which of the methods, soil or foliar analysis, better describes site fertility. However, the results clearly indicate that, in spite of the limitations to he pointed out in this paper, soil analysis gives more reliable results. When the differences between the index groups were examined, it was noticed that a significant difference was much more often found in the amounts of nutrients in the soil than in the nutrient contents of the needles. In particular, the amounts of nitrogen, manganese and phosphorus proved to be important local indicators of site fertility. In addition, the correlation found in central European stands between the site index and the nitrogen level of the soil shows more definitely the superiority of soil analysis for the evaluation of site fertility.

The present results indicate that nitrogen best describes the soil fertility. This study does not settle the mutual order of importance of the other nutrients. This order is probably different in different circumstances, e.g., on different soil formation types. Thus, in a scientific investigation as many nutrients must be analyzed as possible. When the local conditions are adequately known, an analysis of a few nutrients may suffice for the evaluation of forest site fertility.

These studies indicate that manganese frequently shows a strong correlation with site fertility. We assume at present that plants only need manganese in very minute amounts, but trees may have some unknown needs. It is evident that much more attention must be paid to the role of manganese, and perhaps of many other nutrients, too, in tree growth.

When the data for central European pine and spruce stands on the same soil type were combined, the correlation coefficient between the site index and the quantity of nitrogen was of equal significance (0.1 percent level) to that for one species of tree only. This would indicate that these species have similar growth at similar nitrogen levels of the soil in this area.

It was estimated that only one sixth to one half of the rotating nutrients of the soil are exchangeable. On the other hand, nitrogen is totally determined by Kjeldahl analysis. There are indications that plants may be able to use nonexchangeable nutrients and, through contact feeding, even mineral particles of the soil. This may be one of the reasons why nitrogen analysis here most universally reflected the soil fertility.

Many scientists consider that needle analysis gives more reliable results in the evaluation of site fertility. However, this method has mainly been used in young stands or plantations and the present study deals with mature stands. The greatest value of needle analysis is perhaps in revealing the deficiency of certain nutrients. It seems evident that in the stands studied here the nutrient content of the needles has usually been above the limit of deficiency, with the possible exception of the nitrogen content of the Scandinavian pine stands. The needles never showed chlorosis.

The citric acid cycle of metabolism contains several organic acids and certain plants contain other acids too, for instance oxalic acid. It is apparent that the function of the cations of cells is in large measure to balance the organic acids and inorganic anions such as phosphate, sulphate and chloride (e.g., DeKock 1958). For this reason it is hardly to be expected that analysis of foliar cations would satisfactorily reflect site fertility.

Soil analysis

The correlation between the site index and the nutrient quantity was calculated not only for the surface soil but also for the subsoil and for the total layer analyzed. It proved that the correlation coefficient was generally most significant for the surface soil. These results confirm the earlier observation that it is most important to know the nutrient level of the upper mineral soil. The exact depth of the soil layer that should be analyzed probably varies in different countries, but ocular observations in connection with the field work indicate that the depth of the roots of either species of tree usually varies within quite narrow limits. Thus, the depth of the surface soil layer to be analyzed for the evaluation of site fertility could probably he agreed universally without much risk.

For soils with a water-holding layer or ground water at a depth that the roots can reach, an exception to the usual procedure is required. These soils must be studied in some cases to a depth of several meters. The investigator must be prepared for this in all alluvial soils.

In studies on soil fertility the exchangeable nutrients are commonly determined. The amount exchanged depends on several factors and one of the least studied of these despite its probable importance is the function of vegetation, and of organic matter in general. It is known that many organic acids make chelates, complexes or precipitates with polyvalent cations and prevent the exchange of these cations.

Random analyses indicate that these "inhibitory compounds" are present in much greater amounts in spruce litter than in pine litter. Perhaps this partly explains why the correlation between the site fertility and the mineral nutrients of the soil was poorer in spruce stands. As an example of the amounts of an inhibitory compound it may be mentioned that a spruce litter sample recently analyzed in our laboratory contained about 50 mille equivalents per 100 grams of oxalic acid. Calcium oxalate is sparingly soluble and this calcium cannot be extracted with ammonium acetate. The soils of the present study also varied greatly in the percentage exchangeability of the nutrients of the humus layer. For these reasons more fundamental work needs to be done on the chemical methods of soil analysis. This work must be done on forest soil samples and not on agricultural soil, as hitherto.

Phosphorus was determined in this study by two methods, extraction with ammonium acetate and with calcium lactate. The amounts extracted were usually smaller with lactate than with acetate. However, in all the Polish soils, but especially in the pine stand soils, the amounts extracted with lactate were much larger. A significant difference between the index groups was usually detected simultaneously with both methods.

A negative correlation between site fertility and the basic mineral index was found in the pine stands of north Finland. Some of these sample plots were on a soil derived from granulite, and the heavy minerals of granulite (specific gravity > 2.68) are almost worthless as a source of plant nutrients. Thus, the applicability of this method in fertility studies must be confirmed by mineralogical analyses.

A highly significant (0.1 percent level) correlation was found between the water-holding capacity and the content of fine soil (< 0.02 and < 0.002 millimeter). Thus, this simple method serves very well as a rapid measure of soil texture.


Both soil and foliar analyses are widely used as means of evaluation of site fertility. Therefore we must try to determine the conditions in which one of these methods has more advantages and fewer shortcomings than the -her.

It was mentioned earlier that foliar analysis may reveal deficiency of a nutrient. It might also have advantages on slopes where the moving soil waters continuously bring new nutrients. On stratified soils, too, foliar analysis may, perhaps, have advantages. However, the foliar sampling season is very short. Moreover, the results are greatly dependent on the weather before sampling, for instance on the rainfall, and a stand must be sampled over several years (Wehrmann 1959). The results may also depend on metabolic disturbances (DeKock 1958). And finally, foliar analysis does not allow evaluation of the fertility of a treeless area.

Soil analysis, on the other hand, probably gives more reliable results in mature and young stands, and will do so all the more when the nutrient requirements of trees in different soil conditions are better understood. In sapling stands and plantations the two methods may be of equal value. However, it is evident that the results of foliar analysis are more difficult to interpret than those of soil analysis (Wilde 1958).

The final conclusion to be drawn from these studies is that, of these two methods, soil analysis should be the standard method in the critical evaluation of forest site fertility. When the defects of soil analysis are overcome, it will tend to supplant foliar analysis. In exceptional cases, soil analysis may, perhaps, be supplemented by foliar analysis.


AALTONEN, V. T. 1950. Die Blattanalyse als Bonitierungsgrundlage des Waldbodens. Commun. Inst. For. Fenn., 37 (8). 41 p.

AALTONEN, V. T. 1956. Die Blattanalyse als Bonitierungsgrundlage des Waldbodens. Commun. Inst. For. Fenn., 45 (2). 21 p.

DEKOCK, P. C. 1958. The nutrient balance in plant leaves. Agricultural Progress, 33:88-95.

LEYTON, L. 1958. Relationship between the growth and mineral nutrition of conifers. In Thimann, K. V. Physiology of forest trees. New York, Ronald Press, p. 323-345.

STREBEL, O. 1960. Mineralstoffernährung und wuchleistung von Fichtenbeetänden (Picea abies) in Bayern. Forstw. Centbl., 79:17-24.

TAMM, C. O. 1953. Growth, yield and nutrition in carpets of a forest moss (Hylocomium splendens). Medd. Skogsforskn Inst., Stockholm, 43 (1). 140 p.

VIRO, P. J. 1951. Nutrient status and fertility of forest soil. 1. Pine stands. Commun. Inst. For. Penn., 39 (4). 54 p.

VIRO, P. J. 1956. Investigations on forest litter. Commun. Inst. For, Fenn., 45 (6). 65 p.

WILDE, S.A., VOIGT, G. K. & STOECKELER, J.H. 1958. Response of coniferous seedlings to soil applications of calcium and magnesium fertilizers. Soil Sci. Soc. Amer. Proc., 22:343-345.

WITTICH, W. 1958. Bodenkundliche und pflanzenphysiologische Grundlagen der mineralischen Düngung im Walde und Möglichkeiten für die Ermittlung des Nähr stoffbedarfes. Allg. Forstzschr., 13 (10): 121-124.


A major step toward a soil map of the world was taken in June 1961, when an advisory group met in Rome to approve the timetable for a joint FAO/UNESCO undertaking which will extend over a four-year period.

This first soil map of the world should provide a realistic picture of the distribution of major soils and, by co-ordinating national and international effort, will allow the exchange of knowledge between one region and another. It will enable conclusions to be drawn as to how different soil areas can be used. The small-scale world map will derive from larger-scale regional maps, such as the European soil map, and will also lead to larger-scale regional versions and eventually to soil resources maps which will interpret soil properties in terms of agricultural production potential.

Co-ordinated by FAO's Land and Water Development Division, the first phase of the work will be the preparation of regional maps bearing an integrated legend: Australia in 1962; North America, South and Central America, Europe and the Near East in 1963; north and central Asia (which together with Europe contains the area of U.S.S.R.), south and southeast Asia, and Africa in 1964. In this way a series of continental maps and a first look at the world soil map will be available for the next Congress of the International Society of Soil Science to be held in Romania in 1964. The second phase will be the preparation of a second draft of the world map in 1966, with a view to final publication in 1968 in the form of a loose-leaf atlas comprising some 40 sheets.

Previous Page Top of Page Next Page