Calcium, magnesium and potassium in mineral soils from the southern half of Finland

,

0.1 % in the sand and fine sand soils to 2.74±0.21% in the heavy clay soils of the sur- face layers and to 3.10±0.07% in those of the deeper layers.
The portion of exchangeable Ca was relatively high: in the groups of surface soils from more than one tenth to one third of the total amount.The corresponding average amounts released by even the more drastic treatment with acid were not markedly higher.
Only a few per cents of total Mg were exchangeable and slightly higher amounts were dissolved by 0.1 N HCI, whereas the treatment with N HCI at 50°C released about half of the total Mg.Exchangeable K and K dissolved by 0,1 N HCI did not exceed 1 % of the total K, except slightly in the heavy clay soils; the average amounts released by N HCI ranged from 5 to 18 % of the total K.
The plant availability of these nutrients was discussed.
In a previous work (Kaila and Ryti 1968), characteristic differences in the total contents of Ca, Mg and K in various particle size fractions of some Finnish mineral soils were found.In all the 30 samples analysed, the Ca content was in the clay fraction (< 2/<m) distinctly lower than in the silt (2-20 /on) and fine sand (20 -200 /urn) fractions.The clay fraction again, was richest and the fine sand fraction poorest in total Mg.The K content varied in a way similar to that of Mg but differences between these three fractions were less marked.If this observation, which in respect to Ca seems to be contrary to most results from other countries, would be generally valid in Finland, it would mean that our soils rich in clay would be poorer in total Ca and richer in total K and, particularly, in total Mg than those of a coarser texture.
In order to test this assumption, a larger material of various kinds of mineral soils from the southern half of Finland were analysed for their total contents of Ca, Mg and K.More information about the occurrence of these nutrients was tried to get by comparing the portions exchangeable by am- monium acetate and the portions released by extraction with dilute mineral acid or by a more drastic treatment with acid at a higher temperature.

Material and methods
The material consists of 210 samples of mineral soils collected from various places in the southern half of Finland (between 60°and 65°).There were 62 samples from the plough layer and 76 samples from the depths of 20 cm to 70 cm of cultivated lands, and 20 samples from the surface layer and 32 samples from the deeper layers of virgin lands.With very few exceptions the samples originated from areas with a bedrock of granite, granodiorite, migmatite, quartz feldspar schist or mica gneiss.
The samples were air-dried at room temperature and crushed to pass a 2 mm screen.The mechanical analysis was performed by a pipette method, the content of organic C was estimated by wet combustion and the pH-values measured in 0.02 N CaCl 2 -suspension with the ratio of soil to solution of 1 to 2.5.Sodium carbonate fusing was used for the determination of the total contents of Ca, Mg and K. Exchangeable cations were extracted by washing with N ammonium acetate at pH 7. The amounts of Ca, Mg and K which would represent easily acid-soluble part of these nutrients were extracted by shaking 10 g of soil in 100 ml of 0.1 N HCI for one hour, and the more drastic treat- ment was the procedure developed by Schachtschabel (1961) for the deter- mination of reserve K: 5 g of soil was treated with 50 ml of N HCI at 50°C for 20 hours.
The determination of Ca and Mg in the extracts was performed by Perkin Elmer atomic absorption spectrophotometer 290 and that of K by an EEL flame photometer.

Results
Because of the rather low number of samples in some of the textural classes, no statistically significant differences in the analytical results were found between the sand and fine sand soils, between the silt and loam soils, and between the clay loam, sandy clay and silty clay samples.Therefore, these groups of soil samples were pooled, respectively.In tables 1 to 4, the sand and fine sand soils are samples containing more than 50 % particles larger than 0.02 mm in diameter, silt and loam soils are composed of at least 20% the fraction 0.02 0.002 mm and of not more than 50 % the fraction 0.2 0.02 mm, all with less than 30 % clay (< 0.002 mm).The group of coarser clays contains 30 to 60 % clay, and the heavy clay soils are characterized by more than 60 % clay.  1 show that pooling the samples in this way resulted in groups with a mean clay content of about 10, 20, 40 or at least 70 % both in the samples of surface soils and those from the deeper layers.There is no significant difference in the mean content of organic C between the groups of surface soils, on the one hand, or between those from the deeper layers, on the other hand.The pH-values not recorded in this table ranged from pH 3.9 to pH 7.0 with an average of about pH 5.2 -5.4 in all the groups.
The total content of Ca was exceptionally low in a sand soil from Satakunta, from an area with the bedrock of sandstone: 0.38 % in the sample of plough layer and 0.28 % in the subsoil.Without these samples the average content of total Ca in sand and fine sand soil was in the surface soils and l.lSdiO.12% in the deeper layers, ranging in both groups from 0.7 to 1.6 %.In the samples of heavy clay the lowest content of total Ca was about 0.5 % and the highest 1.3 %.In the other groups the minimum content was about 0.7 % or 0.8% and the maximum contents almost 1.9 %.
Also the average content of total Ca tends to be in the heavy clay soils lower than in those of a coarser texture.
On the basis of the mean values, the total content of Mg distinctly in- creases within the surface soils and within the samples from the deeper layers with the increase in the clay content.Except in the sand and fine sand soils, the surface soils seem to be somewhat poorer in Mg than the corresponding group of the samples from deeper layers.The lowest con- tents, about 0.2 % Mg, were found in several virgin or cultivated sand or fine sand soils, and the highest results among these soils were about 1.1 %.
In the samples of heavy clay soils the content of total Mg varied from 1.0 to 2.1 % in the surface samples and from 1.3 to 2.3 % in the subsoil samples.
The minimum values of the other groups were distinctly higher than that of the sand and fine sand soils and their maximum value was markedly lower than that of the heavy clay samples.
The average content of total K also increases with the increase in the clay from group to group, although in the surface samples the difference between the K contents of coarser clay and heavy clay soils is not statistically significant because of the large variation and the small number of samples.
There is no difference between the mean K content of sand and fine sand soils from the surface layer and from the subsoil, but in all the other groups the surface soils tend to be somewhat lower in total K.The lowest value was found for a virgin fine sand soil, 0.8 % K, and the highest for a heavy clay sample from the depth of 50 to 60 cm 3.4 % K.
In all these 210 samples, the mean content of total Ca is %, that of total Mg %, and that of total K O7 %.In tire 95 samples of clay soils the average Ca content is %, in the 115 non-clay samples it tends to be somewhat higher, 1.13±0.05%.The mean content of total Mg in clay soils, %, is distinctly higher than that in the non-clay soils, %.
Also the mean content of total K in the clay soils, %, is significantly higher than that in the non- clay samples, 2.20T;0.09%.
The dependence of the total content of Mg on the clay content of all samples is also proved by the total linear correlation coefficients between these variables which is r = o.Bl***.The correlation coefficient for the relationship between the total K and clay contents is somewhat lower, or r = o.73***.On the other hand, a low but significant negative relation- ship between the contents of Ca and clay is indicated by the correlation coefficient r = -o.32***.In the 95 samples of clay soils this relationship seems to be somewhat closer with r = -o.4B***.
In each group of the soil samples (Table 2) the mean content of exchange- able Ca is markedly higher than that of Mg or K, and it also represents a considerably larger portion of the total content than do the exchangeable form of the other two cations.In some heavy clay samples about one half of the total Ca was found to be exchangeable by N NH 4 OAc, while not more than about one tenth of the total Mg and less than 2 % of the total K were exchangeable.On the other hand, in some of the sand and fine sand samples of the deeper layers only 0.1 % of total Ca and 0.02 % of total Mg or K were exchangeable.Both the amount of these cations as mg/100 g of soil and the portions they represent of the total contents tend to increase from non-clay soils to clay soils more distinctly within the samples of deeper layers than within those of the surface soils with their higher content of organic exchangers.The results in Table 2 show no signi- ficant difference between the heavy clay samples from the topsoils and sub- soils.In the other groups, the mean contents of exchangeable Ca and K and also their percentage of the total contents are higher in the surface soils.This is the case with Mg only in the sand and fine sand soils.According to results in Table 3, the amounts of Ca and Mg dissolved by 0.1 N HCI tend to be somewhat higher than the amounts in exchangeable form.In most cases, however, the difference between the correspond- ing mean values is not statistically significant.Only from the non-clay and coarser clay samples from the deeper layers, more Ca was extracted by this dilute acid than exchanged by NH 4 OAc, and this is the case also with Mg in the clay soils from the deeper layers.Several samples gave somewhat lower amount of acid-soluble K than exchangeable K, though because of the rather large variation, the difference between the corresponding average values is not statistically significant.Treatment of the soil samples with N HCI at 50°(Table 4) seems to dissolve Mg very effectively, but Ca to a not significantly higher extent than 0.1 N HCI at room temperature.The average amounts of Mg released by the more drastic treatment are 7 to 18 times as high as those extracted by 0.1 N HCI, except in the sandy soils of deeper layers in which the corresponding ratio is 70.The average amounts of K released by N HCI at 50°from the surface soils are 6 to 14 times and from the samples of deeper layers 15 to 25 times as high as those dissolved by the more diluted acid.
With increasing clay content there is a distinct tendency towards higher amounts of Ca, Mg and K dissolved, though, particularly, the means of the relative release seem to differ between the adjacent groups less signi- ficantly.The part of total Mg released ranges from 3 % to 60 % in the sand and fine sand soils and from 40 % to 70 % in the samples of heavy clay.The corresponding limits for dissolution of K are 0.3 % to 12 % and 8 % to 24 %, respectively, and for the release of Ca 3 % to 37 % and 25 % to 53 %, respectively.

Discussion
The present results indicate that the total content of Mg in Finnish mineral soils seems to be fairly closely connected with the clay content of the soil.The same appears to be true also in respect to the occurrence of K, though to somewhat lesser degree.The relative poorness of clay fraction in Ca observed in the previous paper (Kaila and Ryti 1968) is revealed in the lower average content of total Ca in the heavy clay soils than in the samples with less than 60 % clay, and also in the low but statistically significant negative correlation between the Ca and clay contents of these soils.
According to an old estimation (Sederholm 1925) the rocks of Finland contain averagely 2.4 % Ca, 1.0 % Mg and 3.0 % K.In these 210 soil samples analysed, the mean content of Ca was 1.1 %, of Mg 1.0 %, and of K 2.4 %.Even if allowance is paid to the fact that soils formed from basic rocks of northern Finland were not included in the present material, it seems that losses of Ca have been rather high during the relatively short age of these soils.Data collected by Soveri (1956) of the chemical composition of our clay sediments show a higher mean content of Ca, 1.55 %, with the range from 0.8 to 2.8 %.
Yet, it is likely that these results are from deeper layers of soils than those analysed in the present study, and therefore from less weathered material.
The Mg content of the surface layers of our mineral soils seems to be high as compared with data from many other parts of the world.In a German textbook (Scheffer and Schachtschabel 1970) the range of the usual total content of Mg is reported to be from 0.05 % to 0.5 %, and according to Jackson (1956) the soils generally contain less than 0.6 % Mg.In this material even the mean content of sand and fine sand soils was more than 0.5 % and that in the heavy clay soils 1.5 % in the surface samples.
Only a few per cent of total Mg was exchangeable and watersoluble, but, as an average, about one half could be released by N HCI at 50°C.Though it is supposed that K dissolved by this treatment would represent also the potential- ly available sources of K to plants (Schachtschabel 1961), it is not likely that it is the case with Mg.It is more likely and even proved by unpublished results of some pot experiments on exhaustive cropping, that an extraction with 0.1 N HCI would give a more reliable estimation of the Mg-sources of our soils than the more drastic treatment.The diluted acid dissolved from the surface soils Mg in amounts which were not, or only slightly, higher than the content of exchangeable Mg in the corresponding samples.
If the weight of the soil of the plough layer is taken to be 2 500 000 kg, the average amount of Mg dissolved by 0.1 N HCI would correspond to 350 kg/ha in sand and fine sand soils, to 550 kg/ha in loam and silt soils, to 850 kg/ha in the coarser clay soils and to almost 2400 kg/ha in the heavy clay soils.As compared with the annual uptake of about 10 kg Mg/ha by cereal crops, these sources seem to be fairly large.Yet, losses by leaching must be taken into account, and also the fact that there are sand and fine sand soils with less than 100 kg/ha, and loam and silt soils with no more than 150 kg/ha of this «potentially available» Mg.
The sources of exchangeable K in the present soils would support cereal crops from 2 to 15 years.As far as the non-exchangeable part of the K ex- tracted by N HCI at 50°C would be released in the field rapidly enough, these sand and fine sand soils would supply 75 kg K/ha for 10 to 90 years, loam and silt soils for 15 to 110 years, the coarser clay soils from 50 to 140 years and the heavy clay soils from 70 to 200 years.Leaching of K will, of course, cause losses, but on the other hand, the deeper layers may contribute, and their contents of both potentially plant-available K and Mg tend to be even higher than that of the more weathered and leached material of the surface layers.
It seems that Ca exists in our soils either as exchangeable and water-soluble ions or as part of the mineral lattice which is not dissolved by the acid treat-

Table 1 .
Contents of clay, organic C, Ca, Mg and K in groups of soil samples (Mean values with the confidence limits at 95 per cent level)

Table 2 .
Exchangeable Ca, Mg, and K in groups of soil samples /Mean values with the confidencelimits at 95 per cent level)

Table 3 .
Ca, Mg, and K extracted by 0.1 n HCI at 20°C {Mean values with the confidence limits at 95 per cent level)

Table 4 .
Ca, Mg, and Kextracted by 1 n HCI at 50°C (Mean values with the confidence limits at 95 per centlevel)