0 RS Proceedings of thwk .I dt"ll th6lnternat ...... held at aoligb"M I k5
[ON TNR TrRwn N LAM, V p m [i r-iMmVEENM rv-'rv 0D
...... Internatio*0 'Bern, S, ...... Development of K-Fertilizer Recommendations 22nd Colloquium of the International Potash Institute
Soligorsk, USSR June 18-23, 1990
Development of K-Fertilizer Recommendations
International Potash Institute, CH-3048 Worblaufen-Bern/Switzerland P.O. Box 121 Phone: (0)31/58 53 73 Telex: 912 091 ipi ch Telefax: (0)31 58 41 29 © All rights held by: International Potash Institute P.O. Box 121 CH-3048 Worblaufen-Bern/Switzerland Phone: (0)31/58 53 73 Telex: 912 091 ipi ch Telefax: (0)31/58 41 29
Design: Mario Pellegrini, Bern Printing: Lang Druck AG, Liebefeld-Bern Proceedings of the 22nd Colloquium of the International Potash Institute
Contents
Opening Session Page N. Cello Welcome address ...... 9 A. Podlesny Welcome address by the Director General of Byeloruskali ...... 13
Session No. 1 Potassium demand in cropping systems J Breburda Development of agricultural yield levels and soil K-status in Eastern and Western E urope ...... 17 A. van Diest The position of K in nutrient balance sheets of the Netherlands ...... 41 M. Kerschberger and Records of soil fertility in the GDR 55 D. Richter M.A. Florinsky and Agrochemical monitoring of exchange- E.N. Yefremov able potassium in arable soils of the U SSR ...... 63 U Kafkafi The functions of plant K in overcoming environmental stress situations ...... 81 V V. Prokoshev Coordinator's report on the 1st Working Session ...... 95
Session No. 2 Soil types and K-dynamics V V Prokoshev and Soil properties and potassium be- TA. Sokolova haviour ...... 99 H. Grimme Soil moisture and K mobility ...... 117 M. De Nobili, L. Vittori Antisari and P Sequi K-uptake from subsoil ...... 133
3 S. Feigenbaum, A. Bar-Tal Dynamics of soil potassium in multi- and D.L. Sparks cationic systems ...... 145 L Szabolcs K-status and dynamics in salt affected soils ...... 163 LM. Bogdevitch Coordinator's report on the 2nd Work- ing Session ...... 173
Session No. 3 Soil and plant test methods and their calibration for long-term sustainability of soil fertility A. E. Johnston and The use of plant and soil analyses to K. W. T Goulding predict the potassium supplying capac- ity of soil ...... 177 G. Wessolek Estimation of root density in modelling nutrient requirements ...... 205 PB. Barraclough Modelling K uptake by plants from soil 217 J. Decroux Leaf analysis for standardising soil ana- lysis ...... 23 1 Y.M. Heimer, A. Golan- Tissue protein as indicator for the K + Goldhirsh and H. Lips nutritional status of plants ...... 247 A. van Diest Coordinator's report on the 3rd Work- ing Session ...... 257
Session No. 4 Evaluation of field experiments and fertilizer recommendations 0. Jourdan and Field experiments for fertilizer advice - P Villemin design - execution - interpretation .. 263 D. Collin Soil and plant test data in computerised fertilizer recommendations ...... 279 E. Andres Soil fertility data banks as a tool for site- specific K-recommendations ...... 291 R. Czuba Experience with fertilizer recommenda- tions in Eastern Europe ...... 307 M. Perelli Experience with fertilizer recommenda- tions in Italy and in Southern Europe 317 A.A. Sobachkin and Experience with fertilizer recommenda- LM. Bogdevitch tions in USSR ...... 335
4 Ch. Pieri Coordinator's report on the 4th Work- ing Session ...... 345
Session No. 5 Implementation of fertilizer recommen- dations with special reference to potash fertilizers P Hotsma Channels to reach the farmer ...... 349 H. Vandendriessche Adjustment of fertilizer dressings to M. Geypens and J. Bries achieve high quality and optimum price for potatoes and sugar beet ...... 359 TE Gately and Need of improved livestock production W.E. Murphy from farm produced feed and the conse- quences for fertilizer application .... 373 H. Vis The range of various K-fertilizers and their agricultural use ...... 385 H. J Scharf Environmental aspects of K-fertilizers in production, handling and application 395 K. Mengel Coordinator's report on the 5th Work- ing Session ...... 403
Closing address N. Celio Closing address ...... 405
5 Opening Session Welcome Address
Dr. N. Celio, former President of the Confederation of Switzerland, President of the International Potash Institute, Bern/Switzerland
Ladies and Gentlemen, I welcome you to our Colloquium in Soligorsk. It is gratifying to see that the potash producers and relevant authorities are represented and that we have with us the gentlemen of the press. I welcome especially all those whose efforts have made this event possible. Those of you who have attended earlier meetings of the International Potash Institute will have noticed with regret the absence of a familiar figure. I refer to our esteemed Director and Organizer Heinz Ktinzli who died a few days before Christmas last year. It is fitting that we remember him on this occasion. I am especially grateful to the Director General of Byeloruskali, Mr. Pod- lesny, for his kind invitation and I wish to thank him and his coworkers for the great efforts they have made in organizing the 22nd Colloquium of IPI. It is a privilege to hold our Colloquium in this great country which, unhappily, is not well known to us. We hope we shall have the opportunity here to repair this omission by learning something of a small part of your country and its history and that we shall be able to gain some idea of the cultural traditions of the Soviet Union. This is the first time that an IPI Colloquium has been held in the Soviet Union and in a place where a major potash producer is situated. This is one result of the decision of the group of Soviet producers to join IPI as an affiliate member; members of this group are Uralkali, Byeloruskali, Sylvinit and the Institute of Halurgy. It is a great pleasure to have the delegates of these compa- nies in our midst. There is a number of reasons why IPI membership is impor- tant. First, all the potash producers of Europe and the Near East are now united in a common aim so we can say that, in IPI, the vision 9 uium entitled ((Development of K-fertilizer recommendations >. Numerous aspects of this subject will be covered by eminent scientists drawn from all parts of Europe. I would like to expand a little on the responsibility of the producer for his product; the potash producer obviously has a responsibility for the quality of his material but he must also be concerned that what he supplies is used in an efficient manner. Rational use of potash is only possible through sound research and education and this is the affair of both governments and the industry itself. So long as there is a need to improve crop yield and quality these partners have a common aim and their cooperation must be fruitful. The potash producers of Western Europe have a long tradition of support to research and education; for more than 100 years this has benefited both industry and farming. This tradition was the bedrock upon which the Interna- tional Potash Institute was founded almost forty years ago with its head- quarters in Bern. Many of you may not yet be familiar with our publications nor may you have attended earlier IPI Congresses and Colloquia so it may be helpful if I briefly outline the activities of IPI which are now also supported by the potash producers of the Soviet Union and by those of the German Democratic Republic and Jordan. The task of the IPI is to collect and evaluate the results of research on the major plant nutrient potassium from all over the world and to disseminate the information to research and agricultural extension workers in all countries. This task is approached in the following ways: - Regular and occasional publications are sent to practically every relevant research and experimental establishment in the world. The POTASH REVIEW appears in English in alternate months and is directed primarily to research workers. Another regular publication is the INTERNATIONAL FERTILIZER CORRESPONDENT, a quarterly in English and Spanish. Other publications which are much appreciated are books and booklets on specific topics in potassium research and on individual crops which ap- pear under the general heading <«Fertilizing for high yields .Mr. Sychevsky has kindly arranged to display a selection of IPI publications at the front of the hall. - Our scientific meetings, Colloquia and Congresses are arranged in working cycles by I P1's Scientific Board. The 13 members are scientists from 13 Euro- pean countries who decide the topics to be discussed and select the speakers. Following 3 Colloquia on closely defined subjects and with restricted par- ticipation (100-150) a Congress is held to evaluate the series of Colloquia in front of a larger audience (200-300). The two Colloquia preceding this meeting were devoted to Methodology of K-research in the soil (held in Austria) and in plants (Belgium). In addition to these international meet- ings IPI has also arranged Workshops on Potassium on a national basis in several countries of Asia, Africa and Latin America. - We have access to a computerised information bank in which all the litera- ture relating to potassium in plant nutrition is digested. This forms a basis 10 for our own work and publications and offers a service to all specialists who approach us for information. - A competition for young research workers under forty years of age is held every three or four years. The last of these, in 1990, attracted 122 entries from 59 research workers from 26 countries. The submissions are put before a jury drawn from IPI's Scientific Board and the winner receives a cash prize. - An important element of our activity is the stationing of qualified agronomists in overseas countries. Our delegates are now established in the Mediterranean area and in Argentina, while joint missions in cooperation with the Potash and Phosphate Institute of the USA and Canada are active in South and East Asia and in Brazil. Agronomists from IPI or from IPI members pay frequent o'~erseas visits in order to foster cooperation with local research and agricultural establishments. - For the furtherance of our aims it has been necessary to form close links with other international organisations and I would mention especially in this connexion the Potash and Phosphate Institute in Canada and USA, the European Fertilizer Manufacturers' Association in Zurich and the Inter- national Fertilizer Association in Paris; we also cooperate with many other organisations concerned with fertilizers. - The Food and Agriculture Organization of the United Nations in Rome occupies a special place in our international relations. In passing, it is good to hear that the Soviet Union is considering membership of FAO. IPI is proud to have been one of the three founders of the FAO Fertilizer Programme in 1961; this programme has been responsible for fertilizer ex- periments and education in more than 50 developing countries. Recent evaluation of over 100000 trials and demonstrations has shown that I kg NPK fertilizer produces 10 kg cereal grain. This report shows that much work at all levels in agricultural development and extension is needed if agricultural production is to be maintained or, better, improved. We are all quite aware that potassium is a most important factor in the im- provement of crop yield and quality and there can be not doubt that the sub- ject of this 22nd IPI Colloquium «Development of K-fertilizer recommenda- tions is of immediate interest and importance in very many countries, not least in the USSR. The opportunity presented for a stimulating exchange of ideas is now more important than ever for all partners in Europe. There is another reason for our happiness in visiting your country. We are today the guests of a country from which has emerged and expanded to a half of Europe the greatest peaceful revolution of the century. The fear engen- dered by the balance of atomic weapons has been replaced by the desire for peaceful co-existence and improved relations between different societies and cultures. It would, however, be a mistake to think that there are no more obstacles and that the morrow will be entirely cloudless. To change the orientation of your economy towards the free market is no easy task; such a transformation II cannot be realised overnight and we must be careful not to delude ourselves into believing that major problems can be solved by putting the difficulties to one side. It is better to proceed slowly, step by step rather than to hurry with the risk of producing chaos. We sincerely express our best wishes for your country and your authorities and for their future. A first step in mutual agreement has been achieved by the Soviet Union, but the problems of world politics concern not only the Soviet Union but Europe and the rest of the world. We are perfectly well aware that without the cooperation, mutual understanding and good wishes of the superpowers there is little hope of realising true peace. Even more than political change and its economic consequences, it seems to me that the development of inter- national confidence is the key to the door to developing an effective policy from which the whole world can benefit. Power-block policies belong to the past; ideologies no longer separate nations and peoples and there is real hope of fulfilling the dream of mankind to live and work in a truly peaceful society. Truly, we are faced with difficulties, especially in the East; the pressing problems of those who need help can be solved only by joint efforts and solidarity. It is fifty years since the Russian people shed their blood in the greatest tragedy the world has seen - the second World War. We must now seize the opportunity given us by the newly established confidence and show that we are ready to support your efforts. Your scientific effort in striving for the better utilisation of your soils to the end that your people's food needs can be met is very important. Liberty may be the first right of human society but this cannot be complete unless the life offered to mankind is worthwile. We are well aware that the world is changing around us and that day by day the work of the research workers is more and more important for our future. In your work as scientists and managers, you offer a great service to all humanity. Our Institute offers sincere thanks to you and acknowledges the contribution of your scientists. 12 Welcome Address Mr. A. Podlesny, Director General of Byeloruskali, Soligorsk, USSR Allow me to offer to you, participants in the 22nd Colloquium of the Interna- tional Potash Institute, a cordial welcome on Byelorussian soil. It is a great honour for us that such a representative international scientific meeting is taking place here in Soligorsk. We shall do everything in our power to make your stay at the Byeloruskali Combine useful and enjoyable. Our enterprise, and no less the city of Soligorsk itself is based on the Starobin potash deposit which was discovered in 1949. The first detachment of builders came on site here in July 1958 at which time the area was farming land. The birthday of the city of Soligorsk was 10th August in the same year, when the first stone was laid. The building of the city and the potash installa- tion proceeded hand-in-hand and the first stage of the potash operation was commissioned in December 1963. Construction of two more installations fol- lowed and the three were united in 1970 and in 1975 they were given the name < 14 Chairman of the 1st Session Dr. V V Prokoshev,, Laboratory of Potash Fertilizers, Institute of Fertilizers, Lenin- skyprospect 55, 117333 Moscow, USSR 1st Session Potassium Demand in Cropping Systems Development of Agricultural Yield Levels and Soil K-status in Eastern and Western Europe J. Breburda* Summary I n 1988, Soviet fertilizer output exceeded the production of the United States, France, West Germany, and Great Britain combined, and the USSR exports much of this production. But partly because of flaws in fertilizer quality, composition, and allocation, farm produc- tion continued to respond disappointingly. Imports played an important role, although some success in improving domestic production was registered. The use of chemical fer- tilizers and pesticides became increasingly controversial from the point of view of protecting the Soviet environment. Soviet fertilizer output reached 37.1 million tons in 1988, although the growth rate (2.2 percent above 1987) was the lowest in several decades. At the same time, the deliveries of fertilizers to farms declined by 1.1 percent (312000 tons), reducing the ratio of deliveries to fertilizer production. Underlying these trends were ecological concerns and pressures on farms to lower costs and on the country to gain hard currency by increasing fertilizer exports. Chemical fertilizer subsidies valued at 2.9 billion roubles were eliminated in 1988 and 1989. Overall, the use of fertilizers increased 2.6 times from 1970 to 1986. Application rates per hectare of sown area approached 118 kilograms in 1986, compared to the US rate of 92. The average Soviet application rate had first exceeded the US rate in 1983. The economic effectiveness of agricultural chemicals has been disappointing and their increased application has been subject to rapidly diminishing returns. Between 1975 and 1985, gross agricultural output in comparable prices increased by only 20 percent, while the volume of fertilizer deliveries and pesticide use increased by nearly 50 percent. The environmental imbalances which are observed in the West are accentuated by the economic imbalances of the USSR. Nitrogen fertilizer is produced in abundance, without phosphate and potassium complements, leading to low yields. The resulting vegetables, for instance, have nitrate levels 2-5 times the norm. The Soviet Union is peculiarly characterized by not replenishing the organic matter in its soils. According to Soviet scientists, the Soviet Union is experiencing a catastrophe in declining soil fertility. Some tentative conclusions on the recent performance of East European agriculture are as follows: Agricultural output in the 1971-1975 period grew at an average annual rate of 3.9 percent for the whole region, or more than double the rate for the previous five years. In 1976-1980, the average annual rate of growth was only 1.6 percent, and in 1981-1985 it was 1.4 percent. 1986 was a very good year with a 4.1 percent growth. In gross and net product the best results in 1975-1980 were achieved in Hungary, Romania, and Yugoslavia, * Prof. Dr. J Breburda, Institute of Soil Science, Justus-Liebig-University, Otto-Behaghel- strasse 10/D, D-63 Giessen, Federal Republic of Germany 17 the worst in Bulgaria and Poland. In 1981-1986 the best results were obtained in Poland and Romania. Progress and mechanization of agriculture and use of fertilizers has been good in Eastern Europe, but its level, except in Czechoslovakia and the GDR, is still signifi- cantly behind that of Western Europe. Overall, the yields were still substantially below those of the Federal Republic of Germany. 1. Present situation, aims and problems of Soviet agriculture The most important tasks of Soviet agriculture are to ensure reliable supplies of foodstuffs to the population and of raw materials to industry. The Twelfth Five Year Plan targets for agriculture were derived from the 1982 foodstuffs programme. Table I below shows the plan figures for certain products and the actual results achieved for the period 1981-1990. Table I. Soviet agriculture plan targets and actual production for the period 1981-1990 Product 1981-1985' 1986-1990' 1990 (million tonnes) Plan Actual Plan2 Actual 3 Plan4 Crop products (gross): Grain 238-243 180.3 250-255 205.5 250-255 Sugar beet 100 76.4 102-103 85.9 92-95 Potatoes 87-89 78.4 90-92 75.3 90-92 Raw cotton 9.2-9.3 8.3 - 8.3 9.1-9.4 Sunflower seed 6.7 4.9 7.2-7.5 5.9 7.0-7.1 Vegetables and waterfruits 33-34 29.25 37-39 29.4 40-42 Fruits and grapes 18.5-20 17.5 24-26 15.5 - Animal products: Meat (carcass wt.) 17-17.5 16.2 20-20.5 18.0 21 Milk 97-99 94.6 104-106 104.1 106-110 Eggs (billion) 72 74.4 78-79 82.7 80-82 Notes: I. Annual average 2. Plan according to the foodstuffs programme of 24/5/82 3. Annual average for 1986-1988 4. Main guidelines of the XXVIIth Party Congress 1986 5. Excluding waterfruits Sources: Kapustyan [1986]; Norodnoe Khozyaistvo SSSR za 70 let [1987]; Sel'skaya Zhizn' [1988]; Stepanov [1983] It can be seen from the table that actual agricultural production remains far below the Plan figures, especially in crop production. As early as 1986 in the main guidelines of the XXVIIth Party Congress the Plan targets were reduced, in that the average production volumes that had originally been planned for the period 1981-1990 were not to be fully achieved until 1990 (Gitz-Coenenberg et aL [1987]; Kapustyan [19861). Whether they will actually 18 succeed even in 1990 must remain a matter of doubt in view of the large dis- crepancies between planned and actual production. The difficulties now appearing everywhere in agriculture and associated branches of the economy, evidence of which can be found in many scientific publications and policy documents, would argue against the likelihood of the targets being achieved (especially in crop production). Soviet agriculture has to cope with complex physical and climatic condi- tions. The weather cannot be blamed as the sole cause of the unsatisfactory development of the agricultural sector however, because there are still some agricultural units (kolkhoz and sovkhoz) that have achieved significant in- creases in yields despite unfavourable weather conditions (Klimbowskii et al. [19881). The yields of wheat, mean annual precipitations and temperatures in Ukraine and Kazakhstan are shown in Figures 5.1. and 5.2. These improve- ments are based on using production techniques and organization appropriate to the soil and weather conditions. What is more serious, there has been evi- dence of increasing soil damage over the last few years. The reduction of the natural fertility of the soil due to soil erosion, insufficient organic fertilizing, soil compaction and increasing salinity in irrigated areas are all reflections of this trend. A further reduction of soil fertility is also to be expected over the next few years, for according to Soviet reports the impoverishment of the soil in humus, which is a most important component and determines to a large extent the fertility of the soil, is reaching alarming proportions. In many cases the addition of nutrients through organic and mineral fertilizers is not enough to compensate for the increased nutrient removal caused by the culti- vation of more demanding crops. The resulting nutrient deficiency is made up by nutrient delivery from the soil, which necessarily leads to nutrient im- poverishment and a reduction in the organic matter reserves of the soil (Kuz'michev [1983/). The continuing reduction of soil fertility through more intensive use is certainly an important problem at present, but once again it is not the main reason for the unsatisfactory trend in crop production. An increasingly negative influence on agricultural production is being exercised by mistakes in cultivation methods. There are problems with the choice of varieties, which is closely connected with the limited and qualitatively poor seed supply. What is more, the pattern of crop rotation does not always follow the principle of a change of crop type taking account of the existing condition of the soil and what it can actually take. Crop rotation damage results, with the frequent appearance of fungal diseases, plant pests and weed infestation. Under these conditions the effect of increased fertilizer application is cor- respondingly reduced. In addition, fertilizer delivery, storage and distribution are all problem areas. In this connection it is interesting to turn to the ( 19 of farm machinery. On the other hand, the greater power of the agricultural machinery stock can lead to soil damage through the improper use of tractors and implements (Ryabov [1988]). The cultivation problems outlined above are the main reasons for the un- satisfactory trend in crop production, all the more so when they mutually reinforce one another or have a combined effect and magnify the influence of meteorological risk factors in crop farming. Furthermore, there are a number of factors that have a negative effect on production under certain circumstances only, such as: - Melioration and irrigation associated with excessively high investment in relation to the expected yield, appearance of salinization in irrigated areas; - Fall in the workforce; - Work organization and interest, etc. Lastly, attention should be drawn to the losses of agricultural products be- tween the farm and the consumer, the reduction of which is an obvious way to increase the supply of foodstuffs. If these losses were eliminated, deliveries to consumers could be increased by as much as 20 percent, and even 30 percent in the case of certain products. The expenditure involved in eliminating losses is less than the cost of a corresponding increase in production by a factor of 2 to 3. The experience of the 1987 grain harvest (over 210 million tonnes), for which the weather conditions were bad, showed that the grain drying and cleaning capacity was inadequate. Although at the beginning of the harvest there was a total available capacity of over 1 million tonnes per hour, substan- tial quantities of grain had to be stored in the open before processing. The only remedy here would be storage and processing facilities in the immediate proximity of the grain fields (Deeva [1987]). 1.1 Land use for agriculture and forestry in the Soviet Union The total area of the Soviet Union is about 22.3 million square kilometres or 2.2 billion hectares. Topographical and climatic factors have a lasting in- fluence on land use however, especially in the fields of agriculture and forestry. It is well known that more or less risk-free agricultural activity is possible only in the «agricultural triangle of the Soviet Union, an imaginary triangle whose sides join the Baltic, the Black Sea and Lake Baikal. In 1986, the total agricultural area was 559 million hectares, amounting to 25.1 percent of the area of the country. Of the total agricultural area, 227.4 million hectares (40.7 percent) was arable land and 326.5 hectares (58.4percent) grassland, of which 33.7 million hectares (10.3 percent) meadows and 292.8 million hectares (89.7 percent) pastures. The latest statis- tics available for forests date from 1983, when there were 810.9 million hectares covering 36.4 percent of the area of the Soviet Union. [For purposes of com- parison, the corresponding figures for the United States for 1985 were: agricul- tural area 431.4 million hectares (47.1 percent of the total area of the United 20 States), of which arable land (including gardens, fallow land and perennial crops) 189.9 million hectares (44.0 percent of the agricultural area), forests 265.2 million hectares (28.9 percent of the total area of the United States).] While the size of arable land has stabilized at 227 million hectares, the share of low fertility soils has risen. Over 152 million hectares, or two-thirds of total arable land, have experienced reduced fertility owing to water and wind ero- sion. For example, 17.7 million hectares of land in the Ukraine, or about half of all Ukrainian arable land, are affected by erosion. In the last 15 years alone, the area of eroded soils increased by 55 million hectares. The academician A. Kashtanov stated that today erosion reduces yields by 30 percent and is the main cause of harvest losses. Gully formation is increasing by 100000 hectares annually, and close to half of available water and soil nutrients are captured by weeds. As a consequence of natural salinization and improper irrigation, 105 mil- lion hectares of agricultural land in the southern USSR have degenerated into cracked solonetz (salinized soils), which cover extensive territories in Central Asia, the Lower Volga Basin, Northern Caucasus, Western Siberia, Southern Urals, and other regions. This problem can be alleviated through gypsum treatments, the leaching of phytotoxic salts, soil reclamation, drainage, and manuring for improving soil structure and increasing permeability. However, gypsum treatment and reclamation are insufficient and are carried out on an area less than the size of the annual increase in salinized area, while larger amounts of water than are available are needed to accomplish the necessary salt leaching. Also, more than 51 million hectares of arable land in the USSR consist of acid soils. Lime treatment should be provided on 17 million hectares annually, with 160 million tons of lime materials. Yet, actual deliveries are only one-third that level. 1.2 Impact of cultivation on organic matter content and humus supply The importance of organic matter, with its many soil functions, for the fertility of agricultural areas, and the related question of humus supply has been very thoroughly discussed in the Soviet technical literature. For some time now there has been increasingly extensive research in the Soviet Union into the trend in humus content over the years. It cannot be denied that the present content of Soviet soils, in particular in the black earth regions, is causing considerable concern. The causes of the substantial loss of humus in arable areas are firstly the intensive, regular mixing and airing of the soil through tillage, and second a relatively limited return of organic matter through stubble, etc. As a result of cultivation, a substantial proportion of the original humus content of virgin land can be lost in this way (Kiryushin [1987]). Table 2 shows that the switch to arable use has led to a reduction of the humus content. In the European part of the USSR, a humus loss of up to 69 percent of the original humus content has been observed in the course of 100 years since 21 the grasslands (steppe) came under cultivation (Chesnyak et aL [1983]). Ac- cording to this estimate, the black soils of this region lost between about 51 and 270 tonnes of humus per hectare in the top 30 cm of soils in the 100 years between 1881 and 1981. But this does not at all mean that the loss of humus continues indefinitely with arable use. Depending on the farming methods used and hence the amount of organic matter returned to the soil in the form of stubble and roots from the crops grown, and the annual humus breakdown, a new humus level is established after a number of years or decades. Table 2. Humus content in virgin and arable soils (percent) Administrative area Use Soil depth (cm) and type of soil 0-20 20-40 0-40 North Kazakhstan Virgin soil 6.90 5.30 6.10 Carbonate black soil Arable: 20 years 5.34 4.37 4.80 Arable: 50 years 4.80 3.93 4.36 Semi-palatinate Virgin soil 4.70 3.30 4.00 Dark brown soil Arable: 67 years 2.40 1,70 2.00 Karaganda Fallow land 2.40 2.50 2.40 Dark brown soil Arable: 50 years 2.60 2.40 2.50 Aktjubinsk Virgin soil 2.70 1.60 2.10 Dark brown soil Arable: 18 years 1.70 1.60 1.60 Urals Virgin soil 3.18 2.65 2.91 Dark brown soil Arable: 14 years 2.42 2.24 2.33 Arable: 18 years 2.25 2.02 2.13 Alma-Ata Virgin soil 1.38 0.81 1.10 Grey soil Arable: 15 years 1.07 0.81 0.94 Arable: 20 years 1.02 0.77 0.89 Source: Yumagulova [19861 Intensive cultivation with increasing yields per hectare but without proper compensatory soil management measures, however, leads to further humus loss. The Soviet black earth regions, with the most fertile soil in the world, were over-worked in this way for many years. These black soils, which cover an area of 190 million hectares in the Soviet Union of which, 100 million hectares are arable land, form the backbone of Soviet agriculture, especially crop production. When this land was first cultivated it was assumed that its natural fertility was inexhaustible and that conservation measures such as or- ganic fertilizing and manuring were not necessary. This has resulted not only in falling humus content in the black soils, but also structural damage and increasing soil erosion, and hence low rates of increase in the yields of impor- tant crops, particularly over the last 20 years (Kovda [1987]). The analysis of copious data based on soil surveys and assessment of farming methods in sovkhoz and kolkhoz has revealed substantial humus loss in the arable areas of all regions of the Soviet Union as a result of 20 to 25 years of intensive farming. In the non-black earth zones, the humus stock declined over this period'at the rate of 0.5 to 0.7 tonnes per hectare per year. In the central black 22 earth zone and the North Caucasus, losses of 0.6 to I t/ha per year and even more have been recorded. In the RSFSR, an annual average humus im- poverishment of 0.6 t/ha is observed. As a result of 20 to 30 years of agricul- tural activity, the humus content of half the arable land of the Lithuanian SSR on the sandy derno-podzols has fallen from 1.9 to 1.8 percent and on the loamy soils from 2.16 to 1.8 percent (Popov [1988]). An effective way of providing humus is the application of stable manure or compost, but average nationwide use of them is only about 5 tons per hec- tare, a figure which obscures the fact that, while some technologically ad- vanced farms apply from 20 to 100 tons per hectare, the great majority of farms have to make do with only a few kilograms. 1.3 Mineral balance and fertilizing In attempting to solve the present problems of Soviet agriculture, one of the factors to which great importance is attached is fertilizing. Fertilizer applica- tions in the Soviet Union have been steadily increasing in recent years, as shown in Table 3 below. Table 3. Production and deliveries of fertilizers to agriculture, USSR (1000 metric tons') Year Total Nitrogen Phosphate Ground Potash Trace phosphate elements rock Production 1966-70 average 10379 4210 2030 955 3177 7 1971-75 average 17877 7248 3451 1032 6138 8 1976-80 average 23328 9283 5300 828 7910 7 1981-85 average 29294 12573 6747 774 9192 8 1986 34737 15200 8540 788 10200 9 1987 36300 15700 8900 791 10900 9 19882 37100 16000 9100 801 11190 9 Deliveries 1966-70 average 8452 3520 1847 857 2221 7 1971-75 average 13802 6209 2978 904 3703 8 1976-80 average 18063 7632 4460 827 5137 7 1981-85 average 22156 9790 5766 774 5817 9 1986 26514 11475 7567 787 6677 8 1987 27412 11787 7800 764 7052 9 19882 27100 11700 7700 7873 6904 9 1 Nutrient weight basis. Nitrogen: 20.5% N, phosphates: 18.7% P20, ground phosphate rock: 19% P20, potash: 41.6% K20 2 Estimates except total and footnote 3 3 Sel'skaya zhizn', 1/14/89. 23 However, not all arable areas receive adequate supplies of mineral fertilizers. High levels of application (e.g. for cereals) are found on the farms of Byelorus- sia (165 kg/ha), the Baltic republics (up to 280 kg/ha) and the Moscow area (up to 300 kg/ha) among others. This has made it possible to increase grain yields in these areas. However, in many other areas very little mineral fertilizer is applied. In the Altai region and areas such as Volgograd, Saratov, Orenburg, Kurgan and Novosibirsk, applications amount to only 30 to 50 kg/ha (Vashchukov [1988). Overall, the use of mineral fertilizers increased 2.6 times from 1970 to 1986 (Table 4). Application rates per hectare of sown area approached 118 kilo- grams in 1986, compared to the US rate of 92. The average Soviet application rate first exceeded the US rate in 1983. Fertilizer application rates tend to be higher in Western Europe, where fertilizer is more effective because of the climate, than in the USSR (Tables 5.3. and 5.4.). Table 4. Application of mineral fertilizer to selected crops, USSR' Year Grain Corn Cotton Sugar Potatoes exclud. corn for grain beet Rate (kg NPK/ha): 1975 ...... 42 155 391 399 280 1980 ...... 51 215 417 438 274 1985 ...... 72 200 376 455 293 1986 ...... 86 226 390 443 304 1987 ...... 89 209 410 419 284 1988 ...... 89 207 395 420 274 Share fertilized (%): 1975 ...... 48 94 99 99 93 1980 ...... 57 95 94 99 93 1985 ...... 71 94 98 99 95 1986 ...... 73 97 99 99 95 1987 ...... 72 95 100 100 95 1988 ...... 73 94 96 99 93 1 Nutrient weight basis. Source: Vestnik stalisiki, various issues. The economic effectiveness of agricultural chemicals has been disappoint- ing and their increased application has been subject to rapidly diminishing returns. Lemeshev [1988aand b] points to situations in which greatly increased fertilizer application has not increased yields (e.g., sugar beet in Moldavia). Between 1975 and 1985, gross agricultural output in comparable prices increased by only 20 percent, while the volume of fertilizer deliveries and pesticide use increased by nearly 50 percent. 24 1.4 Nutrient content of Soviet soils The nutrient content of Soviet soils varies greatly. It has been estimated (Romanenko [1988]) that 33 percent of the land has too little phosphorus, 30 percent a medium content and 37 percent an optimum or high content. About 10 percent of Soviet arable land is deficient in potassium, while 90 percent has a medium or high content. The latest published data of December 1988 tend on the whole to confirm these figures (Tolstsousov [1988]). Table 5 shows the nutrient content of agricultural land broken down according to land use. The figures refer to 1986. The nutrient content situation in the individual republics of the Soviet Union is shown in Table 6. Table 5. Breakdown of Soviet agricultural land according to phosphorus and potassium content (percentage of total area) Land use Phosphorus Potassium Low Medium High Low Medium High Arable ...... 34.7 37.0 28.3 10.0 21.0 69.0 Meadows ...... 44.8 28.5 26.7 29.9 24.0 46.1 Pastures ...... 43.2 30.7 26.1 22.1 20.7 57.2 Perennial crops ...... 42.5 28.6 28.9 20.0 21.4 58.6 Source: Derzhavin el aL [1987] Table 6. Percentage of arable land with low phosphorus and potassium contents in the individual republics of the Soviet Union Republic Phosphorus Potassium USSR total ...... 34.7 10.0 RSFSR ...... 33.7 9.0 Arm enian SSR ...... 49.5 4.1 Azerbaydzhan SSR ...... 67.1 48.5 Byelorussian SSR ...... 15.5 18.4 Estonian SSR ...... 37.9' 40.3 Georgian SSR ...... 66.4 38.7' Kazakh SSR ...... 69.7 1.9 Khirghiz SSR ...... 23.8 13.6 Latvian SSR ...... 51.6 27.7 Lithuanian SSR ...... 71.6 46.6 M oldavian SSR ...... 51.7 0.1 Tadzhik SSR ...... 37.2 58.1 Turkm en SSR ...... 77.1 42.9 Ukrainian SSR ...... 16.1 6.6 Uzbek SSR ...... 70.6 28.9 I Data for 1985. Source: Derzhavin et AL 11987] 25 What is more, according to Chumachenko [1986], 72 percent of the arable land of the Soviet Union has less than 10 Mg P205/100 g of soil, whereby it is to be noted that a phosphorus content of less than 10.1-15.0 mg P205/100g of soil on carbonate chernozems, kashtanozems and serozems is insufficient to ensure average plant yields. 1.5 Problems in fertilizer use The possible reasons put forward by scientific circles for thb inadequate nutrient supply and the associated low crop yields in the Soviet Union may be summarized as follows: - Deliveries of fertilizers to Soviet agriculture are inadequate with respect to the crop production requirements. There are many complaints regarding the quality of the fertilizers (physical properties and chemical composition); - The present quantitative ratio of nitrogen, phosphorus and potassium of 1:0.7:0.6 in total production and fertilizer deliveries does not correspond to the needs of the soil or the requirements of crops, which is one reason for the low yields. The optimum nutrient ratio would be 1:1:0.85; - The lack of knowledge on the part of the managements of agricultural un- dertakings regarding the necessary quantity of fertilizers and the proper time to apply them for particular crops leads on the one hand to under- supply of the soil when too little is applied, and on the other to relatively high losses of potassium, magnesium and other nutrients through leaching when too much is applied. Not achieving optimum plant nutrition causes fungal diseases in particular and also has negative effects on the phytosani- tary condition of the soil and the plant stock; - The effectiveness of increasing fertilizer applications is limited because of humus losses, soil structure damage, weed growth in the fields, the appear- ance of fungal diseases and plant pests, the low proportion of high-yielding varieties and hybrids, and lastly through erosion. Higher fertilizer applica- tions therefore frequently turn out to be uneconomic in proportion to the yield obtained. In the Eighth, Ninth and Tenth Five Year Plan (1966-1980) fertilizer applications on sugar beet more than doubled, reaching 460 kg/ha of NPK, but the yield increased by an average of only 800 kg/ha (4 percent) (Romanenko [1986]; Markov [1987]). According to the decisions of the XXVIIth Party Congress of the CPSU, in order to enhance soil fertility in the Soviet Union, the ((chemicalization of agriculture)) will be continued at a higher level in the Twelfth Five Year Plan. It is planned to supply agriculture with 30-32 million tonnes of fertilizers and 440000 to 480000 tonnes of plant protection products in 1990. In assessing the influence of fertilizers on the soil and the natural environ- ment, the special role played by nitrogen must be stressed. Nitrogen is involved in processes in the soil different from those in which phosphorus and potas- 26 sium play a part. Unlike P and K which are fixed by the soil and can therefore even be applied as 2. East Europe In the last two decades, the agricultural sectors in most East European coun- tries have made tangible though uneven progress. This has taken place in the context of varying systems of management. In Poland and Yugoslavia, the ownership and management of farms continues overwhelmingly in private hands; less than one fourth of agricultural land in each country is in state and collective farms. In Hungary, the ( 27 From 1975 to 1986, the greatest increase in agricultural output was achieved by Romania with an increase of 40 percent, followed by Yugoslavia, Hungary, Czechoslovakia, the GDR, and Bulgaria, in descending order. Poland ex- perienced a 4 percent increase in output. Most of the East European countries made rapid progress toward increased useof fertilizers in recent years. Table 7 shows that by 1983-1986, consumption of fertilizers per unit of land was exceeding the West European level in Czechoslovakia, the GDR, and Hungary. Bulgarian and Polish consumption. per hectare were getting close to the level of Western Europe, and they were roughly at or above the average for Eastern Europe in the same period. The heavily increased application of fertilizers is already reflected in significantly increased yields in Eastern Europe, but the rationality of the increased use is not immediately evident. A calculation would require a «good set of rela- tive prices of production factors, other inputs, and agricultural products. Table 7. Consumption of commercial fertilizers per hectare of agricultural land kg N + P205 + K 20/ha Eastern Europe= 100 1973 1976 1979 1983 1973 1976 1979 1983 -76 -79 -82 -86' -76 -79 -82 -86' Bulgaria ...... 105 119 151 152 75 78 97 98 Czechoslovakia ...... 218 244 248 260 156 160 158 168 GDR ...... 287 273 270 250 205 179 172 161 Hungary ...... 201 221 225 226 143 145 144 146 Poland ...... 176 189 186 176 126 124 118 114 Romania ...... 69 87 91 90 49 57 58 58 Yugoslavia ...... 49 57 65 68 35 37 41 44 Eastern Europe ...... 140 152 157 155 100 100 100 100 Western Europe ...... 176 197 210 229 126 130 134 148 1 Data for 1986 are preliminary. Sources: Calculated from statistical yearbooks of respective countries and FAO yearbooks and monthly statistical bulletins, assuming commensurability of the pure nutrients per kilo- gram The adoption of high-yielding crop varieties helped to increase yields per unit of input in all the East European countries. Research on improvement of seeds has been stepped up by the agricultural research institutes, partly under the coordination of the CMEA (Council for Mutual Economic Aid) Permanent Commission on Agriculture. Irrigation and drainage of agricultural land on a large scale is increasing the productivity of land in all East European countries. The recent develop- ment in Eastern Europe of agro-industrial complexes is increasing the overall efficiency of labor use through local processing of agricultural products, em- ploying seasonally idle agricultural labor, and diffusing technical knowledge in rural areas. An international comparison of agricultural outputs shows that Eastern Europe as a whole accounted for about 53 percent as much output 28 as the USSR and about 39 percent as much as the USA in 1981-1985; the USA's output was about 37 percent larger than the USSR's. In per capita levels of agricultural output in this period, the USA ranks highest, followed by Hungary, the GDR, Bulgaria, Czechoslovakia, Romania, the USSR, Poland, and Yugoslavia, in descending order. The development of wheat and sugar beet yields in Eastern Europe is shown in Tables 5.5. and 5.6. With the purported trend toward rational use of resources in Eastern Europe, leaders there, as elsewhere, may wish to ponder the influence of sys- tems of management on productivity. Concern with agricultural efficiency has prompted improvements in motivation through higher producer prices, higher profit, more freedom of action, control of resources, and other per- sonal incentives. To emulate Hungarian success in agriculture, the govern- ments in other East European countries have laely indicated more favourable policies toward private farmers and owners of private plots. They have started to help private farmers directly with incentives to increase output and produc- tivity. It remains to be seen what further impact these new policies will have on East European agriculture. 3. Changes in agricultural practice in Western Europe During the Second World War, there was great pressure to increase food production. Since then, agriculture has intensified greatly, partially due to economic pressures and Government subsidies, and more recently because of support from the European Economic Community. Some results of this intensification and pressure for higher crop yields have been the greater use of inorganic fertilizers, pesticides, development of new methods of cultiva- tion, increasing monoculture, larger sizes of fields, precision drilling of some crops to a stand, burning of cereal straw and stubble after harvest, and short- term leys instead of old pasture for grass production. The development of wheat yields in the European Community is given in Table 5.7. As animal production has become increasingly specialised and intensified, the availability of animal wastes as organic sources of nitrogen (N), phospho- rus (P), and potassium (K) has diminished. Organic manures are bulky in terms of their nutrient content. In relation to their economic value as nutrient sources, they are also costly to transport when the animals are a long way from crops. Hence, the use of inorganic fertilizers, particularly as sources of nitrogen, has increased rapidly. The use of pesticides in agriculture has been increasing steadily during the last 4 decades. New synthetic herbicides and insecticides were introduced first, but more recently there has been increasing use of fungicides. More autumn planting and earlier drilling in autumn have been achieved partly by increased mechanization and also by streamlined cultivation sys- tems. Use of cultivators rather than ploughs, shallow cultivations (less than 10 cm), and zero tillage have all contributed towards more work in autumn. Straw burning is a necessary part of these streamlined systems, both through 29 time saving and through avoiding the loss of yield associated with surface and shallowly incorporated straw. If the practice of straw burning declines substantially, farmers will need to make extra overall investment in tractor power and in other machinery. Since 1980, there has been a return from reduced cultivation systems towards ploughing and deeper cultivation. This change was brought about partly because of grass weed problems, but also through the development of better cultivation equipment. An important requirement of mechanized agriculture is adequate control of water tables in slow draining land. Otherwise, the risk of low bearing capac- ities becomes unacceptable, and soil physical conditions and crop yield suffer. This interaction between water regime of land and mechanized agriculture has been a major factor in stimulating improved field drainage. High economic yields can only be obtained with an adequate nutrient supply. In subsistence and low input agriculture the farmer relies to a large extent on the natural nutrient reserves of the soil, that is, he is mining the soil. And in most cases he has to content himself with a low yield level. In highly productive modern agriculture fertilizers are an essential input in order to attain and maintain the economically necessary high yield levels. Fertilizer recommendations are based on soil test results and crop demand having in mind two objectives: raising and maintaining the nutrient level in the soil above a given threshold level and replacing the crop offtake. An example of officially recommended soil test levels and K fertilizer rates for southern Ger- many is given in Table 8. Table 8. Soil K (lactate extractable) indexing according to texture (sand <8% clay, loam 8-25% clay, clay loam 26-40% clay) and K recommendations, Sand Loam Clay loam Recommended Index (mg KzO/100 g) fertilizer rate A very low < 8 < 12 < 15 Of ftake+ 100 kg K/ha B low 8-12 12-25 15-30 Offtake+ 50 kg K/ha C optimum 13-25 26-40 31-50 Offtake D high 26-40 41-60 51-70 Offtake x 0.5 E very high >40 >60 >70 0 Source: Informationen for Pflanzenproduktion, Baden-Wurttemberg, Ministerium for l ndlichen Raum. Landwirtschaft u. Forsten, 1987. It is, however, not possible to generalize such recommendations because there is a number of modifying factors that affect the plant availability of K. The K level in a soil is only one of the factors that govern K supply to plant roots. Table 9 shows this very dramatically. The data are taken from a set of 41 field trials which were run for up to 20 years (Beringer and Orlovius [19851). At the beginning the soils were adjusted to increasing K levels (original 30 K level, I mg K/g clay, 1.5 mg/g clay') by applying the appropriate quantity of K fertilizer. In the following years the K-fertilizer rate was adjusted to offtake. There was no relationship between response and soil K level. This finding explains why so many opinions exist about the optimum soil K level (because usually only a limited number of data with a limited variation of yield determining factors is at the disposal of the respective researcher). Ap- parently contradictory results from field trials are reported showing either good (Jankovic and Nemdtlh [19741, Nemdth and Harrach[1974]) or no corre- lation (Schachtschabel[1985]) between soil K status and crop response, de- pending on whether the combination of yield determining factors in the trials being compared was similar or different. One has to live with the fact that no universally applicable threshold exists, and this is a challenge to research. Table 9. The effect of increasing the K-status of soils and then adjusting the K-fertilizer rates to offtake of the respective crop (rotation: 2 years cereals I year sugar beet) on yield (Beringer and Orlovius [1985]) Texture Soil Duration K in control Relative yield over control group (years) (mg K20/ K1 KL.5 K2.0 (076clay) 100 g) 5.1-12 Fluvisol 2 8 - 103.5 99.3 12.1-17 Luvisol I 7 101.8 93.2 100.4 17.1-25 Cambisol 2 8 98.2 101.8 99.2 Luvisol 10 29 - 101.5 104.8 25.1-35 Cambisol 2 7 98.2 99.8 99.5 Cambisol (gleyic) 17 63 - 102.3 104.8 > 35 Fluvisol 20 38 - 104.3 110.7 Cambisol (gleyic) 5 28 112.4 116.3 Therefore, a new approach taking into account all relevant site specific fac- tors and grouping fields with similar characteristics of soil, parent material, climate, etc. is called for to supersede the convential approach (Andres [1990]). Global balance sheets of removal and input for provinces or even countries as a basis for generalized fertilizer recommendations cannot do justice either to the requirements of the individual farmer striving for his livelihood, nor to the requirements of agriculture in general striving for optimum economic production. I A soil with 2007o clay would then have 30 rg/100 g of exchangeable K. Clay has proven to be an important modifier of K availability (Grimme et aL. [1971]). 31 4. References Abuzov, M.L: lspol'zovat' ovrazhno - balochnye zemli (Use of the erosion gorged and ravined soils). Zemledelie, No. 2, Moscow, 41-42 (1985) Andres, E.: Soil fertility data banks as a tool for site specific K recommendations. Proc. 22nd Coll. Int. Potash Inst. 291-305 (1990) Averin, L G. andBaturin, V Y: Khimizatsiyu zemledelia - na novye rubezhy (Chemicaliza- tion of agriculture - towards new horizons). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 6, Moscow, 3-6 (1981) Beringer, H. and Orlovius, K.: Aufbau und Nutzbarkeit von Vorraten an pflanzenverfog- baren Kalium in Ackerboden. VDLUFA Schriftenreihe 16. Kongrellband, 159-169 (1985) Bolling, L andSohne, IV: Der Bodendruck schwerer Ackerschlepper und Fahrzeuge. Land- technik No. 37, 54-57 (1982) Bondarev, A. G.: Fizicheskiye svoistva pochv kak teoreticheskaya osnova prognoza ikh up- lotneniya sel'skokhozyaistvennoi tekhnikoi (Physical soil properties as theoretical basis for a prognosis of agricultural techniques). Vliyanie sel'skokhozyaistvennoi tekhniki na pochv (Effects of agricultural techniques on the soil). Moscow, 3-9 (1981) Bondarev, A. G., Rusanov, V.A. and Poljak, A. Y.: Problema obostryaestsya (The problem gets worse). Zemledelie, No. 2, Moscow, 23-25 (1985) Breburda, J: Bodenerosion und Bodenerhaltung. Frankfurt am Main, 1983 Breburda, j, Jaehne, G., Kellner, P., Pospelova, G., Schinke, E. and Wodekin, K. E.: Die Stagnation der sowjetischen Agrarproduktion seit 1978. Osteuropa, No. 6, Stuttgart, 425-442 (1985) Brune, H.: Schadstoffeintrag in Boden dutch Industrie, Besiedlung, Verkehr und Land- bewirtschaftung. Bodenbewirtschaftung, Bodenfrucbtbarkeit, Bodenschutz, Kongress- band. Darmstadt 85-102 (1985) Buldei, V R. and Vosnyuk, S. T: Osushitel'nye melioratsii i okhrana prirody (Drainage and nature conservation). Lvov, 1987 Chesnyak, G. Y, Gavrilyuk, F Y and Krupenikov, L A.: Gumusovoe sostoyanie cher- nozemov (Humus content of the chernozems). Russkii chernozem - 100 let posle Dokuchaeva (Russian chernozems - 100 years after Dokuchaev). Moscow, 1983 Chumachenko, L M.: Ispol'zovanie syrykh fosfatov v zemledelii (Use of raw phosphates in arable farming). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 2, Moscow, 20-22 (1986) Deeva, V.: Vazhnyi istochnik popolneniya zernovykh resursov (Important resources for in- creasing the grain stock). Planovoe khozyaistvo (Planned economy), Moscow, 77-80, 1987 Derevyagin, VA.: Puti transformatsii organicheskogo veshchestva v pochvakh (Transfor- mation paths of organic substances in the soil). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 3, Moscow, 49-51 (1986) Derzhavin, L. M.: Soderzhanie v pochvakh sel'skokhozyaistvennych ugodii SSSR podvizh- nogo kaliya i effektivnost' kaliinych udobrenii (Content of the soils of the agricultural land of the USSR in available potassium and the effectiveness of potassium fertilizers). Vestnik sel'skokhozyaistvennoy nauki (Agricultural science journal), No. 7, Moscow, 30-37 (1987) Derzhavin, L. M., Florinshil, M. A., Pavlichina, A. V and Leonova, L N.: lzmenenie soder- zhaniya povizhnykh pitatel'nykh veshchestv v pakhotnykh pochvakh SSSR (Changes in the available nutrient content in the arable soils of the USSR). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 11, Moscow, 54-57 (1987) Gotz-Coenenberg, R., Hohmann, H. H. and Seidenstecher, G.: Sowjetische Wirtschafts- entwicklung 1980-1990: Bilanz und Perspectiven. Teil 11: Der 12. Funfjahrplan 1986-1990. Bericht des BlOst Nr. 34 (1987) 32 Gzovskii, V.: lspol'zovanie vozobnovlyaemykh prirodnykh resursov v stranakh SEV (The use of renewable natural resources in the CMEA countries). Voprosy ekonomiki (Eco- nomic issues), No. 5, 108-118 (1983) Jezhevshkii, A.: Die technische Umristung der Landwirtschaft. Internationale Zeitschrift der Landwirtschaft. No. 5, Berlin, 379-383 (1987) Kapustyan, L K.: Sel'skoe khozyaistvo v dvenadtsatoi pyatiletke (Agriculture in the Twelfth Five Year Plan). Sel'skoe khozyaistvo (Agriculture) No. 9, Moscow (1986) Khachaturov, T. Aufgaben bei der Intensivierung der Bodennutzung. Sowjetwissenschaft, No. I, 23-29 (1986) Kiryushin, V.J.: Upravlenie plodorodiem pochv v intensivnom zemledelii (Soil fertility in intensive arable farming). Zemledelie (Arable farming), No. 5, Moscow, 2-6 (1987) Klimbovskii, A. A.; Isaikin, L L and Moiseev, A. A.: Pogoda odna, a rezul'taty raznye (One weather, but different results). Zemledelie (Arable farming), No. 10, Moscow, 12-13 (1988) Kotlyarova, 0. G.: Razrobotka i vnedrenie kompleksa pochvozashichitnykh meropriyati v Central'no-chernozemnoi zone (Elaboration and introduction of complex soil- conserving measures in the central black earth areas). Pochvozashchitnoe zemledelie na skonakh (Soil conserving tillage on sloping land), Moscow, 112-119 (1983) Kovda, V.A.: Pochvennyi pokrov, ego ulushchenie, ispol'zovanie i okhrana (Topsoil, its improvement, use and conservation), Moscow, 1981 Kovda, VA.: Proshloe i budushchee chernozema (Past and future of the chernozems). Russkii chernozem - 100 let posle Dokuchaeva (Russian chernozems - 100 years after Dokuchaev), Moscow, 1983 Kovda, V A.: Kompleksnye melioratsii chernozemov v sel'skom khozyaistve SSSR (Com- plex melioration on the chernozems in USSR agriculture). Vestnik sel'skokhozyaistven- noy nauki (Agricultural science journal), No. 4, Moscow, 36-41 (1987) Kravchenko, V. L: Uplotnenie pochv mashinami (Soil compaction by machines), Alma Ata, 1986 Ksiniaris, V: lspol'zovanie vtorichnykh materialnykh resursov v narodnom khozyaistve (The use of recycled products in the economy). Planovoe khozyaistvo (Planned econ- omy), No. 6, Moscow, 38-44 (1979) Kuz'Mitsev, V.P: Osvoyenie i ispol'zovanie chernozemov i dinamika ikh proizvoditelnosti (The taking over and use of the chernozems and the dynamics of their productivity). Russkii chernozem - 100 let posle Dokuchaeva (Russian chernozems - 100 years after Dokuchaev), Moscow, 1983 Lemeshev, M.: Vestnik sel'skokhozyaistvennoi nauki, No. 6, 130-133 (1988a) Lemeshev, M.: Vestnik sel'skokhozyaistvennoi nauki, No. 10, 23 (1988b) Marisinkevich, G. J: Ispol'zovanie prirodnykh resursov i okhrana prirody (Use of natural resources and nature conservation), Minsk, 1985 Markov, V.A.: Na sluzhbe plodorodiya (In the service of fertility). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 4, Moscow, 77-80 (1987) Melua, R. A. and Kiver, V F.: Deistve i posledestvie uplotneniya (Effect and consequences of soil compaction). Zemledelie (Arable farming), No. 2, Moscov, 29-31 (1985) Mengel, K.: Ernalhrung und Stoffwechsel der Pflanze. Stuttgart (1984) Miller, R. F: Agrarpolitik von Breshnev his zu Tschernenko. Osteuropa, No. 6, Stuttgart, 405-424 (1985) Mineev, V G.: Optimizatsiya primeneniya udobrenii i ekologicheskie aspekty sovremen- nogo zemledeliya (Fertilizer optimization and ecological aspects of arable farming). Vestnik sel'skokhozyaistvennoy nauki (Agricultural science journal), No. 6, Moscow, 23-30 (1987) Mineev, V. G.: Okologische Probleme der Agrochemie. Internationale Zeiischrift der Land- wirtschaft (No. 3, Berlin, 209-212 (1988) 33 Morgun, FT and Shikula, N.K.: Pochvozashchitnoe bespluzhnoe zemledelie (Soil- conserving ploughless cultivation), Moscow, 1984 Murashko, A. L et at: Okhrana sel'skokhozyaistvennykh ugodii u okruzhayushchei sredy (The protection of agricultural areas and the environment), Minsk, 1984 NarodnoeKhozyaistvoSSSR za 70 let (The USSR economy over 70 years), Moscow, 1987 Narodnoe Khozyaistvo SSSR (The USSR economy) (various years), Moscow Nosko, V.S.: lzmenenie gumusnogo sostoyaniya chernozema tipichnogo pod vliyaniem udobrenii (Changes in the humus content of a typical chernozem under the influence of mineral fertilizing). Pochvovedenie, No. 5, Moscow 26-32 (1987) Popov, P.D.: Resursy organicheskikh udobrenii i balans gumusa (Organic fertilizer resources and humus balance). Khimizatsiya sel'skogo khozyaistva (Chemicalization of Agriculture), No. 5, Moscow, 5-8 (1988) Rat von Sachverstondigen fur Umweltfragen: Sondergutachten vUmweltprobleme der Landwirtschaf>), Bonn, 1985 Ryabov, E.L: Dorozhe zolota (Dearer than gold). Zemledelie (Arable farming), No. 9, Moscow, 25-29 (1988) Romanenko, G. A.: Aktual'nye voprosy razvitiya zemledeliya (Current issues in agricultural development). Zemledelie (Arable farming), No. 7, Moscow, 2-6 (1986) Romanenko, G.A.: Kak tvoe zdorov'e zemlya? (How are you, land?). Zemledelie (Arable farming), No. 6, Moscow, 21-25 (1988) Rozanov, B.G.: Rasshirennoe vosproizvodstvo pochvennogo plodorodiya (nekotorye teoreticheskie aspekty) [Further restoration of soil fertility (some theoretical aspects)]. Pochvovedenie, No. 2, Moscow, 5-9 (1987a) Rozanov, B. G.: Rasshirennoe vosproizvodstvo pochvennogo plodorodiya v usloviyakh in- tensifikatsii zemledeliya (Further restoration of soil fertility under conditions of inten- sification of cultivation). Vestnik sel'skokhozyaistvennoy nauki (Agricultural science journal), No. 7, Moscow, 25-30 (1987b) Rusanov, V.A.: Trebovaniya k tekhnikie (The need for technique). Zemledelie (Arable farm- ing), No. 7, Moscow, 20-23 (1987) Saiko, V.F and Tarariko, A. G.: Pochvozashchitnye sistemy zemledeliya na sklonovykh zemlyakh Ukrainy (Soil conserving tillage systems on sloping land in the Ukraine). Vest- nik sel'skokhozyaistvennoy nauki (Agricultural science journal), No. 4, Moscow, 19-28 (1986) Scheffer, F and Schachtschabel, P: Lchrbuch der Bodenkunde, Stuttgart (1982) Sel'Skaya Zhizn: Sotsial'no-ekonomicheskoe razvitie SSSR v 1988 godu (Social life (1988) Socio-economic development of the USSR in 1988], No. 18, 1-3 (1988) Sel'Skoe Khozyaistvo SSSR: Statisticheskii sbornik (USSR social economy - statistical handbook), Moscow (1971) Shikula, N. K., Morgun, F K. and Tarariko, A. G.: Pochvo-zashchitnoe zemledelie (Soil- conserving tillage), Kiev, 1983 Shil'nikov, L A. and Lebedeva, L. A.: lzvestkovanie pochv (Liming the soil), Moscow, 1987 Shil'Nikov, L A. and Meshchanov, V.N.: Probleme izvestkovanie pochv (Problems of lim- ing the soil). Khimizatsiya sel'skogo khozyaistva (Chemicalization of Agriculture), No. 5, Moscow, 2-5 (1988) Shin, 1 V.: Itogi raboty obedineniya 34 Statistisches Jahrbuch far die Bundesrepublik Deutschland. Federal Statistical Office, Wiesbaden (1964 and 1988) Siepanov, A. J.: Zadachi i puti realizatsii prodovol'stvennoi programmy (Tasks and ways of realizing the foodstuffs programme). Sel'skoe khozyaistvo (Agriculture) 9, Moscow (1983) Tarariko4 A. G.: Pochvozashchitnye meropriyatiya na sklonovykh zemlyakh Ukrainy (Soil conserving measures on sloping land in the Ukraine). Pochvozashchitnye zemledelie na sklonakh (Soil-conserving tillage on sloping land), Moscow, 189-199, 1983 Tolstousov, V.P: Kompleksnaya programma povysheniya plodorodiya pochv (Complex programme for increasing soil fertility). Zemledelie (Amble farming), No. 12, Moscow, 2-5 (1988) Umarov, M. U. and Kurvanlaev, R.: Povyshenie plodorodiya oroshaemykh pochv putem regulirovaniya ikh fizicheskikh svoistv (Increasing the soil fertility of irrigated land through regulating the physical properties of the soil). Tashkent, 1987 Vashchukov, L.: Die Entwicklung der Getreidewirtschaft in der UdSSR. Internationale Zeitschrift der Landwirtschaft, No. 5, Berlin, 376-378 (1988) Vashchukov, L.: Ulushchit sokhrannost' i ispol'zovanie zemelnykh i vodnykh resursov strany (Improvement of the conservation and use of soil and water resources). Vestnik statistiki (Statistical handbook), No. 7, Moscow, 39-44 (1988) Vashchukov, L. and Pusakov, V.: Die Entwicklung des Meliorationswesens in der UdSSR. Internationale Zeitschrift der Landwirtschaft, No. 4, Berlin, 303-305 (1986) Vasilev, N: Die Melioration als Mittel zur Leistungssteigerung in der Pflanzenproduktion. Internationale Zeitschrift der Landwirtschaft, No. 2, Berlin 90-95 (1988) Vorobev, G. Y.: Berech pochvu ot pereuplotneniya tekhnikoi (Protecting the soil from com- paction). Zemledelie (Arable farming), No. 9, Moscow, 15-17 (1987) Wadekin, K. E.: Sowjetische Landwirtschaft in der Stagnation. Osteuropa, No. 2, Stuttgart, 89-100 (1983) Yagodin, B. A.: Intensifikatsiya zemledeliya i sovremennye problemy agrokhimii (Intensifi- cation of arable farming and agrochemical problems). Khimiya v sel'skom khozyaistve (Chemistry in Agriculture), No. 4, Moscow, 2-5 (1987) Yumagulova, A. N.: Plodorodie nepolivnykh pochv Kazakhstana (Fertility of the rainless soils of Kazakhstan). Zemledelie (Arable farming), No. 3, Moscow, 27-28 (1986) 35 yieldsWinter t/ha)wheat u -4.5 .4.2 -3.9 -3.6 -3.3 -3.0 -2.7 Precipi- 2.4 tations Temp. -2.1 (mm) (0 C) .1.8 500-- 20 -1.5 Jan. 400- 16 -1.2 to 3 00 12 July 200 8 100 4 0 1956 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 Aug. 25 and 75 Sept. 25 Yields of winter wheat (t/ha) ..... Jan. to July precipitations (mm) Mean annual temperature (0 C) - Aug. and Sept. precipitations (mm) of preceding year Compiled from computer excerpts of "Deutsches Seewetteramto, Hamburg, 1989, and from N.CH. SSSR, volumes 1956-1988 5.1. Yields of winter wheat, mean annual precipitation and temperature in Ukraine (1956-1988) Summer wheat yields (t/ha) 1.92 -1.5 -1.4 -1.3 -1.2 -1.1 1.0 0.9 Precipi- 0.8 tations Temp. -0.7 (mm) (0 C) -0.6 200 A I_ _ .7 -0.5 7 -0.4 Jan- 120 to 80 •5 -0. July , 40 - 86 88 70 72 74 76 78 80 82 84 1956 58 60 62 64 66 68 151 My45 -- -_ - Yields of summer wheat (t/ha) -----Jan. to July precipitations (mm) Mean annual temperature (0 C) - May precipitations (mm) Compiled from computer excerpts of Deutsches Seewetteramt,", Hamburg, 1989, and from N.CH. SSSR, volumes 1956-1988 Kazakhstan (1956-1988) -J 5.2. Yields of summer wheat. mean annual precipitation and temperature in 5.3. Application of fertilizers to small grains in the USSR, 1975-1987 Year kg NPK/ha 1975 ...... 42 1976 ...... 47 1977 ...... 48 1978 ...... 5 1 1979 ...... 49 1980 ...... 51 1981 ...... 5 1 1982 ...... 54 1983 ...... 58 1984 ...... 65 1985 ...... 72 1986 ...... 86 1987 ...... 89 Source: Vestnik statistiki, various issues 5.4. Fertilizer application to small grains in the USSR and major Republics, 1981-1987 (kg NPK/ha) 1981 1982 1983 1984 1985 1986 1987 USSR ...... 51 54 58 65 72 86 89 RSFSR ...... 47 49 55 62 69 84 87 Ukraine ...... 76 79 85 94 106 126 136 Byelorussia ...... 219 217 213 217 221 247 265 Kazakhstan ...... 10 11 13 18 23 25 25 5.5. Development of wheat and sugar beet yield in Eastern Europe (/ha) Crop Country Average 1986 1987 1988 1981-85 W heat Bulgaria ...... 3.85 3.84 3.82 4.01 Czechoslovakia ...... 4.68 4.72 5.08 5.26 East Germany ...... 4.13 4.64 4.59 4.83 Hungary ...... 4.63 4.36 4.37 5.44 Poland ...... 3.29 3.70 3.72 3.43 USSR ...... 1.49 1.80 1.83 1.70 Yugoslavia ...... 3.46 3.60 3.62 4.19 Sugar beet Bulgaria ...... 22.5 20.3 18.8 16.1 Czechoslovakia ...... 34.8 35.9 35.4 33.5 East Germany ...... 29.4 34.6 34.7 23.4 Hungary ...... 38.9 36.2 36.1 39.3 Poland ...... 33.1 33.6 33.2 34.1 USSR ...... 21.8 23.3 26.6 26.1 Yugoslavia ...... 42.3 40.5 38.0 34.7 38 5.6. Development of wheat yield (I/ha) in Eastern Europe (without USSR) 0 1962-64 0 1972-74 1979 1980 1981 1982 1983 1984 1985 1.77 3.35 3.05 3.56 3.38 3.69 3.55 4.09 4.01 5.7. Development of wheat yield (t/ha) in the European Community 0 1960-64 0 1970-73 1980 1981 1982 1983 1984 1985 Fed.Rtp.of Germany 3.28 4.38 4.87 5.09 5.47 5.44 6.27 6.08 Netherlands 3.99 4.85 6.54 6.68 7.38 7.05 7.84 6.63 Average EC 2.57 3.45 4.35 4.30 4.60 4.50 5.60 5.06 39 The Position of K in Nutrient Balance Sheets of the Netherlands A. van Diest* Summary Traditionally, the Netherlands has been a country exporting agricultural products having relatively high monetary values and relatively low nutrient contents. Consequently, nutrient balance sheets have always been positive, in the sense that more nutrients are imported than exported. Excessive discrepancies in nutrient balance sheets started to develop when 30 years ago many smallholders became engaged in intensive livestock production, for which large quan- tities of feedstuffs needed to be imported from overseas countries. These smallholders live predominantly in the Eastern and Southern parts of the country where sandy soils, low in natural fertility, prevail. The younger soils in the West and North are mainly of marine origin, with excellent physical characteristics but also rather low in natural fertility. Consequently, fertilizer use has been relatively high since the 1930's. Presently, however, imports of NPK in feedstuff form exceed those in fertilizer form. Due to excessive application of liquid manure to the sandy soils of the smallholders, leading to environmental pollution, the government has set limits to the quantities of liquid manure allowed to be applied. The P20, contents of the manures are used as criteria for thequantities of manure that can be applied. Theobjective of such legislation is that eventu- ally the quantities of P applied in manure form will match the quantities of P withdrawn by the crops. Presently, the quantity of manure that can be applied to grassland is high enough to evoke the risk of hypomagnesaemia in cattle. In 1991, the maximum quantities of manure permitted to he applied will be reduced, leading to a substantial increase in the quantity of liquid manure a farmer can no longer use on his own soil. In pilot plants, surplus liquid manure is now being dried, with the intention to sell the resulting dried animal manure on foreign markets. High costs associated with the drying process created an interest in cheaper technology in which ammoniacal N is nitrified and the resulting NO, is subsequently denitrified. Phosphates can be precipitated with calcium. All potassium will be present in the remaining effluent which can most cheaply be disposed of by channelling it into the river Rhine and eventually into the North Sea. 1. Traditional farming in the Netherlands In the recent past, large discrepancies between the quantities of nutrients im- ported into and exported from the Netherlands have resulted in a surge of attention given to nutrient balance sheets for that country. Traditionally, the * Prof. Dr. A. van Dies, Department of Soil Science and Plant Nutrition, Agricultural University, Dreijenplein 10,6703 HB Wageningen, Netherlands; member of the Scientific Board of the International Potash Institute 41 Netherlands has been a major exporter of agricultural products, but the quan- tities of nutrients contained in the dairy products, vegetables and flowers exported were small in comparison with the nutrient quantities present in the raw materials imported from overseas, mainly grains, needed for the dairy production. Many of the dairy commodities were produced by peasants making a living on smallholdings usually not exceeding 5 ha in size. Most of these smallholders lived in the part of the country East and South of a diagonal that can be drawn across the country from the Northeast to the Southwest (Figure 1). , ,,.,, q ...... Figure 1. Map of the Netherlands, with the diagonal roughly separating the Pleistocene sandy soils from the Holocene clay loam and low-moor peat soils. 42 To the right of this diagonal we find predominantly fine-textured sandy deposits of Pleistocene origin, laid down by wind action during glacial and interglacial periods. To the left of the diagonal young marine deposits of Holocene origin are to be found, interspersed with areas of low-moor peat. Many of the marine soils were formed from sediments on the bottoms of inland lakes and of an inland sea. These were drained after the construction of dikes leading to the formation of so-called polders. Other marine soils were created after the construction of new dikes around young marine deposits laid down beyond the last built dike. Such deposits became reclaimable when they had reached an elevation making them fall dry during periods of low tide. 2. Natural fertility of Dutch soils A notable difference exists in natural fertility among soils on either side of the drawn diagonal. The sandy Pleistocene soils of aeolian origin are inher- ently poor in nutrients. Before the fertilizer era, permanent agriculture on such soils was possible only when manure, produced by pigs and cows foraging in forests and, later on, by sheep grazing on heather (Calluna vulgaris) - co- vered fields, was collected in pens and barns used for nightshelter, and was applied to small arable lots. In this manner, a patchwork of fertile « 43 soil-structural conditions. However, contrary to general belief, the natural fer- tility status of these soils is not exceptionally good. The organic matter contents shortly after reclamation lie around 3%, and the quantities of N annually released due to mineralization of soil organic N are too low to satisfy the needs of the crops, mainly cereals, potatoes and sugar beet. Likewise, the P contents of the young soils are inadequate to allow high yields to be obtained over a considerable period without inputs of fertilizer P. The situation is different for K. The mineral reserves of this nutrient are high enough to meet the K requirements of crops over a period of many years. This is shown in Figure 2. The results were obtained in an experiment conducted on the University Farm, in which since 1975 arable crops are grown with and without the commonly recommended quantities of NPK. For a period of around 10 years prior to 1975 the land had been managed by a governmental agency supervising the proper transition of lake-bottom sediment to polder soil. In this period occasional use was made of small quantities of fertilizers, but for all practical purposes it can be stated that at the start of the fertilization experiment, the soils contained only natural supplies of nutrients. tons of beets/ha Sugar beet 80,17 40 9979 Th8 tonsO [mo3o,ll 10+ I1:++ 2988 1 tons of tubers/ha n ' j 2e use 40s Figure 2. Sugar44, beet and potatoP, and yieldsK fertilizer on newly applications reclaimed sincepolder 1975. soil, as affected by 3. Soil testing in the Netherlands The results presented in Figure 2 make it clear that high yields can only be obtained when from the start use is made of N fertilizers. A need for sup- plemental fertilizer P appears to arise after 7 years for potato, a crop with a rather weak root system. No response to P was yet to be observed for sugar beet after 9 years, but a response became visible after 13 years of farming. Each year, soil samples are collected from each plot for determination of available quantities of P and K. The method used for P is one in which the soil is shaken with water, the so-called P. method. With this extraction proce- dure, information is obtained on the P-intensity factor, and not on the P- capacity factor. For the past 45 years, most Dutch soils have annually received larger quantities of fertilizer P than those withdrawn by the crops. The result has been that these soils have acquired considerable pools of labile soil P and consequently score high in a soil-testing procedure with emphasis laid on P-capacity factors. For high yields, not only high levels of potentially available P (capacity factor), but also of immediately available P (intensity factor) are necessary. Since Dutch soils vary more in P intensity than in P capacity, the currently employed water-extraction method supplies information primarily on the P intensity factor. It can be observed in Table I that in the soil of the zero-P plots the level of water-extractable P fell considerably in the 1974-1982 period, and has remained rather constant since then. The fact that only since 1986 sugar beet and potato have started to respond to added fertilizer P (Figure 2) casts doubt upon the usefulness of water-extractable P as index of P avail- ability in these newly reclaimed marine soils lacking a long P fertilization history. Table 1. Changes in water-extractable P and adjusted HCI-extractable K-values, averaged over all plots without and with annual P and/or K fertilizer applications, in a 15-year period of a long-term fertilization experiment Extractable quantities (K in mg K20 per 100 g soil, P in mg P205 per I soil) 1974 1982 1985 1889 Fertilizer treatments P K P K P K P K -P, -K ...... 26 26 15 25 13 21 13 18 +P, +K ...... 26 26 23 28 29 23 33 20 Dutch soils, in general being light-textured, do not possess high capacities for retaining potentially available K. For that reason it cannot be assumed, as in the case of P, that long-term K fertilization practices have led to large pools of potentially available K. Consequently, a soil-testing method is used which is intended to supply information on both immediately and potentially available K. This is thought to be accomplished by shaking soil with a 45 0.1-molar HCI solution, expected to extract not only the exchangeable K, but also a fraction of the non-exchangeable K. In practice, it has been observed that the K-HCI values thus obtained are not the most suitable indices of plant-available K. The correlation between K uptake by crops and estimated soil K availability can be improved, when account is taken of the influences of soil texture and of pH on the availability of soil K. It was experienced that the availability of HCI-extractable K to crops decreases with increasing content of fine soil particles. In the Netherlands it is customary to characterize soils texturally by their percentage of particles smaller than 16 gm. In the formula b K-value = K-HCI value x 0.15 pH (KCI) - 0.05 the numerical value of the correction factor b is determined by the percentage of soil consisting of particles < 16 pm. The higher this percentage, the lower is the value of b which attains a value of unity when the percentage of soil particles <16 tm is 35. With this empirically determined correction factor, account is given of the practical experience that a certain quantity of extracta- ble K is to be valued lower in terms of plant-available K in heavy-textured than in light-textured soils. This experience was obtained with the use of hundreds of field experiments conducted in the period 1945-1965. In the above formula it is also expressed that soil pH affects the relationship between extractable K and plant-available K. Again here, by empirical means it was observed that a certain quantity of extractable K is to be valued higher in terms of plant-available K in acid than in alkaline soils. One of the causes underlying this relationship appears to be the nature of the dominant cation in exchangeable position, complementary to the cation whose availability is under consideration. The phenomenon, which is now known as the complementary-ion effect (Black [1967]) was first observed and described by Russian investigators before the second world war (Jarusov [1937], Ratner [19381). In alkaline soils, Ca as usually dominant cation on exchange com- plexes is held less tightly than is Al as usually dominant cation on exchange complexes of acid soils. The complementary-ion principle dictates that the ease of release of e.g. K from an exchange complex increases with increasing bonding strength of the complementary exchangeable cations. In practical terms, a certain quantity of exchangeable K is more easily released to the soil solution in an acid than in an alkaline soil, assuming types of clay mineral and CEC values to be identical. The data of Table I show that after 15 years of farming without K applica- tion the quantities of available soil K have not yet fallen to values at which yield depressions occur, in spite of the fact that yield levels on the young ma- rine soils are high and K withdrawal is therefore substantial. The conclusion can therefore be drawn that marine sediments converted into polder soil in the Netherlands can be valued very highly on account of their excellent physi- cal characteristics and their high levels of availability of Ca, Mg, K and trace elements. They are, however, naturally low in available N and P. 46 The low-moor peat soils in the western part of the country are particularly known for their high level of potentially available N. When farming started on these soils about 1000 years ago, the peat deposit was around 10 m thick. Generally, 8 meters are remaining now, which means that per year 2 mm disap- pear due to mineralization. In this process, annually 150 kg N are released for absorption by the grass predominantly grown as permanent pasture on these peat soils. 4. Fertilizer use in the Netherlands The data presented in Table 2 show that before the second world war the con- sumption of potash fertilizers exceeded that of N- and P fertilizers. After 1950, K was surpassed by N. The annual consumption of K stabilized around 60 kg K2 0 per ha, whereas the N consumption climbed to values over 200 kg/ha. Although such values may seem high in comparison with those pertaining to Eastern European countries, they nevertheless represent only fractions of the total quantities of nutrients annually entering Dutch agricultural soils. Increasing percentages'of total imports of nutrients enter the country in the form of feedstuffs, as in shown in Table 3. Table 2. Consumption of N, P205 and K20 in fertilizer form in the Netherlands in the period 1920-1988 Year Quantities in 1000 tons Quantities, kg per ha agric. land N P20 5 K20 N P205 KO 1920 ...... 26 49 54 12 23 26 1930 ...... 45 103 99 19 45 43 1938 ...... 93 98 130 40 41 55 1950 ...... 155 120 156 65 51 67 1960 ...... 224 112 138 97 48 59 1970 ...... 405 t10 135 193 51 63 1978 ...... 443 81 107 211 39 52 1983 ...... 467 87 117 233 37 52 1985 ...... 505 89 125 250 43 62 1988 ...... 458 80 98 227 40 49 Table 3. Quantities of feedstuffs and feedstuff-K imported in the Netherlands Total quantity Cassava as % Total K % of K of feedstuffs of total imported in imported imported (1000 tons) feedstuffs imported feedstuffs (1000 tons) in cassava 1970 1988 1970 1988 1970 1988 1970 1988 7000 14874 7.507o 19% 61 167 14% 2907o 47 5. The use of feedstuffs in the Netherlands It was mentioned earlier that around 25 years ago many smallholders shifted from mixed farming to more intensified forms of farming usually consisting of intensive livestock production. With the aid of bank loans they started to build sheds for housing pigs, calves and poultry. In this manner, the origi- nally small farming enterprises were converted into meat factories. The cereals, fodder beets and mangolds these smallholders used to grow had to be supplemented with feedstuffs mainly imported from overseas countries. The major feedstuffs are soybean press cake, cassava meal, maize gluten, grains (primarily maize) and citrus pulp. The data in Table 3 show that the quantity of feedstuffs imported has more than doubled since 1970. One of the important developments of the last 20 years has been the percentage rise in the import of cassava, a feedstuff with a relatively high K content. It can be seen that presently almost one-third of all K in feedstuffs enters the country in the form of cassava, imported almost exclusively from Thailand. The 167 000 tons of K imported in feed- stuffs is a quantity substantially exceeding the 97 800 tons of K imported in fertilizer form in 1988, as mentioned in Table 1. For the year 1983 it was calculated (van Diest [1986/) that 447 000 tons of K entered Dutch soils in the form of animal excreta, which is a quantity still considerably higher than the combination of imported feedstuff-K and fertilizer-K which for that year amounted to 258 000 tons of K. In other words, the K reserves in Dutch soils are still the major suppliers of K to the growth of crops and, indirectly, to the livestock which consumes roughage mainly in the form of grass, hay and silage maize, grown inside the Netherlands. 6. Potassium balance sheets It can therefore be stated that the bulk of K is involved in a nutrient cycle in which the soils perform a buffer function, in the sense that they are the major suppliers of K to crops and are the largest storehouse of K annually returning to the soil, primarily in the form of animal manure. The exceptional situation for the Netherlands lies in the fact that annually such large quantities of K are added to the storehouse, in the form of feedstuff- and fertilizer-K. In itself, this situation would not be a matter of major concern, if it were counterbalanced by large exports of K in the form of agricultural commodities, or otherwise. The Netherlands is a major ex- porter of agricultural products, with a 1986 value of 19.2x 109 US $ to be compared with a value of 12.2 x 109 US $ for imported agricultural products, among which animal feedstuffs take an important position (F.A.O. [1988]). 9 Three major contributors to the total export value are meat (3.2x l0 US $), dairy products (3.0× I09 US $) and vegetables and fruits (3.1 x 10' US $). As mentioned before, especially the meat is produced by smallholders who them- selves raise only a minor portion of the feedstuffs needed and therefore de- pend heavily on imported materials. 48 The K concentrations in the exported animal products are generally low, so that a very wide discrepancy exists between quantities of nutrients imported and exported. The quantity of K in agricultural export products for 1983 was calculated to be 23 000 tons, which value is more than 10 times smaller than that of K in import products, for that year being 258000 tons. As a summary, the following data are brought together for the year 1983 (in 1000 tons): 1. K im ported in fertilizers: ...... 106 2. K im ported in feedstuffs: ...... 152 3. K applied in various animal manure forms: ...... 447 4. K exported in agricultural commodities: ...... 23 K in item no. I is used primarily for the production of arable and horticultural crops, mainly grains, potato, sugar beet, vegetables and flowers. The quanti- ties of K applied are more or less in balance with the quantities withdrawn by the crops. The K in produce consumed by humans eventually enters the surface water via waste disposal treatment plants. The K in crop residues may be returned to the soil or may serve as animal feed. K in item no. 2 primarily enters livestock, but here a balance is far out of sight. The average Dutch cow annually produces over 6000 liters of milk containing 9 kg K, a quantity representing only a few per cents of the annual K intake. Practically all K is to be found back in the excreta, and will thus make a contribution to the K pool of item no. 3. 7. Use of fertilizer K and animal-manure K One way to stem the flow of nutrients entering the country could be the replacement of fertilizers by animal manures. It has already been known for a long time that the efficiencies of utilization by crops of K applied in animal excreta and in fertilizer are similar. However, broadcast application of liquid manure gives rise to extensive volatilization of'NH 3 from the manure and for that reason in the future liquid manure is to be injected, not only on arable land but also on permanent pastures. For the latter, the injection unit consists of a pipe set behind a duckfoot and a disc coulter. The liquid manure, flowing from a tank drawn by a tractor, is thus dispensed into narrow furrows at a depth of 15-20 cm, with usually a 50 cm distance between the furrows. For P it has been shown that such a row-application technique for liquid- manure leads to an uneven distribution in the soil underneath the sod and consequently to uneven P nutrition of the grass and to yield decline. In Table 4, results are presented of field trials conducted in the Netherlands on the influences of similar doses of K in fertilizer- and in liquid-manure form on dry-matter production and K percentage of grass. When liquid ma- nure is injected, this is done once per growing season, in early spring. In the experiment reported (van Dijk [1989]), the recommended fertilizer K was split over 6 dressings, each one applied immediately after the grass was cut. 49 Table 4. Dry-matter production and K content of grass with the recommended quantity of K applied either in fertilizer form or in liquid-manure form. Results presented are means of 7 experimental fields Control K broadcast in K injected as (no K applied) split dressings basal dressing as fertilizer* in liquid manure dry-matter production, tons per ha 11.2 13.5 13.1 K0o in dry-matter cut I ...... 2.2 3.5 3.2 cut 2 ...... 2.1 3.8 3.3 cut 3 ...... 1.7 3.6 2.6 cut 4 ...... 1.5 3.6 2.2 cut 5 ...... 1.3 3.2 1.9 cut 6 ...... 1.5 3.3 1.8 * average total application approx. 250 kg K per ha. It can be seen that a single basal application of liquid-manure K results in a gradual decline in K content of the grass with slight consequences for the production of dry matter. The conclusion of this experiment can be that liquid-manure injection has less negative consequences for the K nutrition than it has for the P nutrition of grass. However, the damage that injection does to the sod makes it undesirable to inject more than once per growing season, which implies that nutrient contents of the grass may vary widely within one growing season. A general reluctance noticeable among arable farmers to use liquid manure is based on a number of inherent characteristics of liquid manure, such as - the liquid form, requiring investments for new storage facilities and dispens- ing equipment, - the possible presence of weed seeds, nematodes and other deleterious or- ganisms, - the variability in nutrient content of the manure, - the impossibility of applying nutrients at any desired moment during the growing season, - the impossibility of applying only one nutrient at a time, - the repulsive odors. 8. Legislation aimed at environmental protection A lack of interest among arable farmers for the large surpluses of liquid manure in the Netherlands has resulted in irrresponsibly large quantities of manure applied by the intensive livestock producers to the few hectares of land they possess. Such injudicious use of manure has led to eutrophication of surface water and pollution of deep drinking-water reservoirs. To prevent further deterioration of environmental quality, the government has set limits to the quantities of liquid manure that can be applied annually. The P20 5 50 contents of the manures were chosen as criteria for the maximally allowable quantities of manure to be used. The overall objective of such legislation is to return to a situation in which the quantities of P applied in manure form will match the quantities of P withdrawn by the crops grown. Since 1987 the maximum quantities of P20 5 annually allowed to be applied per ha in manure form are: arable land, excluding maize land: 125 kg permanent pasture: 250 kg land to be used for maize culture: 350 kg Potassium as a pollutant certainly does not attract as much attention as given to phosphate and nitrogen. It is not a serious cause of eutrophication and does not constitute a hazard to the health of human beings. For livestock the matter is different. In the recent past hypomagnesaemia has been a serious problem in the Netherlands. Although it was shown to be a multiple-cause disorder (Kemp [19821), hypomagnesaemia can result from a disturbed K/Mg ratio in the roughage consumed by cattle. An excessive ratio can occur in grass, when 250 kg P20 5 is applied per ha in the form of liquid cattle manure con- taining 0.44% N, 0.18% P20 5 and 0.55% K20, which can be viewed as representative concentrations. In such a situation, 770 kg K20 and 620 kg N are applied per ha along with 250 kg P20 5. These quantities of K and N can both be considered as exceedingly large especially when applied annually over a number of years. They can endanger the health of cattle. In 1991, the maximum quantities of P20 5 permitted to be applied annually per ha will be reduced to: arable land, excluding maize land: 125 kg permanent pasture: 200 kg maize land: 250 kg Although still far out of step with the quantities of P annually withdrawn by crops (approx. 70 kg P20 5 by arable crops and 80 kg P20 5 by grass), the new P20 5 quantities allowed to be applied will further increase the already existing surpluses of liquid manure that can no longer be applied by livestock producers to their own land. The total quantities of animal manures produced in 1988 are presented in Table 5 (Centraal Bureau voor de Statistiek [1989J). Table 5. Quantities of animal manures produced in the Netherlands in 1988 Cow manure ...... 60 x 106 tons of which on pastures ...... 32.5 x 106 tons in stables ...... 27.5x 106 tons Calf manure ...... 4 x 106 tons Pig m anure ...... 21 x 106 tons Poultry manure ...... 2 x 106 tons Total ...... 87 x 106 tons Per ha agric land ...... 43 tons 51 Especially for pig production it has become customary to add extra phosphate to the feed. Recent research has shown, however, that a proper growth rate of pigs can be sustained with less phosphate added to the feed than was prac- ticed in the past. One consequence of a reduction in the use of P additives is that per unit P 20 5 more liquid manure and more N and K are applied. Thus, a reduction in the permissible rate of P 20 5 application will not neces- sarily lead to corresponding reductions in N and K applied. Most of the surplus manure is produced on smallholdings with intensive pig production and with little land available for use of the manure. Practically the only crop that can tolerate excessive application of liquid manure without serious yield reduction, is maize. As a result, maize has become the most widely grown arable crop in the Netherlands. When, therefore, through reduc- tions in P content of pig manure, substantial reductions in liquid manure ap- plication can be circumvented, the excessive quantities of N and K applied will at least not affect the health of cattle, since most of the pig manure is applied to maize. 9. Technologies for processing liquid manures Nowadays, much attention is being paid to technologies aimed at creating dried products out of liquid manures. The activities of a few pilot plants have resulted in dried materials which can be granulated or pelleted. The organic matter contents lie around 50% and the N, P20 5 and K20 contents around 707o for each of these nutrients, with variations depending mainly on varia- tions in manufacturing techniques. In comparison with chemical fertilizers the transportation costs per unit of nutrient are high. To keep the prices com- petitive, pig producers are therefore expected to pay relatively high prices for the right to have their manure collected for transportation to the processing plant. The processing costs will be reduced considerably when drying techniques can be avoided. When, for instance, aerobic treatment of the liquid manure causes nitrification of NH 4 and, subsequently, the resulting NO 3 will undergo an anaerobic treatment in which it is denitrified, much of the nitrogen can be returned to the atmosphere as N2 . Next, phosphates in the manure can be precipitated through addition of lime. The remaining effluent still contains all the K. Due to the high solubility of K salts, this K could only be recovered by evaporating off most of the liquid phase, which in this scenario is considered prohibitively expensive. Alterna- tively, the effluent could be disposed of into the river Rhine, and from there into the North Sea. It should then be brought to the attention of environmen- talists objecting to such a disposal that an annual disposal of K present in lOx 106 tons of pig manure into the river Rhine would, as salt load, be equiva- lent to the salt entering the Netherlands every two days with Rhinewater pol- luted upstream. 52 10. Conclusions Nutrient balance sheets in the Netherlands show wide discrepancies between quantities of nutrients entering and leaving the country. The imbalance has developed as a result of imports of low-value, but nutrient-rich agricultural bulk products and exports of high-value, but nutrient-poor specialty products. For too long, emphasis was placed primarily on the importance of agricultural exports to the national economy. Presently, the damage done to environmental values by the exorbitant presence of nutrients in soil, water and atmosphere receives primary attention. Excessive production of liquid manure is one of the major causes of dis- turbed nutrient balance sheets. For a return to less extreme differences be- tween imports and exports of nutrients, three types of measures will be needed. These are 1. measures to stimulate the interest shown by arable farmers in the Netherlands in the use of liquid animal manures, 2. measures to increase the export of nutrients to other countries, in the form of liquid manure or of dried products, and 3. measures to reduce the dependence of intensive livestock production on imports of feedstuffs. If a combination of these three types of measures will not yield the expected results, in the form of a return to less extreme ratios between imported and exported nutrients, the only remaining solution will be a drastic reduction in intensive livestock production. In strategies aimed at reducing the flow of nutrients to Dutch soils, surface waters and atmosphere, potassium assumes a special position. Unlike nitrogen and phosphate, potassium does not promote eutrophication of Dutch surface waters. Unlike nitrogen, potassium cannot be eliminated from liquid manures in gaseous form. Unlike phosphate, potassium cannot be easily precipitated from liquid manure. Disposal of potassium in surface water in an effluent from which nitrogen and phosphate is eliminated, can be seen as a rather harmless and therefore acceptable way to alleviate the problem of disturbed nutrient balance sheets in the Netherlands. 11. References Black, CA.: Soil-Plant Relationships, 2nd edition. John Wiley, N.Y., London, Sydney, (1968) Centraal Bureau voor de Statistiek Kwartaalberichten Milieu 89/4 (1989) Diest, A. van: The social and environmental implications of large-scale translocations of plant nutrients. Proc. 13th IPI-Congress, Reims, France, 289-299 (1986) Dijk, TA. van: Het gebruik van dierlijke mest op grasland. 4. Kali-werking van in het voorjaar geinjecteerde dunne rundermest in her jaar van toediening. Meststoffen, no's 2/3, 10-14 (1989) F.A.O.: Trade Yearbook, vol. 42, (1988) Jarusov, S.S.: On the mobility of exchangeable cations in the soil. Soil Sci. 43. 285-303 (1937) 53 Kemp, A.: The importance of the chemical composition of forage for optimizing animal production. Proc. 12th IPI-Congress, Goslar, Fed. Rep. of Germany, 95-116 (1982) Rainer,E. J.: The availability for plants of exchangeable cations in connection with chemi- cal amelioration of soils. Bull. Acad. Sci. U.R.S.S. Classe Sci. Math. Nat., Set. Biol. 1153-1183 (1938) 54 Records of Soil Fertility in the GDR M. Kerschberger and D. Richter* Summary Within the system of plant nutrition and fertilizer application great importance is attached to characterizing the soil's nutrient supply available to plants, and to soil analysis in particu- lar. In the GDR the necessity for soil analysis covering all farm land was realized and taken up at an early date. Since the early fifties, at approx. four-year intervals, all agricultural land has been tested for pH, P, K and later for Mg (systematic soil analysis). Regarding the development of soil K content, in the early Fifties approx. 40 to 50 percent of all soils were still in poor condition in terms of K supply. Due to the purposeful and systematic application of K fertilizers the K content of the soils was progressively increased. In 1989 only 12 percent of the arable land and 18 percent of the grassland were left with low and very low levels of K supply. The DL method (Egner & Riehm) is used in systematic soil analysis for determination of available K. Tests have shown the DL method to be equal or superior to other methods and it will therefore continue to be used in future for both P and K determinations. The Agro-Chemical Centres are responsible for fertilizer distribution to the arable farms. K fertilizer recommendations are based on the results of field tests. The parameters deter- mined for K fertilization have been incorporated in an EDP programme named 1. Organization and results of systematic soil analysis (SSA) Great importance is attached to characterizing the soil's nutrient supply avail- able for plants, and to soil analysis in particular. Soil analysis can boast more than one hundred years of history. Since the epoch-making discoveries by Sprengel and Liebig, soil analysis has gained both scientific and practical relevance. Even though chemical soil analysis had to endure plenty of criticism, it has gained acceptance world-wide. The necessity for soil analysis for detecting the soil's nutrient reserves has been sufficiently proved by the scientific findings of agricultural chemistry. In the GDR this aspect was taken up at an early date, and legal stipulations on this were made as early as 1950, stating, among other things: * Dr. sc. M. Kerschberger and Prof. Dr. sc. D. Richter, Institute of Plant Nutrition and Ecotoxicology Jena of the Academy of Agricultural Sciences of the GDR, Naumburger Str. 68, Jena 9, DDR-6909, GDR 55 ((Soil analysis shall ... be carried out. It is supposed to impart precise knowledge of the nutrient contents of our soils, and thus to allow fertilizers to be used in a rational way as a precondition for maintaining and improving soil fertility. In the year 1952 the following decree was legally issued: « Table 1. Results of systematic soil analysis for the K content of the soil in approx. ten-year intervals (Wein each class) K supply level Year I (high) II (medium) Ill (low) Arable land (1989: 4.48 million ha) 1953' 26 36 38 1960 31 43 26 1970 44 37 19 1980 60 35 5 Grassland (1989: 1.07 million ha) 19532 1960 24 28 48 1970 44 24 32 1980 50 30 20 1 Arable land plus grassland 2 Not available It can be seen from Table I that in the early fifties almost 40 percent of all arable land, and approximately 50 percent of all grassland had a low, i.e. insufficient, K supply. Over the years, with the growing supply of K fertilizers to our country's farms, the low-nutrient soils could be given higher amounts of fertilizers in a most purposeful way, which fact applies particularly to the arable land. Table 2 indicates that the balance (K fertilizer-uptake) has always been nega- tive, with an average of -30 to -40 kg/ha over the years. However, when the amount of K obtained from organic fertilization is included in this calcula- tion, positive K balances will mostly result in the above-mentioned periods, 56 because a significant amount of roughly 30 to 50 kg of K/ha is to be expected from organic fertilization every year. Although in detail it should be noted yet that, particularly on the low K sand and sandy soils which account for most of the soils on K supply level II1, a more intensive K fertilization was carried out than on the high-sorption soils rich in K. This selective use of K fertilizers led to an almost complete elimination of low K supply on arable land and, with some restriction, on grassland. Table 2. Development of crop yields and K fertilization in GDR agriculture (according to Wiler [19841) Period Crop pro- Applied in Uptake Difference (K duction, fertilizer (kg K/ha) fertilizer minus grain equi- (kg K/ha) K uptake) valents (kg K/ha) (t/ha) 1950-55 ...... 3.17 50.2 88.8 -38.6 1956-60 ...... 3.49 63.7 97.8 -34.1 1961-65 ...... 3.14 72.9 87.9 -15.0 1966-69 ...... 3.69 71.8 103.3 -31.5 1970-73 ...... 3.82 80.5 107.0 -26.5 1974-77 ...... 3.96 83.0 110.9 -27.9 1978-82 ...... 4.19 68.7 117.3 -48.6 1983-85 ...... 4.76 84.6 133.2 -48.6 1986-88 ...... 4.88 92.6 136.6 -44.0 2. Calibration of soil analysis At the beginning of the eighties our institute was confronted with the task of revising the limit problem, i.e. classifying the soils' K contents in different supply levels, with the aim of determining five such levels. The verbal defini- tion of these supply levels (SL) is as follows: SL I very high content in the soil SL 2 high content in the soil SL 3 medium content in the soil SL 4 low content in the soil SL 5 very low content in the soil The soil K content for deciding fertilizing strategy is indicated by K supply level 3 which is defined as follows: Medium K contents of the soil=range of the nominal value of soil fertility; optimum content to be aimed at in the soil. To achieve highest crop yields, fertilization is required to fully counter- balance uptake by the plants, with the DL soluble content in the soil being 57 kept constant. To compensate for leaching on certain soils a small extra amount is required, in addition to replacing uptake. The experimental basis for deriving critical values consisted in a multitude of static tests with increasing K levels. The experiments were planned as long term field experiments with mostly 4 treatments (up to 10 maximum) and 4 replications in each case. On typical sites 65 experiments have been carried out in several experimental series since 1954 on arable land (560 crop years) and 53 experiments on grassland (310 crop years). The experiments included a control treatment and increasing rates of K fertilizer. In order to establish 5 supply levels of the soil, the relative yield as percent of the maximum annual yield has been determined each year for all treatments. Then treatments were distributed in groups according to their annual K balance (K fertilization minus K uptake). With 7 balance groups the variability of K balances of the treatments could be put in such an order that every balance group was occupied in a representa- tive way by pair comparisons. Afterwards the relation between K content of soil and relative yield was calculated for every K balance group. The point of contact of the regression straight line with the relative yield of 100 percent was used for determining critical soil K values. The data was presented by soil type. Plant analysis was not used for determining critical values. The new five- step ranges of critical values are presented in Table 3. Table 3. Critical values (mg K/100 g soil; Egner, Riehin, DL method) for soil groups and supply levels Soil Type of particle size, Supply level group geological origin I 2 3 4 5 Plough land I Sand ...... > 16 15-11 10- 7 6- 4 3 2 Loamy sand ...... >20 19-12 I- 8 7- 4 (3 3 Sandy loam ...... >23 22-14 13- 9 8- 5 (4 4 Loam ...... >27 26-16 15-11 10- 6 5 4.4. Loess-loam ...... >23 22-13 12- 9 8- 5 4 5 Clay ...... >40 39-23 22-16 15-10 (9 6 Moor ...... 25 24-17 16-13 12- 7 (6 Grassland I Sand ...... 19 18-11 10- 6 5- 3 2 2 Loamy sand ...... >23 22-12 11- 7 6- 4 <3 3 Sandy loam ...... >26 25-13 12- 8 7- 4 3 4 Loam ...... >29 28-16 15- 9 8- 5 (4 4.4. Loess-loam ...... >26 25-13 12- 8 7- 4 <3 5 Clay ...... >30 29-16 15- 9 8- 5 (4 6 Moor ...... >25 24-16 15-Il 10-, 7 (6 After introduction of critical values for K since 1988 (Kerschberger,Richter and Witter [19861) the results of the systematic soil analysis for arable and 58 grassland have been classified in a new way. The K supply of the soil according to the new ranges of critical values is shown in Table 4. Table 4. Results of the systematic soil analysis for the K content of the soil (0 by class, 1989) K supply level 1 2 3 4 5 Plough land ...... 12 41 35 II 1 G rassland ...... 19 38 25 13 5 According to Table 4 the GDR soils' K supply is variable, but can on the whole be described as satisfactory. Only one percent of all arable land has a very low K content and needs systematic and urgent K fertilization. The same applies to five percent of the grassland. On the other hand, high K fertilization carried out for decades led to a considerable proportion of land with very high K supply. With the exception of light, sandy soils on arable land, K fertilization could be omitted completely, since K fertilization of land with very high K content of the soil will, for the most part, produce no increase in yield (Table 5). This will also comply with the ecological demands for reducing the amount of materials carried into the soil which, for example, is important for protecting our water resources. Table 5. Mean increase of yield achieved by K fertilization in long term field experiments Percentage increase in yield compared with no K fertilization K supply level Plough land Grassland I ...... 0 0 2 ...... 7 7 3 ...... 10 12 4 ...... 14 20 5 ...... 29 32 The data in Table 5 further indicate that on the average of all test results the effect of K fertilizers is closely connected with the K supply of the soil. In the case of supply level 5 a yield increase of about 30 percent is achieved by means of K fertilization, i.e. heavy crop losses are to be expected should K fertilization be omitted. The same still applies, to a considerable extent, to K supply level 4. With the soil's K supply growing, yield increases due to K fertilization are becoming substantially smaller, and are mostly accidental and rarely of 59 statistical relevance. K fertilization in the case of supply level 2 serves to stabi- lize the desired level of output, rather than increase the yield. In this connec- tion the level of K fertilization is adjusted as to add less K to the soil than is extracted by the crop, and to achieve a gradual decline in the soil's K content toward K supply level 3. 3. Comparison of soil analysis methods The present system of basing K fertilizer recommendations on the results of soil analysis calls for an answer to the question as to whether the double lactate (DL) extraction yields representative results. Does the method largely reflect the crop's response to soil nutrient content? Richter [1978] compared various methods of extraction with the results given in Table 6. Table 6. Coefficient of regression (r2 ) between K uptake by plants and K content of the soil in pot experiments with different types of soil Extraction method Authors Coefficient (r2) Low High K content K content in the soil in the soil DL method Egner, Riehm ...... 0.59 0,90 AL method Egner, Riehm, Domingo ...... 0.88 0.88 0.5 N-NH4-acetate Bailly ...... 0.90 0.92 2 N HCL Milceva ...... 0.90 0.92 Table 6 indicates that the DL method reflects just under 60 percent of the plants' K uptake when the K contents of the soil are only low. In contrast to this, accuracy is as high as 90 percent when the soil K contents are medium or high. Under the latter conditions the other extraction methods tested yield no better information. As most soils in the GDR boast medium or high K and soils with very low K are to be found only on a very small scale (Table 4), adherence to the DL extraction method was considered justified. The fact that apart from K also P is determined by means of the DL extract, which is of considerable economic advantage, is another reason for retaining the DL method. In the GDR no raw phosphates were used in P fertilization. Therefore the evidence of the DL method is satisfactory. The advantage of applying a certain extraction method for a long period is to provide comparable test results, as well as the possibility of precisely evaluating the development of the nutrient contents in the soil. Thus it is pos- sible in the GDR to trace and present over decades the development of the K content in the soil (as well as P, pH, Mg) for territorial units such as rural communes, districts and counties. 60 4. Fertilizer distribution by the Agro-Chemical Centres (ACC) The ACCs distribute K fertilizers for all arable farms, and the fertilizer- spreading equipment is also concentrated in those enterprises. First of all the ACCs accept the fertilizers delivered for arable farms and store them till the date arranged for spreading. The practical work of the ACCs is inspected and supervised by the arable farms. The quantity of fertilizers to be spread is based on the fertilizer recommen- dations annually calculated on every crop-producing farm by computer pro- grams. This EDP project ((Fertilizing System was developed by agricultural scientists and first employed in 1969 to calculate fertilizer recommendations for crops. The system offers plot-related recommendations for fertilization with macro- and micro-nutrients. The further developed EDP project ( Table 7. Determination K fertilizer requirement using an EDP-fertilizer-system (DS '87) for winter wheat (8 t grains/ha) on loam soil (weathered limestone) Plot K supply K K balance K Minera K 2 level uptake + sheet value • delivery = fertilization (kg/ha) (kg/ha) factor' I 5 136 + 100 1.0 236 2 4 136 + 50 1.0 186 3 3 136 ± 0 1.0 136 4 2 136 - 50 0.9 77 5 1 136 - 136 - 0 determined by a special soil analysis method the K amount from organic fertilization must be subtracted * K in above-ground biomass (grain + straw or beet + tops). 61 5. Experience in cooperating with crop-producing farms There is intensive and many-sided cooperation between scientific institutions and crop-producing farms. For example, annual meetings on fertilization are held at district and county levels and numerous publications (also in the daily press, among other things) provide information about the state of art etc. The work done by the Advisory Body of the Department of Agrochemical Analysis and Advice of the Institute of Plant Nutrition and Ecotoxicology Jena is particularly intended for cooperation with the crop-producing farms. A qualified staff of about 100, organized in 14 county work groups, is respon- sible for giving agrochemical advice to all crop-producing farms and ACCs. In this connection the advisory work of the Department of Agrochemical Analysis and Advice is more and more aimed at the introduction and handling of microcomputer programmes. These programmes give recommendations on soil and plant testing, as well as on mineral and organic fertilization, and the crop-producing farms buy them. What can be pointed out and stressed about our many years' co-operation with the crop-producing farms is that it is practised with a great deal of mutual interest. Discussions on critical aspects stimulate our thinking and serve the further development of both science and practice. 6. References Kerschberger, M. andRichter, D.: Erminlung von Bilanzkoeffizienten fir die P-Dungerbe- messung. Arch. Acker- u. Pflanzenbau u. Bodenkd., Berlin 22, 9, 559-567 (1978) Kerschberger, M., Richter, D. and Witter, B.: Neue Versorgungsstufen for die P-, K-, Mg-Ge- halte und den pH-Wert des Bodens auf Acker- und Groinland. Feldwirtschaft 27, 8, 366-368 (1986) Richter, D.: Charakterisierung der K-Bindungsformen im Boden dutch unterschiedliche Extraktionsverfahren. Arch. Acker- u. Pflanzenbau u. Bodenkd., Berlin 18, 7, 503-515 (1974) Richter, D. and Kerschberger, M.: Pflanzenverfflgbares P und K in Ackerboden als Grund- lage fOr die Dingerbemessung. Tag.-Ber., Akad. der Landw.-Wiss. DDR, Berlin 166, 295-301 (1978) Witter, B.: Entwicklung, Stand und Ergebnisse bei der Durchfohrung der systematischen Bodenuntersuchung in der DDR und Vorschlage fur die ktinftige Verfahrensweise. Akad. der Landw.-Wiss. d. DDR, Berlin, Diss. B, 128 and Appendix, 1984 62 Agrochemical Monitoring of Exchangeable Potassium in Arable Soils of the USSR M.A. Florinsky and E.N. Yefremov Summary Agrochemical monitoring in the USSR includes periodical examination of agricultural and cropped lands for the determination of labile phosphorus, exchangeable potassium, humus, micronutrients and lime requirement as well as constant observation for soil fertility on special sites. The aim of monitoring is to diagnose soil-status, to reveal the change of soil- status resulting from the regular use of mineral and organic fertilizers and other agrochemi- cals, to predict soil fertility in the near future. In the past 20 years the area of arable soils low in exchangeable potassium has reduced by 17 mio. ha. The rate of change in soil characteristics differs greatly in various regions of our country. The dependence of area change of the soils low in K on the amount of fertilizer applied has been established. 5-15 times as much potassium fertilizer is used for the correction of exchangeable potassium content in soils of arid zones as in the humid regions. Negative tendencies have been increase in the area of low exchangeable potassium soils in the Trans-Caucasus and Central Asia. This is the result of underestimating the sig- nificance of potassium. At the same time an increasing area with very high levels of exchangeable potassium exceedingagrochemical and ecological optimum was recorded in the zone of intensive appli- cation of potassium fertilizer (Byelorussia, Ukraine). A forecast of soil distribution in 2000-2030 with speical reference to different content of exchangeable potassium has been made. 1. The programme of agrochemical monitoring The problem of soil conservation and soil fertility increases as well as ecologi- cally safe fertilizer use can only be solved on the basis of the system especially developed for soil-ecological monitoring. In the USSR the agrochemical monitoring of the soil fertility status was established in 1964. It provides for: - periodical information about the content of labile phosphorus, exchange- able potassium, organic matter and the degree of acidity; * M.A. Florinsky and E.N. Yefremov, Central Institute of Agrochemical Services, CINAO, Pryanishnikov St. 31/2, 127550 Moscow, USSR 63 - data in regard to changes of soil fertility due to regular use of mineral and organic fertilizers as well as chemical ameliorants; - determination of tendencies and relations of soil fertility changes; - soil-status forecast for the near future; - formalization of quantitative informations for soil fertility maps. The State National Programme of Agrochemical Monitoring provides all land users, irrespective of their organisation and structure, with periodical data on potassium determination in croplands, grasslands and in perennial stands. 205 agrochemical laboratories make soil tests on an area of 38 mio. ha annually. Since 1974 a system of control sites has been developed where observations are made during the whole vegetation period on the changes of exchangeable potassium, labile phosphorus and micronutrients in the arable layer and in subsurface horizons. These control sites are situated on the fields of collective and state farms all over the main agricultural zones of the USSR and the cultural practices are the same as on the farms. Soil moisture content is also measured. To reveal the tendencies of soil fertility changes and their relations with regular fertilizer application the annual official statistics are used on mineral and organic fertilizer application with special reference to the information about fertilizer type and quality. 2. Methods of investigation Agrochemical soil monitoring is based on soil sampling, analyticaltest and mapping. Agrochemical maps and field reports are compiled in accordance with the results of these analyses. The unit of an agrochemical investigation is an elementary area the size of which can be changed depending on native and economic conditions not only within the boundaries of one zone but in the farms also due to the soil cover variability and different nutrient levels. In zones with podzolic and sod-podzolic soils as well as hydromorphic soils with badland and rolling relief and different parent rocks it is 1-3 ha, in plains 5-8 ha. As to chernozems in steppe and wooded steppe with rugged relief and mixed soil cover the site size is 3-6 ha (normally 5 ha); in wooded steppe with broken relief and relatively similar soil cover 5-10 ha; in steppe with flat relief and similar soil cover 10-25 ha. In mountain regions the sizes vary within 0.5-3.0 ha because of great soil heterogeneity. In the region of irrigated cropping the area of such site equals to 2-5 ha which corresponds with the typical irrigated territory. Depending on the cropping system the elementary area increases for arable crop rotation and decreases for vegetables (1-2 ha). 64 Each elementary area is represented by one mixed sample consisting of 20-40 separate probes taken evenly from the whole area. While taking probes their representativity for the given field is strictly taken into account. Soil samples are taken during the whole vegetation period either from non- fertilized areas or from low fertilized areas (45-60 kg/ha active ingredient). As to the fields where such high rates as 90 kg/ha nutrients are used mixed samples are taken 1.5-2 months after fertilizer application. In fields with ma- nure soil sampling is carried out during the whole vegetation period. To determine exchangeable potassium in the USSR soils the following methods are used (Figure 1): Exchangeable potassium is extracted from podzolic and sod-podzolic soils as well as from gray wooded soils by Kirsanov's method (0.2 M HCL solution in the ratio 1:5 soil: solution for mineral horizon and 1:50 for organic horizon) with subsequent flame-photometric determination. In the zone of dark gray wooded, leached, ordinary typical chernozem ex- changeable potassium is determined by Chirikov's method based on its extrac- tion from soil with 0.5 M C2H4 0 2 solution in the ratio 1:25 soil: solution. In the regions with chernozems, chestnut, brown, gray-brown and gray soils exchangeable potassium is determined by Machigin's method based on extrac- tion with 1% NH 4CO 2 solution in the ratio 1:20 soil: solution. In the Baltic region exchangeable potassium is extracted from sod-podzolic and sod-calcareous soils by DL- and AL-methods. In the first case with 0.04 M calcium lactate solution; in the second case buffer solution is used (pH 3.7) containing lactic and acetic acids and ammo- nium acetate. Soil: solution ratio 1:50 for DLmethod and 1:20 for AL- method. On the relatively small area of humid subtropics (krasnozem) exchangeable potassium is extracted by Oniani's method (0.05 M H 2 SO 4 solution in the ratio 1:25 soil: solution). The test data obtained are represented by large scale maps or field plans. The field plan is a document form with average exchangeable potassium con- tent and reserves together with the other agrochemical data. Test results obtained in each of 3005 administrative regions are supplied to the data bank in Central Institute of Agrochemical Services (CINAO) on special record cards to be summarized and transformed into information bulletins and maps which are then sent to the agricultural and ecological authorities. 65 Figure 1. Methods of determination of exchangeable potassium and labile phosphorus used by agrochemical services in the USSR 180 240 300 360 420 48 540 560 . ,1 -lcla1 e OOf 2. -' KProv inmet nu mber 4 - Prone number 480 540 66 102". 1080, 1140 1200 1260 1320 1380 600 - ~ ~ J *'iji Ns-48 __44& Ala-t Aral. Se / $..I CINO-98 60 660 72 Os'67 3. Fertilizers usage Regular use of fertilizers and manures has two main objects: - to satisfy plant nutrients needs; - to prevent soil impoverishment. The amount of potassium applied as fertilizers differs essentially between republics and regions. More than 75% of all potassium fertilizer is used in the European part of the USSR. In Byelorussia and the Baltic republics 90-140 kg/ha K20 are used annually, in the Trans-Caucasus and Central Asia 25-35 kg/ha K20. Minimum of K20 is used in the steppe regions of Kazakh- stan (0.5-1.5 kg/ha K20). There is a significant variation in the amount of manure applied to the arable soil (from I to 15 t/ha) per year. Besides organic manures have differ- ent chemical composition depending on animal species, structure of cattle stock and native-climatic conditions. As a result of regular controlling manure quality it is established that potassium level in manure in the south and south- east regions is 2.5-3.0 times as large as in the north and north-west. Figure 2 summarises rates of potassium applied per ha of cropland in differ- ent Soviet republics during the past 20 years. Efficiency of potassium fertilizers in different regions of the USSR (yield response by I kg K20 in grain units) is shown in Figure 3. Yield response grows from 2.5 kg in steppe and dry steppe regions up to 6.0 and more in the zone of adequate moisture. 3000- 2 2500- o 2000- 0 -1500 500- 't 0) t C C MC M0 M M0 M C M C ( , 2 u- - m w E w2 nn (n 2J 12 2 5 c <2 Figure 2. Total potash applied in the past 20 years in the Soviet Republics. 68 4. Soil K-status A large proportion of arable soils in the USSR are classified middle or high in exchangeable potassium (Table 1, Figure 4). Grazed and hayed grassland and land under perennial crops are generally of lower K status. However there are large differences between different native-agricultural zones of our coun- try. Thus in the Baltic republics, in the Trans-Caucasus and Central Asia there are 25-5507o low K soils, in Moldavia and Kazachstan 2-3%, in Byelorussia 14.9%, in Ukraine 8.7% (Figure 5). Variability in exchangeable potassium con- tent is typical of the Russian Federation (Figure 6). In the Central Zone of RSFSR where sod-podzolic soils preponderate the proportion of low K soils varies significantly even between the neighbouring territories - 11.9% in the Moscow region and 40.1% in the Smolensk region (Figure 7). Weighted K20 content in arable soils increases from taiga zone to semidesert zone (Table 2). Table 1. The percentage distribution of the soils with different potassium level in the USSR Content of mobile potassium (mg KZO/kg of soil) Year Type of land Low (0-80) Medium (81-120) High (> 120) 1971 Arable land 16.9 23.1 60.0 1990* 9.2 20.0 70.8 1971 Hayland 41.2 23.1 35.7 1990 28.1 24.9 47.0 1971 Pasture 32.5 20.7 46.8 1990 19.5 20.6 59.8 1971 Perennial stands 21.2 28.5 50.3 1990 20.6 21.0 58.4 * Situation on January I Table 2. Exchangeable potassium content found in the soils of the USSR main native agricultural zones K20 content Determination Zone (mg/kg) method Southern taiga zone (sod-podzolic, podzolic-brown, brown forest soils) 76.7-142.1 Kirsanov Forest-steppe zone (gray forest, podzolized, leached and typical chernozem) 104.7-168.9 Chirikov Steppe zone (ordinary and southern chernozem) 182.8-304.8 Machigin Zone of arid steppe (dark chestnut and chestnut soils) 385.0-432.9 Machigin Semidesert zones (light chestnut and brown soils) 465.4-507.0 Machigin 69 Figure 3. Efficiency of potassium fertilizers (yield response in kg grains per kg K20 applied) 18240 300 4250 48 52~4 -. 48C~- 54 522 5-'o -- 2 Black Sea -A 2-2- 70 ° 3-4.1 4g9:5.9 > 5.9 4 Zone1 number -,.. J Province number 3 A - Subprovince index 48 ° 540 70 60- 10n 300 1AA 4 $ ,r U 440 -Average yield j tlw "----. response in 9866 /I . field experiments Scarried out in "-kg KO applied S '-8. 72 -6 0- 71 Figure 4. The distribution of the soils with different levels of exchangeable potassium. W t. 3S H *6.. l r cutl~l,u~ cita¢c a |* V S' #10 Miy -oil ievJ J 4 Azv *1 A X/'a 6 6%AA j Low KO content (> 50%1) (low content 30-50%) KO content (lowMLeMiddle contentand ighl <30%1 K 0 content 9, Middle and high K 0 content (middle content 36-50%) High KO content imiddle content <30%1 High K,0 content (>80%) 72 C'NA I 73 1 0, USSR o North-W0stn00 0 o o 0Ukra0ine o 01,, 1 , 0 I I I II I I II I I I1 I I I I o , -us North RSFSR S North-West o Ukraine I 2 Byelorussia Central a E Uzbekistan Volgo-Vyatka Kazakhstan 0 Black Earth Georgia Azerbajan Povolzhie so o Lithuania N. Caucasus Moldavia C0 " Latvia Ural S-" Kirghizia West Siberia Ei, Tadjikistan C Armenia East Siberia 5C Turkmenistan Far East Estonia 40- 35" 30- 25- 20" 15" 10. 5- 0 C U ' ~ C . > ' E 0 ~ 0' Figure 7. The distribution of low K soils in the Central zone of RSFSR (1990). It has been established with some certainty that as the sum of active tempera- ture (Et > 10°C) increases and the accumulated precipitation decreases in moving from the southern taiga forest zone to semidesert, the proportion of high K soils increases by a factor of 4 while the proportion of low K soils is decreased by a factor of 16 (Figure 8). Distribution of arable soil area Climatic conditions S-Y - y0 High content E t >10C 60. % z C . z --- 2000 0 0 53 S 20 (n 0 20 40 60 80 100 0 650 x=49.9 y=28.2 z=21.9 Low content Precipitation (mm) Figure 8. Distribution of arable soil area with different exchangeable potassium content for main agricultural zones in the USSR 75 y High content W.!5 r) y t >100C N% = - 40- -2400 4 j 0.42 0 ? 20- z x x 0 20 40 60 80 100 0 500 x=13.8 y=27.4 z=58.8 Low content Precipitation (mm) yY High content 2 LD-y Yt >10-C 0 -2850 g 40 c 20 - x x 0 20 40 60 80 100 0 425 x=4.0 y= 18.8 z=77.2 Low content Precipitation (rm) High content 6% t >100 C 60z S .40--- 3175 0 040- 2 0° 0 220- 0.1 o /Z I .i I I,xl/ . % xX x I .. I- 0 20 40 60 80 100 0 300 x=3.4 y=11.7 z=84.9 Low content Precipitation (mm) High content E t 60_ % C F. t >100C c - N zC - --- 3250 - KZ 50 40- 00 0 2. .. ,:,01 0 20 40 60 80 100 0 200 x=2.9 y= 8 .8 z=88.3 Low content Precipitation (mm) 76 Study of factors affecting the distribution of soils with different exchange- able potassium content (sum of active temperatures, sum of precipitation, continentality of climate, coefficient of moisture, biological productivity) has shown that the distribution depends mainly on 2 factors, i.e. sum of active temperatures and sum of precipitation and is described by the following equa- tions: Ys high = -101.0 + 0.057 Et + 0.024EP, R2 = 0.94 Ys middle = 39.8 - 0.01 Et + 0.018 SP, R2 = 0.93 Ys low = 159.9 - 0.047 St - 0.046 EP, R2 = 0.76 where Ys (high, middle, low)= % of soils with high, middle and low content of exchangeable potassium; Et=sum of active temperatures (> 100 C), 0C; EP=sum of precipitation, mm. For more detailed analysis it is necessary to incorporate among the factors being studied the soil texture thus enhancing the reliability of evaluation of K-status in soil, both for large areas (republic or region) and for small territo- ries (individual farm). 5. K-status changes In the USSR the area of soil low in exchangeable potassium decreased by 7.7% as a whole during the past 20 years (Figure 9). The greatest shift took place in Byelorussia where, as a result of regular use of fertilizer and manure at high rates the amount of soils low in exchangeable potassium was more than halved. The areas of such soils in Lithuania and Latvia as well as in the north-west regions of Russia reduced by 20-30%. In the central and south belt of Russia where chernozem and chestnut soils are widespread as well as in Moldavia there was no significant change (Figures 10, 11). Long intensive cropping decreases the level of exchangeable potassium in the soil. In the Caucasus and Central Asia, particularly in Azerbaijan and Tadjikistan, where heavy potassium removal by intensive crops is noticeable the area of potassium impoverished soils has increased greatly. This trend has been steady over the past 10 years. Failure to understand the importance of potassium fertilizer requirements has resulted in the appearance of new areas low in exchangeable potassium in some regions of the USSR (about 1.5 mio. ha). In most regions of the USSR removal of potassium by crops exceeds the amount applied in fertilizers and K-balance is thus negative except in the north-western part of the USSR. 77 The change in area of low K soils over the past 20 years is related to the potassium balance by the following equation: A S=-0.016 z K-8.36, (R2 =0.66) where A S=change in area low in K A K=potassium applied - K removed in crops At the same time the area of soils with high potassium content has increased in Byelorussia and the Ukraine. Altogether in the past 10 years 2.5 mio. ha such soils have appeared in the USSR. This indicates irrational usage of potas- sium fertilizers as a result of which potassium content exceeds the agrochemi- cal and ecological optimum. High fertilizer rates greatly exceeding K removal by crops increase leaching and reduce K fertilizer efficiency. The changes indicated by samples taken at 4 year intervals on the control sites over 15 years obey the objective laws typical for the given agricultural zone. They also confirm the tendency of potassium content in the arable layer to increase, provided fertilizers and manure are used at the recommended rates and the farms are properly managed. 60 551 40- c 35- .2225", 20.! 15- 0 5-V -L 0 MM M 20C C M, _2 E 25 Figure 9. Change in area (0o) of low K soils over the past 20 years. 78 25- 20. 15. 10- 5- UL) . > D F0 (0 t00 E t- 2 Figure 10. Change in area (%M)of low K soils over the past 20 years in the different zones of Russia. 25- 45- 40- 35- 30- 25- 20- 15- 10- 5N 0 U ~ ~ ~ ~Z . 0w E Figure 11. Change in area (%) of low K soils over the past 20 years in the Central zone of RSFSR. 79 6.K-status forecast The results of soil analysis on 220 mio. ha arable and 42 mio. ha grassland carried out in the USSR shows the overall improvement in the K status of soils over the past 20 years. Potassium fertilizer usage for the improvement of exchangeable potassium content in soils in arid zones is 5-15 times as much as in humid regions. Using the recommended rates of potassium fertilizers and taking into ac- count the estimated yield it is possible to increase the K content of im- poverished soils up to a sufficiently high level. In different zones of our coun- try this process could be finished by the year: - 2000 in Byelorussia; - 2015-2020 in RSFSR, Ukraine, Latvia and Estonia; - 2030 in Lithuania. Fertilizer and manure dressings should be adjusted in these zones to allow for release of soil K-reserves and should not be so large as to risk excess potas- sium being leached. Since there are limits to the capacity of soils to retain nutrients, there is a real probability that potassium (at high rates) will start to pollute surface waters and deep groundwater which has serious conse- quences for the environment. The situation in the Trans-Caucasus and Central Asia is more complicated. It is necessary to forecast the extent to which reserves in soil can buffer a deficit of K where the crops are removing more K than is replaced by fertilizers, organic manures and crop residues. If we do not increase the use of potassium fertilizers in these zones there is a real risk that the arable soils will be utterly impoverished. 7. References Mengel, K. andKirkby, E.A.: Principles of plant nutrition. 3rd Edition Int. Potash Inst., Bern, 655 p. (1982) Quernener, J.: Important factors in potassium balance sheets. Proc. 13th Int. Potash Inst. Congr., 41-72 (1986) Schroeder, D.: Relations between soil potassium and potassium nutrition of the plants. Proc. 10th Int. Potash Inst. Congr., 33-66 (1974) 80 The Functions of Plant K in Overcoming Environmental Stress Situations U. Kafkafi* Summary There is some evidence that duxury consumptiono of potassium is actually a good and relatively cheap ( 1. Introduction The general considerations of plant responses to environmental stresses were described in detail by Levitt [1972]. The influence of K nutrition on the response of plants to environmental stress was reviewed by Beringer and Troll- denier [J980]. It was suggested that the concentration of K in the cytoplasm must be maintained at a level of - 150 mM and that in the vacuole at a mini- * Prof. U Kafkafi, Hebrew University of Jerusalem, Faculty of Agriculture, Center for Agricultural Research in Arid and Semi Arid Lands, P.O. Box 12, Rehovot 76-100, Israel 81 mum level of 20mM (Leigh and Wyn Jones [19841). Any further decline below these values causes a severe reduction in plant growth. In the low K concentra- tion range, the cell's growth responds according to the law of limiting factors. When the K supply is abundant the cell accumulates K in the vacuole and as a result the total K content as expressed on dry matter basis increases. This increase in the content of K on the whole plant basis with no further increase in plant growth is usually referred to in the agronomy literature as oluxury consumptiom or «range of high level> (Bergmann and Bergmann [19851). This review aims to bring some evidence that 2. Cold stress The effects of the high K content of the cell in increasing frost tolerance have been related to decrease in the osmotic potential of the cell sap (Beringerand Trolldenier[1980]). One should distinguish between frost and chilling effects. Chilling effects in the range of +5 to +10'C commonly occur in plants in warm climates. The temperature at which a sudden change in membrane fluidity occurs is specific to each cell and dependent on the relative composi- tion of the various phospholipids. The higher the ratio of unsaturated/satu- rated fatty acids in the cell membrane the more tolerant is the tissue to low temperatures (Beringer and Trolldenier [1980]). The effect of increasing K concentration in the fertigation scheme on yield and chilling damage of carnation was studied recently (Yermiyahu and Kaf- kafi [1990]). Calyx splitting of carnation flowers and heavy percentage of brittle stem was observed in the standard carnation cultivar (White Candy) grown on a sandy loam soil. The treatments and yields are reported in Table 1. Table 1. The leaf content of K during the growing season as affected by K concentration in the irrigation water and its effect on yield and brittle stem of carnation CV. White Candy K in irrigation K in leaf Brittle Stem Yield (g m-3) (g kg-') (0/o of total (flowers m) yield) November April 93 ...... 32.4 24.6 31.8a 89.5a 252 ...... 37.2 27.2 30.9a 89.5a 378 ...... 40.5 32.6 16.9b 90.Oa * K content in the 5th leaf from the top of a stem without a flower. The CV. was less than 10%. 82 Three levels of K2S04 were constantly supplied to the irrigation water so that the concentrations obtained were 93, 252 and 378 g m - '. These K levels were checked on a constant level of total N 126 g m - ' in the trickle irrigation. The effect of the type of the nitrogen source was compared at the highest potassium level. Two N fertilizers were compared: 1) KNO3 and 2) a liquid 7-3-7 commercial fertilizer to which K2S0 4 was added. The ratio N-NO3/ N-NH 4 in the irrigation water was 1/0 with the KNO 3 fertilizer and 4.3/3 with the 7-3-7 commercial fertilizer. The supply of the highest level of K in the daily irrigation resulted in the lowest percentage of broken stems during 4.5 months of flowering (Figure 1). The percentage of brittle stem shows two peaks on 21.2.1989 and on 7.3.1989. These peaks are observed 4-5 weeks after a cold night event below 80C followed by a clear sunny day. When the K content in the 5th leaf from the apex was below 4%, about 31% brittle stem was recorded with heavy losses to the grower. It was only 18% loss when the K content in the plant 162 days after planting was greater than 4%. The highest yield of flowers was obtained in the KNO 3 treatment (105 flowers m -2), about 170% above the other treatments that contained ammonium-N in the irrigation water. The flower quality as measured by calyx splitting and brittle stem was improved by the highest level of potassium irrespective of the K-fertilizer form KNO3 or K2SO4. 70- 60- E 40- 0 S30- Cm 20- 10- 31.1 7.2 14.2 21.2 28.2 7.3 14.3 21.3 29.3 4.4 12.4 Date, 1989 Figure 1. Broken stems of carnation cv. White Candy during the growing season on three levels of potassium in the irrigation water: -E 93 + -252 -- 378 g Km-'. (Cold night events below 80C occured on January 5 and February 6, 1989). Brittle stem susceptibility was significantly reduced by supplying high levels of potassium in the irrigation water at the same level of total N concentration 83 from the beginning of the growing season. As long as the temperatures did not drop below 120C the high doses of K would have been regarded as ((luxury>) or waste. The economical saving to the farmer after only one night event of low temperature is more than total fertilizer cost for the whole season. The fact that the K effect in reducing extreme events in brittle stem is detected only about 5 weeks after the cold events probably explains why such K effects are not often reported in the literature. The continuous supply of potassium at a much higher concentration in the soil solution than that regarded as sufficient for maximum yield might prove an insurance against unexpected climatic events. High K content in the plant reduced the brittle stem of the carnation. The point of breakage occurred in the tissue connecting the 5th or the 6th stem nodes from the flower. The position of the weak point on the stem suggests that an irreversible change in the tissue structure occurred during the development of this node and was affected by the cold night event. The weak point most probably is found in the fiber-like structures that later break-off even at very small tilt. It is possible that the cell wall structure is affected but this hypothesis must be further checked. The actual physiological mechanism that was operating in the carnation case is not clear. The susceptibility to such cold events is very much dependent also on the plant variety. The changes in varietal response to chilling effects might be due to differences in the fatty acid composition of their membranes. The root membrane composition has been reported to have effects on the rates of ion and water transport in the root, carbohydrate content of the plant or translocation of nutrients and metabolites in the plant (Marschner[1986). Of all the potential effects of high K supply to the plants in preventing cold damage, the high K is increasing the metabolite content of the tissue. It is reported for tomato production (Adams [19781), that high levels of K in the root medium result in increased K content in the leaf and fruit. Concentrations of 100 and 300 ppm K in the liquid feed resulted in 4.8 and 7.0% K in the leaves respectively with the high leaf level producing the best quality fruit. High levels of potassium were also recommended to reduce hollow fruit in tomato which is associated with low light and cold nights in the greenhouse (Winsor [1966). It is therefore a safe policy to keep high K concentrations in the soil and in the plant, especially in glasshouse and other intensively grown crops, and thus to increase the soluble carbohydrate content that in the case of a cold stress may act in reducing the damage to plant tissues. The effect of soil temperature around the crown node (transition region) of wheat on growth and nutrient translocation was studied by Boatwright et aL [19761. Careful chilling enabled them to demonstrate that only the crown zone restricted translocation of 16Rb and, by inference, also of potassium to the top (Figure 2). Unlike the crown node, the shoot growing point (meristematic region) was not affected by low surface-soil temperatures. It is the actual transport through the chilled crown zone that influenced the wheat dry matter yield. Their data show that the surface soil temperature in the 2.5 cm depth has an influence on plant development with an optimum for wheat at 19°C (Figure 3). 84 6- 5- E 4- c 3- . " 2- 1 11 0C 4 8 12 16 20 24 Hours subjected to "Rb 86 Figure 2. Effect of crown-zone temperature on translocation of Rb from the roots to the tops of spring wheat (Boatwright et aL 119761). 45- -1.80 Tops, D.W. -1.60 40- 0 _35- 1.40 - - 30 -1.20 _T25"-- a0, 20" - Roots,r D.W .8080 3dleaf> 5 4.60 0 15& • 4th fea~f t le.20 8 12 19 26 Surface soil temperature, 0 C (2.5 cm depth) Figure 3. Effect of surface soil temperature (from 14 to 25 days after wheat emergence) on dry matter yield of roots and tops and on leaf-length growth (Boatwright et al. 11976]). 85 As in the carnation case, the node region is the most sensitive zone to chilling effects. The exact reason for that sensitivity is not clear but its effect on translocation hints that a 3. High root temperatures The rate of K translocation is enhanced by increasing the root temperature up to a maximum above which the roots start to suffer. Kafkafi and Eshet (unpublished results) measured the uptake of 42K by a single root of intact banana plant. The whole plant was grown in nutrient solutions of different temperatures. A single root of about 50 cm length was selected and the 20 cm section from the tip was introduced into a 200 ml glass tube containing the same nutrient solution containing radio active 42K. A 10 cm lead shield protected the Victoreen Geiger counter that measured the radioactivity 40 cm from the root tip. The diurnal pattern of 42K transport in intact banana plant is shown in Figure 4. It can be seen that potassium is accumulating in the 800- -720 SIt ' I / / 0640 '" 340C x 560 / I E - 480. 4BO , , I :". 400" 320- 1 . ,"o ° 3 2 C260 I 26°C > ...... " " o /o • 240' 160- 80.I Night Night Night I ' I * 1 In n I I I I I I lI 0 - 1IIVP" * N . . 24"' I I 4 ' 8 I 12 16 20 24 4 8 12 16 20 24 h " :Labelling 42 Figure 4. The activity of KN03 in the root of intact banana plant during night and day at two root temperatures. 86 root during the night and its level in the root starts to drop with the onset of transpiration on daybreak. Since the countings were recorded continuously at a constant place along the root, they are actually monitoring the flow of radioactive K at that point. After a few hours the 42K flow through the root becomes constant when the solution temperature was 26°C. Its concentration in the root again increases with onset of darkness until a maximum is reached again at the next sunrise. At a root temperature of 34°C a similar pattern is observed, but the 42K concentration in the root continues to decline during the day. This observation suggests that either 42K uptake declined or that the water uptake was relatively 4 2 42 higher than that of K, causing a reduction of K concentration in the root. Long period exposures of roots to high tempera- tures above that optimum for root growth usually cause a sharp drop in plant growth (Cooper [1974/). Leakage of K out of the root might be one of the reasons for that decline in growth as well as shortage in soluble carbohydrate due to excess respiration (Ganmore and Kafkaf [1983/). 4. Salt stress Salt stress usually reduces plant growth. It is usually related to osmotic and specific effects (Levitt [19721). In general, 3 ways are known by which plants exclude Na from reaching the leaves: (1) by controlling Na influx and/or efflux at the plasmalemma of root cells (Jacoby and Hanson [1985]; Jescke [1970]; Nassery and Baker [1972]), (2) by removing Na from the xylem stream and sequestering Na in stellar parenchyma cells of roots and lower stems (Jacoby [19641; Johanson and Cheesman [19831) and (3) by retranslocating Na from shoots to roots via phloem and subsequent extrusion (Jacoby [1979]; Lessani and Marschner [1978]). Any stress that causes K leakage out of the cell will eventually lead to reduction in cell growth. Ben Hayyim et al. [1986] have shown that growth was linearly correlated with K content in callus cells of citrus roots. Increasing levels of Na in the external medium reduced [K] in the cell. Salt tolerant cells were able to hold the K in the vacuole against leakage when [Na] was increased in the external medium. Increasing [Ca] up to 10 mM reduced K leakage at very high [Na]. Similar effects were observed by Cramer et al. [19861. Kafkafl et al [1982/ have shown that increasing chloride concentration in the nutrient solution reduced nitrate uptake by the plant. In his review, Kafkafi [1984] has suggested that salinity reduces the amount of nitrogen transported to the meristematic growing points at the top and coined the term 87 was lower in NaCl-grown plants while the shoot content was the same as that of the control. Proline is known to accumulate in salt-stressed plants. This compound is an osmotic regulator against external decrease in water potential (Levitt [19721). The effect of K on ABA-induced proline production was studied by Pesci [19891. It was found that when the ratio of Cl - /K + in the barley leaf was 3/I proline production was increased. However when potassium content was lower so that the ratio was 5/1 the KCI stimulation was completely sup- pressed. There is no simple and explainable relationship between salt tolerance and K content in the plant at the higher concentration range. Qadar [19881 checked 11 varieties of rice showing tendency to high K accumulations at 4 levels of salinity. Despite wide ratios of Na/K, the contents of these ions in the plant could not explain the differential salt tolerance of these varieties. The same problem was recently studied by Figdore et al. [1989] in low-K stressed tomato plants (0.071 mM K) in nutrient solutions. When K is the limiting factor, the plant dry matter responds to increase of Na' in the solution (Table 2) up to 27.8 mM. This suggests that at low external K concen- tration the specificity of the cytoplasm to K is very high. There is a slight increase in plant dry weight in few varieties (Table 2) at the low [Na]. With increase in [Na] strain no. 576 did not respond beyond 0.45 mM Na while the other strains responded positively to increase in [Na] in the nutrient solu- tion. Na + could replace K + to some extent in its non-specific activity in the plant and this replacement is a characteristic of the variety (Table 3). The amount of dry matter produced per unit K in the plant (K efficiency) between varieties differed only by 32%. When the tomatoes were grown on sufficient K concentration the total dry matter was much higher but the relative yield was reduced by 40-50% due to high levels of [Na]. Comparing the absolute dry matter production at the highest Na level it was clear that increasing the external K concentration from 0.071 mM K to 2.8 mM K increased the yield from about 1.5 to about 5 g (Table 4). At the toxic levels of Na (87 mM), the tomato strains 576 and 571 that showed high substitution capacity under low-K stress did not appear to be the most salt tolerant under adequate K conditions. The nature of the mechanism involved in the accumulation and maintenance of high Na content in some varieties was not determined in that study. Root membrane composition has a profound effect on the prevention of Na 4 and C1- ions from entering the root and subsequently accumulating in the leaves of glycophytes (Kuiper [1968], [1980], [19871). The composition of root-cell membranes not only affects cation selectivity as mentioned above, but it has particular importance in preventing Cl- from entering the root. The ability of five varieties of grapes to tolerate salinity was positively cor- related with the solubility of chloride in the lipids of the membranes (Kuiper [19681). Enrichment of these grape varieties' root-cell membranes with phos- pholipid relative to their monogalactose diglyceride content limited chloride uptake. Story and Walker [1987] have shown that the superior tolerance of 88 Table 2. Mean total plant dry weight of 5 tomato strains grown under low-K stress (71 M K) at 8 Na levels in mM Na (Figdore el al. [19891) Plant dry weight (g)' Strain 0 .0 14 b 0.45 0.88 1.75 3.5 7.0 13.9 27.8 576 L.06a' 1.67a 1.61a 1.71a 1.67a 1.81a 1.70a 1.73a 571 0.94ab 1.32b 1.38b 1.46b 1.46b 1.50b 1.56b 1.67a 349 0.85bc 1.06c 1.08c 1.13c 1.19c 1.29c 1.36c 1.39b 203 0.80c 0.97cd 1.00c 0.99d 1.14c 1.1Od 1.27c 1.28bc 546 0.80c 0.85d 0.87d 0.92d 0.98d 1.02d 1.08d 1.19c a Four replications per strain at each Na level. b Control treatment with no added Na. LSDo.os=0.12 for strain means of individual Na levels from the analysis including the control treatment (comparisons are made within columns). Table 3. Mean Na accumulation per plant, as determin6d by solution analysis, of 5 tomato strains grown under low-K stress (71 M K) at 7 plus Na levels in mM Na.Values in parenthesis are the mg Na added to each pot (Figdore et al 119891) Na/plant (mg)a Strain 0.45 0.88 1.75 3.5 7.0 13.9 27.8 (18) (36) (72) 144) (288) (576) (1152) 576 9.9a 12.3a 19.3a 27.Oa 43.8a 66ab 129a 571 7.7b 11.6a 17.3b 25.3a 43.6a 75a 138a 349 3.7c 5.7b 9.7c 15.4b 30.4b 56bd 109b 203 3.1c 5.0b 8.8c 13.6b 29.1b 53c 108b 546 2.9c 4.9b 9.4c 15.5b 29.4b 56bc I lb LSDo.o 0.9 1.8 1.4 2.7 4.7 12 10 a Four replications per strain at each Na level. Tabel 4. Mean total dry matter dry weight of 5 tomato strains grown under low-K stress (0.071 mM) and adequate K condition (2.8 mM K) at 3 Na levels (mM Na) (Figdore et al [1989]) Plant dry weight (g) mM Na 0.014 0.014 3.9 3.9 87 87 mM K Strain 0.071 2.8 0.071 2.8 .071 2.8 576 ...... 1.43 9.97 2.20 11.87 1.55 4.04 571 ...... 1.14 12.24 1.87 13.33 1.73 4.89 349 ...... 1.16 11.65 1.60 12.37 1.62 5.68 203 ...... 1.26 10.78 1.70 11.44 1.53 5.06 546 ...... 1.09 8.79 1.16 8.62 1.35 4.17 89 citrus root stock Rangpur Lime to long term salinity was highly correlated with Cl exclusion from the leaves and was also associated with high selectivity of fibrous roots for K over Na. The selectivity of biological membranes to ions was shown to be determined on membrane parameters like: size and charge of the polar head group, chain length and degree of the fatty acid chains and interaction of phospholipid with sterol (Blok et al. [1977]). Lipid composition and metabolism is also affected by the presence of salts (Kuiper [1980]; Stuiver et al [1981]). Ion permeability and ATPase activity are both strongly influenced by membrane lipid composition (Cocucci and Bellarin- Denti [1981]; Douglas and Walker [1984]). Therefore lipids have the capacity to regulate ion movement into the roots through their influence on both pas- sive and active transport processes (Douglasand Walker [1984]). It is obvious that the fine structure and lipid composition of the membrane with its bound proteins has a basic genetic code. The expression of that code changes to a certain degree under the influence of salts in the external solution. It was found that a specific phospholipid in a platelet arrangement stimulates plant H * transport and its exchange with K (Scherer [1985]), thus not only the composition but also the fine structure of the phospholipids are involved in ion selectivity by root cells. Salt-induced changes in root lipid composition have been correlated with the relative abilities of different plant species and varieties within species to tolerate or adapt to a saline environment (Erdei[1980]; Stuiver et al [1981]). Douglas and Walker [1984] have demonstrated that changes in the composi- tion of free sterols of fibrous roots correlated well with chloride exclusion capacity of three citrus genotypes. The phospholipid composition of the root membranes has an important role in regulating the amount of Cl - that may enter the root. The energy required by plants to regulate their response to salinity stress was discussed by He/al [1982]. 5. Moisture stress The role of potassium in drought tolerance was reviewed by Saxena [1985. The plant response to drought is a very complex phenomenon that is being studied from various angles and disciplines (Turner and Kramer [1980], Mus- sell and Staples [1977]). Soil moisture influences K uptake by plants by affecting root growth rate and the rate of K diffusion in the soil. Mackay and Barber [1985] tried to resolve the effects of actual root growth as compared with the K diffusion rate as affected by moisture. At the lower side of the optimal soil moisture content, increasing soil moisture increased the effective diffusion coefficient, De, of K by about 2 times and therefore increased K uptake. Increasing the moisture content above the optimum resulted in slow root growth due to oxygen shortage. The reduction in root elongation was reflected in lower K 90 uptake. The rate of root elongation is a crucial parameter in uptake of nutrients that are strongly adsorbed to the soil and their concentration in the soil solution is usually very low. Combined effects of low temperatures and low moisture can be alleviated by increasing the concentration of K in the soil (Nelson [1980). 6. Pathogens The wide effects of fertilizers on plant susceptibility to fungi, bacteria virus and pest diseases was the subject of the 12th Colloquium of the International Potash Institute. There is a general trend observed in many crops to resist pathogen attack by increasing K content in the plant tissue. Reduction of nematode damage was observed in cotton (Oteifa and Elgindi [1976]) and suppression of Alternariasolani in tomato (Kirali [1976]). Increase of wheat tolerance to PucciniastriiformisWest was reported by Kovanci and Colakoglu [19761. Shortage of K in rice is related to many physiological disorders (Ismunadji [1976]) and diseases (Trolldenier and Zehler [1976]). Increased K content in oil palms increased the resistance of this crop to Fusarium (01- lagnier and Renard [1976]). 7. References Adams, P: Effect of nutrition on tomato quality. The Grower, May 18, 1142 (1978) Ben-Hayyim, G., Kafkafi, U. and Ganmore-Newman, R.: The role of internal potassium in maintaining growth of cultured citrus cells on increasing NaC and CaC12 concentra- tion. Plant Physiol. 85, 434-439 (1987) Bergmann, E. and Bergmann, H. W.: Comparing diagrams of plant/leaf analysis present- ing by rapid inspection the mineral nutrient element status of agricultural crop plants. Potash Review, Subject 5, No. 2, 1-10 (1985) Beringer, H. and Troldenier, G.: The influence of K nutrition on the response of plants to environmental stress. Potassium Research-Review and Trends, Proc. llth Congress of the International Potash Institute, 189-222, Bern, Switzerland (1980) Blok, M. C., van Deenen, L. L. M., de Gier, J., Op den Kamp, L A. F and Verleij, A. J: Some aspects of lipid-phase transition on membrane permeability and lipid-protein as- sociation, Biochemistry of membrane transport (G. Semenza and E. Carafoli, eds.) FEBS-Symposium No. 42, Springer-Verlag, Berlin. p. 38-46 (1977) Boatwright, G. 0., Ferguson, H. and Sims, J R.: Soil temperature around the crown node influences early growth, nutrient uptake and nutrient translocation of spring wheat. Agron. J. 68, 227-231 (1976) Cocucci, M. and Ballarin-Denti, A.: Effect of polar lipids on ATPase activity of membrane preparations from germinating radish seeds. Plant Physiol. 68, 377-381 (1981) Cooper, A. J: Root temperature and plant growth. Research Review No. 4, Commonwealth Bureau of Horticulture and Plantation Crops, East Mailing, England (1974) Cramer, G. R., Lduchli, A. and Spolito, V S.: Displacement of Ca2 by Na4 from the plasmalemma of root cells. A primary response to salt stress? Plant Physiol. 79, 207-211 (1985) 91 Douglas, T J. and Walter, R. R.: Phospholipids, free sterols and adenosine triphosphate of plasma membrane-enriched preparations from roots of citrus genotypes differing in chloride exclusion ability. Physiol. Plant. 62, 51-58 (1984) Erdei, L., Seuiver, C E. E. and Kuiper, PJ. C.: The effect of salinity on lipid composition 2 2 and on activity of Ca +- and Mg " -stimulaled ATPase in salt-sensitiveand salt-tolerant plantago species. Physiol. Plant. 49, 315-319 (1980) Figdore, S. S., Gerloff G. C. and Gabbelman, W. H.: The effect of increasing NaCI levels on the potassium utilization efficiency of tomatoes grown under low-K stress. Plant and Soil 119, 295-303 (1989) Gantnore-Newman, Ruth and Kafkafi, U.: The effect of root temperature and 4 NO /NH 4 ratio on strawberry plants. 1. Growth, flowering and root development. Agron. J. 75, 941-947 (1983) He/al, H. M.: Interaction of potassium nutrition and salt tolerance in higher plants. Proc. Int. Workshop Role of K in Crop Production, Cairo 1979. Eds. A. Saurat and M. M. EI-Fouly, Int. Potash Institute, 1982 Ismunadji, M.: Rice diseases and physiological disorders related to potassium deficiency. In: Fertilizer use and plant health, Proc. 12th Coll. of the International Potash Institute, Izmir/Turkey, 33-46 (1976) Jacoby, B.: Function of bean roots and stem in sodium retention. Plant Physiol. 39, 445-449 (1964) Jacoby, B.: Sodium recirculation and loss from Phaseolus vulgaris L. Ann. Bot. 43, 741-744 (1979) Jacoby, B. and Hanson, J B.: Controls on 2Na + influx in corn roots. Plant Physiol. 77, 930-934 (1985) Jeschke, W. D.: Evidence for a K *- stimulated Na + efflux at the plasmalema of barley root cells. Planta 94, 240-245 (1970) Johanson, J G. and Cheesman, J. M.: Uptake and distribution of sodium and potassium by corn seedlings. 1. Role of the mesocotyl in 92 Lessani, H_ and Marschner, H.: Relation between salt tolerance and long-distance transport of sodium and chloride in various species. Aust. J. Plant Physiol. 5, 27-37 (1978) Levitt, .: Responses of plants to environmental stresses. Academic Press, New York, 1972 Mackay, A. D. and Barber, S. A.: Soil moisture effects on potassium uptake by corn. Agron. J.77, 524-527 (1985) Marschner, H.: Mineral nutrition of higher plants, Academic press, 1986 Mussel, H. and Staples, R. C: Stress physiology in crop plants. Proc. Intern. Conference on stress physiology in crop plants, Boyce Thompson Institute, Ithaca N.Y., 1979 Nassery, H. and Baker, D.A.: Extrusion of sodium ion by barley roots. 11. Localization of the extrusion mechanism and its relation to long distance sodium ion transport. Ann. Bot. 36, 889-895 (1972) Nelson, W. L.: Interaction of potassium with moisture and temperature. In: Potassium for agriculture-A situation analysis. Potash and Phosphate Institute, Atlanta, GA, USA, 109-119 (1980) Plesci, P.: Involvement of Cl- in the increase in proline induced by ABA and stimulated by potassium chloride in barley leaf segments. Plant Physiol. 89, 1226-1230 (1989) Ollagnier, M. and Renard, J. L.: The influence of potassium on the resistance of oil palm to Fusarium. In: Fertilizer use and plant health, Proc. 12th Coll. International Potash Institute, Izmir/Turkey, 157-176 (1976) Oteifa, B.A. and Elgindi, A. Y: Potassium nutrition of cotton, Gossipium barbadense, in relation to nematode infection by Meloidogyne incognita and Rotylenchulus reniformis. In: Fertilizer use and plant health, Proc. 12th Coll. International Potash Institute, Izmir/Turkey, 301-306 (1976) Qadar, A.: Potassium status of the rice shoot as an index for salt tolerance. Indian J. Plant Physiol. 4, 388-393 (1988) Saxena, N. P: The role of potassium in drought tolerance. Potash Review 16/102 (1985) Scherer, G. F E.: 1-alkyl-2-acetyl-sn-glycero-3-phosphocholine (platelet activating factor) stimulates plant H + transport in vitro and growth. Biochem. Biophys. Res. Comm. 133, 1160-1167 (1985) Story, R. and Walker, R. R.: Some effects of root anatomy on K, Na and Cl loading of citrus roots and leaves. J. Exp. Bot. 38, 1769-1780 (1987) Sluiver, C. E. E., Kuiper,. PJ C., Marschner, H. and Kylin, A.: Effect of salinity and replace- ment of K + by Na + on lipid composition in two sugar beet inbred lines. Physiol. Plant. 52, 77-82 (1981) Termaat, A. and Munus, R.: Use of concentrated macronutrient solutions to separate os- motic from NaCI-specific effects on plant growth. Austr. J. Plant Physiol. 13, 509-522 (1986) Trolldenier, G. and Zehler, E.: Relationships between plant nutrition and plant diseases. In: Fertilizer use and plant health, Proc. 12th Coll. International Potash Institute, Iz- mir/Turkey, 85-93 (1976) Turner,N. C. and Kramer, P.1: Adaptation of plants to water and high temperature stress. Proc. Seminar held November 6-10, 1978, Carnegie Institute of Washington, Dept. of Plant Biology, Stanford, California, 482 pp., 1980 Winsor, G. W.: Potassium and the quality of the glasshouse crops. Proceedings 8th Con- gress International Potash Institute, Brussels, 303-312 (1968) Yermiyahu, U. and Kafkafi, U.: Yield increase and stem brittle decrease in response to + increasing concentrations of potassium and NO3-/NH 4 in White Carnation CV. Standard. Hassadeh, 90, 742-746 (Hebrew with English summary) (1990) 93 Coordinator's Report on the 1st Working Session Dr. V V Prokoshev, Laboratory of Potash Fertilizers, Institute of Fertilizers, Leningskyprospect 55, 117333 Moscow, USSR The papers presented during this session embrace a wide spectrum of problems, beginning with assessments of the State's efforts to advance the national agriculture, down to the analysis of intricate processes occuring in the structural elements of the plant cell. Nevertheless, all these papers can be united by their one general conclusion to the effect that application of K-fertilizers is not only the major condition for producing high and steady crop yields, but also an index characterizing the general level of farming. Breburda has initiated a difficult task - to analyze the reasons behind the very different efficiency of fertilizers in the countries of West and East Europe, especially in the USSR. Indeed, the traditional allusions to the unfavourable weather conditions prove to be incorrect, because under equal conditions some farms raise crops which are considerably larger than those produced by others. A number of quite sound reasons can be mentioned additionally, but only one general conclusion can be made: the level of crop yields and the efficiency of fertilizers depend upon a great many factors governing the level of farming management. We should agree with the speaker that under the conditions of market econ- omy most impressing results have been obtained in the agriculture of West Europe. Scientists and practical field workers in this country are all too aware of the drastic degradation of soil and losses of organic matter which, as the speaker justly says, reach 0.5 to 1.0 t/ha/year. Where radical measures against these processes have been taken, in particular in Byelorussia, and where large amounts of organic manures have been used, it became possible to weaken these processes and in some areas even to enlarge the organic matter reserves. However, in the main zone of chernozems these negative processes are still going on, badly affecting soil properties. Van Diest cited the example of the Netherlands and went on to develop the idea advanced at the 13th Congress of the IPI by Cooke about the State and intercontinental balance of plant nutrients. Indeed, in the light of the ever expanding trade contacts and specialization of agriculture in terms of the regions, these calculations become especially significant for assessing the total conjuncture of consumption and production of fertilizers. It seems feasible to make such calculations in the USSR in terms of the republics and regions, both for the near and more distant future. 95 The positive K-balance developed in the Netherlands, owing to imported fodders and excessive organic fertilizers, is rather specific, involving regions with exceptionally intensive animal husbandry. Kerschberger and Richter have shown that for the conditions of the GDR, agricultural productivity depends closely on generous use of K fertilizers. The result of following recommendations emanating from the programmed in- terpretation of soil analysis at four year intervals has been a marked reduction in the proportion of low K soils. The available potassium content of arable surface soils has increased; though K applied in fertilizer was less than that removed in crops, the K balance was generally positive if the K in organic manures was included. The Egner-Riehm method is preferred for soil analysis. Florinsky and Yefremov have demonstrated their methods for assessing the dynamics in the variation of exchangeable K content of arable soils in the USSR. Their paper reflects the concept of the integral State agrochemical survey in this country. The results of this survey intended for thought to be taken in siting central fertilizer distribution points and for locating fertilizer production facilities. 25 years of this work has indicated the potential fertility of soils in various regions and shown how fertility has been affected by level of farm management. The paper presented by Kafkafi is an example of the well-known idea that there is nothing more practical than a good theory. Based on his detailed study of the protective role of lipids and their compounds in cells, correlations of cations in tissues and the effects of cations on the consumption, transport and exchange of ions, the author comes to a practical and essential conclusion about the positive, and occasionally most decisive, role of potassium in extreme situations. The latter may be due to variations in the temperature, water, salt and pathogenic regimes and their effects on plant growth. When the K concentration in the nutrition environment is increasing, all factors restricting the rate of root growth, and consequently, augmenting the K concentration in plant tissues, are found to be manifested to a much smaller extent. All-in-all the papers given in this session convince us of the need for regular and strict control of the potassium regime for the production of reliable and high crop yields. 96 Chairman of the 2nd Session I. M. Bogdevitch, Byelorussian Institute for Soil Science and Agrochemistry, Minsk, USSR 2nd Session Soil Types and K-dynamics 97 Soil Properties and Potassium Behaviour V V Prokoshev and TA. Sokolova Summary Application of potash fertilizers can be rational only when based on thorough study of factors responsible for the state of potassium in soil. The content of the main potassium-bearing minerals (illites) contained by the clay frac- tion in the upper horizons is shown to increase steadily from north to south, from podzolic soils to chernozems. This must always be taken into account when planning any application of potash fertilizers. Consideration of changes in the thermodynamic parameters of soil due to intensive con- sumption of soil potassium by plants and long-term application of potash fertilizers leads to the conclusion that it is necessary to calculate more accurately the rates of potash fer- tilizers to be applied so as to preserve the natural mineral structures. The significance of studies revealing the spatial distribution of roots within soil profile is confirmed by various data on migration of potassium and of accompanying elements obtained in soil colums, lysimeters and long-term field experiments. Only on sandy soils did the potassium content vary below I m depth. Potassium migration is shown to depend particularly on the ratio between nutrients in applied fertilizers and on the annual potassium balance. Unbalanced application of fertilizers to sandy soils leads to nonproductive accumulation of echange- able potassium down to the depth of 150 cm. The chloride content was uniform to the depth of 3 m with insignificant but stable distinction according to the rates of potassium chloride applied. Data on non-recoverable potassium losses as a result of erosion processes are presented. Emphasis is placed on the need to construct fertilizer policy on a scientific basis taking zonal-specific features of soils into consideration. 1. Introduction Scientifically based application of mineral fertilizers should solve both eco- nomic and ecological problems of natural regions. One should regard as the optimum such a potassium regime in soil which provides for high fertility of the root-zone at an economic cost. At present, the average productivity of arable soil in West European coun- tries has reached 3.4-4.0 t/ha cereal units; in some countries it is as high as 10 (FRG) and even 15 t/ha (Belgium) (IFC [1988]). Such yields are achieved by using perfect agrotechnical methods and mineral fertilizers, including * Dr. V. V. Prokoshev, Laboratory of Potash Fertilizers, Institute of Fertilizers, Lenin- skyprospect 55, 117333Moscow, USSR and Dr. TA. Sokolova, Faculty of Soil Science, Department of Soil Chemistry, Moscow State University, 117234 Moscow, USSR 99 potash. On average of all crops approximately 10-12 g K20 are needed to produce I cereal unit (c.u.) equally 1 kg of dry matter. On the average, during the recent years in the USSR arable soils yield 1.5 to 1.8 t/ha c.u. and receive about 30 kg/ha K20, the actual figures varying greatly in different regions. Approximately 18 kg of fertilizer is applied to obtain I c.u. (Landwirtschaft der UdSSR [19881). This situation should be thoroughly analyzed on the basis of generalized experimental data, so that it would be possible to provide differentiated rates of fertilizers and predict prospects for further progress in the production of potash fertilizers with regard to specific features of soils in the USSR. The behaviour of potassium in soil is a rather intricate process including a number of particular mechanisms. This paper limits discussion to the fol- lowing problems: potassium-bearing minerals as modified by various factors; some aspects of potassium dynamics in the soil-plant system; migration of potassium and of accompanying elements in the soil profile. Potassium be- haviour in soils in the conditions of arable ecosystems is mainly considered for sod-podzolic soils. The general conclusion is that the problem of effective use of potash fer- tilizers can be solved only when all factors, especially soil factors are taken into account. 2. Potassium-bearing minerals within the zonal series of soils in the USSR In the USSR, soils of the zonal series, on loamy and clay deposits are quite widespread. The main potassium bearing minerals found are finely dispersed illites, most being concentrated in clay fraction. Most illites in soils (Nieder- budde [19751) developed on mantle loess-like deposits, are of secondary ori- gin. They were formed due to the Holocene irreversible fixation of potassium by labile 2:1 clay minerals. As a result, the potassium-fixing capacity of soil, as compared with that of soil-forming rock, is greatly decreased. In most cases the changes occuring in potassium-bearing minerals in soil can bejudged only by using some indirect parameters. Quite useful is informa- tion about the profile distribution of illites and the ratio of the intensities of reflexes 1001:1002 of illites along the soil profile within the zonal series of soils. These data (Table 1) show that from north to south, from gley-podzolic soils to chernozems and solonetzs, one can observe regular unidirectional changes in the profile distribution of illites and in the ratio of their reflex intensities. The ratio of illite content of the clay fraction of surface soil to that of subsoil increases progressively in northern gley-podzolic and podzolic soils. Especially on varve clays with high illite contents, horizons A2 clearly have less illite in their clay fraction, as compared with the underlying horizons. 100 Table 1. Content of illites and correlation between the intensities of reflexes in the first and second orders of illites in some zonal types of soils in the USSR Soil Horizon Depth Content of 1oo of illites (cm) illites* 1oo2 Gley-podzolic on mantle loam A2 6- 10 11/2 1.2 (Zabojeva [1975]) B, 24- 57 12/2 1.7 B2 57- 78 13/4 1.6 Podzolic on varve clay Ai, A2 12- 18 32/4 2.5 (Sokolova [19851) At 18- 24 34/5 3.0 B2 70- 80 39/9 4.1 B3 125-135 42/8 4.4 Podzolic on mantle loam A, 9- 11 24/4 1.3 (Sokolova [1985]) AZ I- 20 26/2 1.6 B2 120-140 25/8 3.0 B3 210-230 25/8 3.0 Sod-podzolic on mantle loam A,, A2 2- I1 26/4 1.7 (Targulyen 11974]) A2 I1- 27 27/3 1.6 B2 120-140 26/7 2.4 B3 190-207 24/7 2.4 Grey forest on loess-like loam A,, A2 0- 6 35/4 5.6 (Zwjagin 119851) A2, B, 24- 38 30/11 3.9 B2 50- 60 30/13 3.8 B3 60- 88 30/12 3.7 Southern chernozem on loess-like Al 0- 8 41/12 3.3 loam (Kornbluim et aL [19721) A.' 8- 19 42/15 3.0 AB 27- 42 41/16 2.5 The illite content was assessed by the amount of K2O in gross chemical composition of clay fraction through multiplication of this value by 10; i.e. from assumption that K20 content in illites equals 10%. The numerator shows % illite in the clay fraction, the denominator shows % illite in the soil including the clay content. In sod-podzolic soils, illites in the clay fraction are distributed uniformly in the profile. In soils of the forest-steppe and steppe zones, namely grey forest soils, chernozems, solonetzs, the illite content in clay is highest in the upper horizons, gradually decreasing with depth. The profile distribution of illites in soil seems to be controlled by several opposing processes. When illites transform into labile structures and are des- troyed, the content of these minerals in the clay fraction, and in the soil as a whole, must necessarily decline. The results are inverse when illite-like struc- tures develop due to irreversible fixation of potassium by labile structures and when illites, composing the larger fractions, are physically crushed down to clay particle size. Such illite-like structures can considerably replenish the reserves of illites both in the clay fraction and in soil. 101 One can suggest that the first two processes are most intensive in acid north- ern podzolic soils, leading to reduced amounts of illites in the upper horizons. Moving from north to the more southern soils which are less acid and in which the biological activity is higher, the process of illitization increases because of the greater quantity of potassium ions. Sources of this potassium are plant residues and the weathering of potassium-bearing minerals of the parent rock. As a result, these horizons accumulate secondary pedogenic illites (Sokolova [19851). The above ideas about changes occuring in potassium-bearing minerals within the zonal series of soils are confirmed by data on the illite intensities ratio Iooi:1002 in the profile under examination. The relationship between these intensities may increase with the decreasing content of interlayer potassium and the increasing amount of iron in octahedral positions. Transformation and destruction of illites in horizons A 2 lead to partial losses of octahedral iron; as a result, the ratios between the first and second orders of illites in these horizons decline. In soils of the forest-steppe and steppe zones the processes of illitization result in less crystallized structures with lower content of potassium. These structures develop in the upper horizons and their potassium content is lower than in primary illites derived from the rock. That is why illites in the upper horizons reveal more profound relationships between the intensities of reflexes of the first and second order. Based on the above, it can be assumed that the discussed features of illite behaviour within the zonal series of soils considering also the texture differen- tiation of profiles in podzolic soils, may help augment the efficiency of potash fertilizers on podzolic soils, as compared with soils in the forest-steppe and steppe zones. 3. Potassium status of soils as influenced by various factors The anthropogenic effect upon different soil properties, including the state of potassium in soil, may be extremely diversified. Such an effect is frequently greater than that of any natural factor. Organic manures may affect the potassium status in soils in various ways. Several mechanisms may act in opposite directions (Goulding et al. [19841). Organic matter is known to be more selective to polyvalent cations than to potassium (Talibudeen [1981]). Together with organic fertilizers soil may receive considerable quantities of K + and NH4 ions. The latter are first sorbed on exchange positions which are most selective to potassium. It is also possible that exchange positions of clay minerals are blocked by films of organic mat- ter. This is particularly likely at pH < 6 when aminogroups of organic matter are partially protonized and are more strongly bound to the exchange posi- tions on clay minerals. All these effects of organic matter will necessarily minimize the selectivity of soils to potassium. 102 Organic matter interacts with clay minerals primarily on the external planar positions. Therefore the relative importance of internal exchange positions on clay minerals, which prove to be most selective to potassium, may increase. As a result, when a soil receives organic manures, its selectivity to potassium is observed to increase (Poonia et aL [1986]). Liming usually decreases the potassium supply to plants. This can be ex- plained by the impact of several unidirectional mechanisms interacting simul- taneously. First, even though the amount of available potassium may remain unchanged, the absolute concentration of potassium may be insufficient for the plant whose growth is improved by liming. Second, liming decreases the relative share of potassium in the cation exchange complex, while the absolute quantity of potassium does not change. In addition, liming leads to an in- crease in the potassium-fixing capacity of soil and of the value PBCK due to the release of interlayer-Al and thus an increase of free negative interlayer charges which are highly selective for potassium (Niederbudde et al. [1981]). Increased values of PBCK due to liming were found in samples of arable horizons in light loamy sod-podzolic soils at the Ramenskoye experimental station. Lime-induced increase in PBCK was observed under all treatments (Table 2) but it was particularly great when no potash fertilizer was applied. Table 2. PBCK in limed and unlimed sod-podzolic soil K PBC Treatment With lime Without lime 0 ...... 37 22 N P ...... 50 33 NPK (potassium chloride) ...... 38 36 NPK (400o potassium salt) ...... 32 28 The impact of fertilizers on the potassium status in soil may differ, depend- ing upon the soil properties, the type and rate of fertilizer. When plants take up potassium, the soil solution may become depleted of K and consequently micas and illites transform into labile structures and release potassium (Moberg et aL [1983]). This, however, occurs at low rates possibly limiting growth rate and crop yield. When potassium concentration in soil solution exceeds a certain limit, one observes a different process, namely that of illitiza- tion due to potassium fixation by labile minerals. In the territory of the USSR this process is most significant in soils of the western and north-western regions of the European USSR where the fine earth contains large quantities of vermiculite (Riabzev [19761). Sometimes the mineralogical composition of fine soil fractions is found to be distinctly dominated by rigid structures, such as kaolinites, weathering- resistant illites, chlorites and chloritized structures in which the interlayers 103 are filled with bands of aluminium hydroxide. In such rigid structures applica- tion of fertilizers may not alter clay minerals. For a long time, scientists have been paying special attention to the effects of potassium depletion and potassium inputs from fertilizer respectively on the potassium potential of the soil, which characterizes the strength of bonds between potassium and the soil absorbing complex, and the value of PBCK. Many authors consider the latter parameter to be sufficiently stable (Beckett et al. [1967]); others provide evidence that it largely depends on the amount of potash fertilizer applied (Schaimuchametov et al. [1987]). The above statement can be supported by the results of our own investiga- tions of samples taken from the plough layer during long-term field experi- ments which investigated the effect of rate and form of potash fertilizer on sandy loam and heavy loam sod-podzolic soils (Table 3) (Orlova et al. fin press]). Table 3. Some parameters of potassium status in soil as affected by potash fertilizers F AKo Treatment AR, . 10' (kcal/mol) (me/100 g) PBCK Long-term field experiment on sandy loam (A): O ...... 4.3 3587 0.12 28 NP ...... 2.2 3590 0.06 27 NPK 270o ...... 9.2 2740 0.23 25 Long-term field experiment on heavy loam (B): O ...... 1.6 3748 0.16 100 NPK 6o ...... 2.2 3590 0.19 86 NPK270 ...... 9.7 2699 0.18 18 Pot experiment on soil from the treatment of the long-term field experiment on sandy loam: a) NP ...... 1.9 3718 0.027 14 NPK 6 ...... 2.8 3478 0.060 21 NPK 6o* ...... 2.2 3590 0.042 19 b) NP ...... 0.7 4285 0.010 14 NPK 6o* ...... 0.8 4190 0.018 23 NPK 6o** ...... 1.3 3900 0.028 22 a) before pot experiment b) after experiment * potassium chloride ** kalimagnesia Sandy loam soils in the Lyubertsy experimental field (A) are dominated by rigid structures, strongly chloritized minerals and dioctahedral illites com- prising the fine fractions. In heavy loam soils of the Dolgoprudnaya experimental station (B) up to 30% of their fine fractions are represented by labile minerals of the mont- morillonite group, the rest being composed by kaolinite, illite and chloritized 104 structures. All soils were low in organic matter, slightly acid and reasonably uniform. Distinct differences were observed in the reactions of the individual soils to long-term potassium depletion on NP treatments and to potassium enrich- ment when potash fertilizers were applied. In the experiment B, because of considerable quantities of labile minerals in the fine soil fractions and high PBCK, the amount of readily exchangeable potassium (A K.) did not increase when soils received potash fertilizers, in contrast to experiment A where it increased. The values of all parameters indicating soil K status changed according as to whether the rate of potash fertilizer was sufficient or insufficient to com- pensate the losses of potassium removed in crops. This was particularly obvi- ous when soils were seriously depleted of potassium in the pot experiment. Kalimagnesia had a more beneficial effect on soil potassium than potassium chloride. On heavy loam soils the value of PBCK undergoes greater changes than on sandy loam ones. Such changes in PBCK can largely be attributed to trans- formation of illites into labile minerals due to potassium depletion of soil by plants, as well as to illitization resulting from irreversible potassium fixa- tion by labile structures when the rates of potash fertilizer applied exceed the requirements of plants or when the competition of labile structures for K is greater than that of the roots. The results of our investigations once again confirm the necessity for a careful approach to assessing the rate of potash fertilizer to be applied. Both excessive and insufficient rates are to be avoided if fertilizer usage is to be rational. 4. Potassium and accompanying elements in soil profile 4.1 Behaviour of potassium in the soil-plant system One indication of the rationality of a chosen policy regarding potash fertilizer is the distribution of available potassium in the soil profile. The soil profile is an integrated complex system and it would be wrong to discuss fertility of any profile layer separately, because individual nutrients, occuring at different depths, usually supplement one another. Quite demon- strative is our observation that potassium uptake by plants from the plough layer and deeper layers of soil depended on their content of available phos- phorus. For example, when these horizons of loamy soil under the NP treat- ment contained equal amounts of exchangeable potassium, our greenhouse experiment showed that equal quantities of potassium were taken up from both layers, while under the NK treatment almost twice as much K was taken up from the plough layer due to the obvious deficit of phosphorus in the lower layer. When assessing a certain soil as an integrated system, it is absurd to examine the availability of potassium reserves in separate horizons of soil profile. 105 In this connection it is practically important to elaborate theoretical esti- mates of spatial distribution of nutrients within the soil profile and to study the progress of the root system (Sokolov [1947]" Mengel et al. [1985]). Such knowledge will enable us to improve both the technology and the machines for application of fertilizers. Hopefully, it will be possible to produce special strains of crops with well developed root systems and to define more accurately the effect of crop rotations involving crops with root systems penetrating to different depths and of different cation-exchange capacity. It was found long ago that roots penetrate to different depth, for instance to 1.5 m in potato, to 2 m in cereal crops, to 3 m in clover (Russel [1950]). As a rule, only 30016 of the whole root system is found below the plough layer. Nevertheless, these roots can provide one third of all nutrients taken up by plants. According to data collected in 17 field experiments with radioactive potas- sium, there were 2697o of roots beyond the 30 cm layer, but they, on the average, provided 35o of potassium uptake, and in some soils - over 50076 (Kuhlmann [1986]). Thus, we cannot say that potassium, enriching the subarable horizon, is lost. Potassium, occuring in a deeper humid layer reduces the susceptibility of crops to detrimental effects of external factors like frost, drought, etc.; it is a reliable source. In view of the above, some ecologists advance theories aimed at minimiza- tion of fertilizer application (van Noordwijk et al. [1986]). They claim that it would be necessary to find 106 reserves? If it has a low K-buffering capacity, one can hardly hope to raise high crop yields. 0 NP L_ NPK "1100 a 4" 80 4B0 v3- 60? 01 ?2- -40 oi 20 II I tillering shooting heading 14 days 15 days 13 days Figure I. Effect of K fertilization on K 20 uptake and daily dry matter production by barley during different development phases (average of 4 years). This was examined by analysing soil samples taken at intervals and we found that the rate of release depends on the soil properties and on the initial content of exchangeable potassium. The results of our research are presented below: Soil Treatment K 1ch. (pp.) K20 released before experiment (kg/ha per day) Loamy soil NP ...... 62 0.57 N PK ...... 119 0.71 Clay soil NP ...... 79 0.71 NPK ...... 142 1.28 Long-term application of fertilizers helped to raise the potassium content in soil and improve its capacity to restore the exchangeable K content of the soil. However, even taking into account the capacity in subarable horizons, rates of potassium release are not sufficient to provide plants with enough potassium during extreme periods. It would be desirable to have for each group of soils some experimental data specifying the capacity of the particular soil to restore the disturbed potassium equilibrium. 107 In this connection, the correlation of the rate of potassium desorption from soil with the rate of potassium uptake by plants is practically important (Nemeth [1972]). Using these data, one could predict the potential abilities of soil and possible crop yields. 4.2 Some problems of potassium migration in the soil The limits to which nutrients can be taken up by the root system of plants depend on the amount of these nutrients in their available forms occuring in the soil profile. There are several factors which affect the content of potassium in the soil profile. Among such factors are: the percent of clay in soil, the duration of fertilization, the rate of fertilizer, the amount and composition of accompany- ing fertilizers, the level of potassium removal by plants and (or) the potassium balance. Soil column experiments have shown that potassium filtration rates on sandy soils are several times higher than those on clay soils (Rhoads [1971/). It is of practical importance that the mobility of all basic cations is increased by the application of nitrogenous fertilizer. This has been confirmed in our experiments with soil colums (Table 4). Table 4. Effect of K and N on cation leaching on different soils in mg/kg soil Applied Leached Treatment (me/1 00 g) Soil K Ca Mg Na O ...... 1.2 28.9 2.1 2.3 2 2.7 32.3 5.3 4.8 3 2.5 54.3 13.8 10.1 K * ...... 2.13 1 20.0 57.9 4.3 4.8 2 18.3 50.0 8.4 7.6 3 11.5 72.3 17.7 13.0 K +Na ...... 5.80 1 38.3 108.1 9.0 7.4 (2.13 + 3.67) 2 32.6 83.0 15.1 10.2 3 20.7 115.1 28.2 16.0 * K,=potassium chloride, Na=ammonium nitrate l=sandy loam soil, 2=loamy soil, 3=heavy loam soil Our long-term experiment on sandy soil has demonstrated that the interac- tion between potassium and nitrogen fertilizers applied to soil is synergistic. When potassium chloride was applied at 60, 120 and 240 kg/ha K20 and the rate of applied N was 60 kg/ha, as urea, the sum of exchangeable cations in soil grew respectively, in terms of each ton of fertilizer (K 20), by 0.23, 0.14 and 0.034 me/100 g (Table 5). At the same rates of potassium, but with nitro- gen applied at 240 kg/ha, the sum of exchangeable cations declined by 0.41, 0.34 and 0.40 me/100 g. Higher rates of N together with higher rates of K decreased the amount of absorbed cations in soil (per I ton N) by 0.57, 0.68 and 0.94 me/100 g respectively. 108 Table 5. Effect of N and K fertilizers on exchangeable cations in a sandy soil during 10years (me/100 g) Treatment Exchangeable Effect from I ton K20 or N cations K fertilizers N fertilizers Nwo 4.19 - - N6O K60 4.33 +0.23 - No K12o 4.36 +0.14 - N6o K240 4.27 +0.034 - N240 3.54 - -0.36 Nuo K"o 3.29 -0.41 -0.67 Nzo K 120 3.13 -0.34 -0.59 N24o K240 2.56 -0.40 -0.94 * kg N and KzO/ha/year Using lysimeters, one can sufficiently quantify the losses of potassium and of accompanying elements due to fertilizer application. Long-term lysimeter studies have been done in practically all countries of Europe. On the whole, the results obtained are rather close, though they depend on the soil properties, intensity of fertilizers and their specific use, the duration of observations over particular crops, and on other conditions. According to the results of experiments carried out in the FRG over 50 years, the losses of potassium, on the whole, were not very great, less than 20 kg K20/ha (Jirgens-Gschwind and Jung [1979]). The filtration experiments on heavy soils in the GDR revealed annual losses of 7-9 kg/ha K20 and up to 25 kg/ha on light soils (Amberger et al. [1979]). Potassium mobility may also depend on the degree of weathering of minerals. Drainage waters from young soils are richer in potassium and show a more variable ion composition of the soil solution. In England, potassium losses from the 90 cm layer of soil under grass were only 1.9 kg/ha K 20. Potas- sium losses are not at all comparable with losses of Ca and Mg which may reach 300 and more kg/ha annually (Cooke [1967]). The intensity of K + and CI - migration, as it depends on soil properties is shown in Table 6 for a lysimeter experiment which we carried out during 4 years, involving cylindrical lysimeters 90 cm in diameter filled to a depth of 72 cm. For the 4 years of the experiment the sum total of precipitation within the frost-free season was 1720 mm, beginning from May when fertilization started. 69% of applied potassium was washed out from clean quartz sand with no plant cover. When sand in the upper layer was substituted by medium loam soil containing 30% clay, the amount of potassium drained was tripled. When the lysimeter was completely filled with soil, only 17% of potassium from potash fertilizers was washed out because K was held in the lower horizons. Chloride was leached out completely; the soil properties affected only the shift of the maximum to a later period. 109 Table 6. Elements leached for different periods without plants in g/lysimeter (mean of 4 years) K Cl Layer Lysimeter (cm) I 2 3 Total 1 2 3 Total I Sand ...... 0-72 28.4 10.9 4.4 43.7 33.2 15.8 4.6 53.6 (69*) (94) 2 Soil+...... 0-20 6.0 4.7 3.7 14.4 25.0 16.6 11.1 52.7 sand ...... 20-72 (23) (92) 3 Soil ...... 0-72 4.7 2.7 3.4 10.8 11.9 14.1 24.9 50.9 (17) (89) Note: I= May to August; 2 = September to November; 3 = December to April = 07oof applied quantity Another group of lysimeters contained plants and received three levels of NPK. The lysimeters were used to calculate the balance of elements and to establish the share between plants, soil and the leachate. The relative distribu- tion of phosphorus, calcium, chloride and sodium was quite constant, though different for each element. The percent of potassium was found to decrease in plants as the fertilizer rate increased, to increase in soil and to decline in filtrate (Table 7). Table 7. Balance of nutrient elements in lysimeters as affected by rates of fertilizers Elements Input Consumption Balance Plants Soil Filtrate g/m2 %7 of consumption NIPK, N ...... 57.0 61.0 - 4.0 59 3 38 P ...... 24.3 21.1 3.2 19 80 1 K ...... 47.4 49.9 - 2.5 61 8 31 Ca ...... 227.5 153.0 74.5 8 67 24 Na ...... 3.3 18.0 -14.7 6 14 80 Cl ...... 45.4 65.9 -20.5 39 15 46 N 2P 2K2 N ...... 110.9 87.4 23.5 53 12 35 P ...... 48.1 34.8 13.3 18 82 0.4 K ...... 92.2 80.8 11.4 50 26 24 Ca ...... 245.5 148.8 96.7 11 50 39 Na ...... 5.6 21.9 -16.3 8 12 80 Cl ...... 85.8 102.0 -16.2 34 24 43 N3P 3K, N ...... 219.1 141.3 77.8 39 23 38 P ...... 95.1 68.4 26.7 13 87 0.4 K ...... 181.8 142.3 39.5 43 48 9 Ca ...... 281.9 197.3 84.6 10 44 47 Na ...... 10.1 29.5 -19.4 7 26 67 CI ...... 166.6 182.5 -15.9 23 31 45 Note: Input= fertilizers+ precipitations+ seeds Consumption =leached by plants+found in soil+in filtrate 110 The same 4 year series of lysimeter experiments studied the leaching of basic elements as influenced by the main forms of potash fertilizers. Leaching of potassium was reduced by 10% and that of calcium and magnesium in- creased when 40% K2 potash salt was applied. The rate of removal of bi- valent cations increased when the rate of chloride was higher. Potassium leaching was increased by the different N fertilizers in ascending order as follows: urea < ammonium sulfate < ammonium nitrate Lk 0-20 cm 0-20 cm a400- x 20-40 cm 0 0 40-60 cm C -300- cc 250-2 9. 200 -- 20_- 150- om 100' X 100 200 400 600 800 K.0 rate (ppm) Figure 2. Intensity of K migration along soil profile as related to the rate applied. Considerable changes in the potassium content of a soil profile can be re- vealed in long-term experiments. At present, at least 30 experiments of this kind are under way in the USSR to investigate the potassium dynamics over the soil profile as a result of fertilization. Data of a study with three sod-podzolic soils of the Central region of the USSR are shown in Table 8. It can be seen by horizontal comparison that potassium from fertilizer has migrated down to 100 cm only on light soils. III Table 8. Migration of exchangeable potassium in the soil profile as found in long-term experiments on soils of different texture (mg K20/100 g) Heavy loamy soils Medium loamy soils Sandy loam soils (20*) (20) (15) Depth (cm) NP + Km,* NP +Ko +1K240 NP + Ko +Klso 0- 20 5.6 7.8 6.5 9.4 13.3 7.5 10.0 20.5 20- 40 4.5 5.7 8.7 10.7 11.8 5.3 10.1 21.8 40- 60 9.6 9.9 12.3 13.3 14.2 2.9 5.6 12.1 60- 80 13.6 13.3 11.8 11.6 13.1 2.4 3.7 6.5 80-100 15.3 15.5 11.2 11.5 12.3 3.2 3.5 4.2 =years of experiment **=average rates of K20 kg/ha In addition to physical properties of soil, which most seriously affect potas- sium migration, the requirements of agricultural crops for the particular ele- ments, both in terms of their rates and inter-nutrient balance are important. Defects in culture, in crop selection, the unjustified application of high rates of fertilizers - all lead to low crop yields and to irreversible losses of potassium, especially on well-drained soils. Soils with higher absorbing capacity are found to sharply increase their potassium content. We analyzed samples of sandy loam soil from the Lyubertsy experimental field of the Institute of Fertilizers. After 15 years the migration of potassium was found to be affected by the rates of fertilizers applied and the balance between the nutrients (Table 9). With deficit in potassium, its content is ob- served to be decreasing throughout the sandy loam soil profile. On the con- trary, when the balance of potassium is positive, it can be accumulated even at depth of 150 cm, as the data for the medium loamy soil shows. When the potassium balance is slightly negative due to K removal by plants, the content of potassium throughout the soil profile is increasing probably due to mobili- sation of less easily available forms. Table 9. Exchangeable K20 content in a sandy soil profile as affected by long-term appli- cation of fertilizers (mg K20/kg) Depth (cm) 0 1-1-0- 1-1-1 1-1-4 1-1-4-* 4-1-0 4-1-1 4-1-4 0-1-1 0-1-4 0- 20 50 34 62 248 174 29 35 138 89 223 20- 40 38 32 51 186 221 25 44 138 76 208 40- 60 30 27 25 103 96 20 28 70 30 94 60- 90 23 24 32 51 32 24 31 58 33 59 90-120 22 19 38 38 32 is 24 47 32 50 120-150 17 17 27 28 23 17 17 32 20 34 Approximate balance of K20 (kg/ha) for 15 years: -938 -1375 -819 +1453 +1466 -1573 -919 +1198 -352 +2005 • Rates, kg/ha N-P 20 5 -K20: 1-1-1: (60-120-60); 4-1-4: (240-120-240) • Potassium chloride substituted by potassium sulfate 112 In every actual case, one will find a particular potassium regime in soil, depending upon the peculiarities of this soil, of the climate, the agricultural crops and the accompanying fertilizers. Potash fertilization can be rational only when all the above factors are thoroughly considered. Similar conclu- sions, but with regard to regional features, were made by numerous authors in the USSR and abroad (Belajev [19711; Trifonenkova et at [1977]; Gamaley [1980]; Hapkina et at [1974]; Schlichting et at [1984]). Practically no losses of potassium, even in long-term experiments, were found on heavy loam soils and on chernozems, though after some time, potas- sium leaching was observed to be on the increase (Chlystovsky et at [19831; Markowsky et at [1973/). The latter may occur when the rate of potassium exceeds the potential fixing capacity of soil (Fotymna et at [1977]). On light sandy soils in zones with temperate precipitation, potassium losses were about 21 kg/ha annually (VOmel [1974]). In practice it is important to remember that potassium migration is much greater in plant-free soils (Baler et at [1971]). In most experiments potassium migration was found to be retarded by liming, especially in the 0-50 cm layer (Rubanov et at [1967/). According to special investigations, downward migration of cations, and of potassium in particular, is increased by soil acidification, sometimes by 3.5 times (Makarov [1987]). When one intensifies fertilization with potash, one should also take measures to minimize potassium losses induced by sur- face run-off due to erosion. A special investigation has shown that loss of potassium in surface run-off is three times greater than by leaching (Petrova [1979]). The composition of surface run-off was examined over 4 years in 4 rivers in central European USSR. In the basin of the Oka river the annual irreversible removal and loss of potassium varied from 80 to 250 kg/ha (Schewtchenko [19841). Taking into account the results of the many years of investigation of potas- sium migration, it would be necessary to review the whole system of fertiliza- tion: rates and forms to be used, timing of application, methods and machines for rational application of fertilizers. 4.3 Problems of chloride-ion migration From an agroecological position, it is very important to consider the dynamics of chloride in soil, this being the basic anion accompanying potassium (Mineev [1984]). Depending on the timing of fertilization and the level of precipitation, chloride in fertilizers may either almost completely enter the plants or practically completely be leached beyond the I m layer of soil. Dur- ing drought chloride may ascend. In our above experiment we measured soil chloride down to 3 m deep; the measurements were made in spring, before fertilizer was applied, and after 12 years of examination. The dose of potassium chloride was varied from 0, 60, 120, 240 to 480 kg K20/ha. Mean chloride contents of 6 samples taken 113 every 50cm were 22±1, 23±2, 28±3, 29±1, 32±2 and 37±1 ppm respectively with insignificant differences between different depths. Thus we can say that long and systematic application of potassium chloride may lead to a certain accumulation of chloride even in sandy loam soil. Many years of investigations on heavy loam soils demonstrated that chloride distri- bution there was practically uniform to a depth of 1 m. The divergence of chloride content from the control treatment was about 20%. Application of kainite and sylvinite augmented the chloride content in soil by 40 to 80% (Tahin et al. [1968/). Today the role of chloride is connected with that of nitrates. Chloride in- hibits nitrification and is antagonistic to nitrate. In our experiments chloride slightly decreased uptake of nitrate. However, we observed this phenomenon only when the level of nitrate was low. Within recent years the chloride content in precipitation, as observed in the industrial regions of the USSR, averaged 11.1 mg/l, which is almost ten times greater than that in purely agricultural areas (1.5 mg/). In view of this, it may be feasible to reconsider the use of various forms of potash fertilizers. 5. References 1. Addiscott, T.M., Rose. D.A. and Bolton, J.: J. Soil Sci., 29, 3, 305-314 (1978) 2. Amberger, A. andSchweiger, P.: Z. Acker und Pflanzenzucht, 148, 5,393-402(1979) 3. Balev, P.M. and Kudeija, P.G.: Publ. Timirjazevskaja Academy of Agricultural Sciences, 3, 101-108 (1971) 4. Bartlett, R. J. and McIntosh, J. L.: Soil Sci. Amer. Proc., 33, 4, 535-543 (1969) 5. Beckett, P. H. T. and Nafady, M. H.M.: J. Soil Sci., 18, 2, 263-281 (1967) 6. Belajev, G.N.: Bodenkunde, 1, 58-69 (1971) 7. Chlystovsky, A.D.: NIUIFs Werke, 242, 17-33 (1983) 8. Cooke, G. W.: The control of soil fertility 520 pp. (1967) 9. Donatora, H.: Sb. VSZ Prase, A, 42, 151-160 (1985) 10. Fotyma, M. and losek, S.: Nowe zol., 26, 7, 21-25 (1977) 11. Gamaley, W.I.: Agricultural Chemistry, 5, 30-38 (1980) 12. Goulding, K. W. T and Talibudeen, 0.: Journal of Soil Science, 35, 3, 397-408 (1984) 13. Hapkina, Z. A., Mejerovsky, A. S. and Silitch, A. N.: Bodenkunde, 9, 79-87 (1974) 14. International fertilizer correspondent: Vol. XXX, No. 4 (1988) 15. Jargens-Gschwind, S. and Jung, 1.: Soil Sci. 127, 3, 146-160 (1979) 16. Kornbloum, E.A. et al.: Soil Science, 1, 107-114 (1972) 17. Kuhlmann, H.: Transactions 13th Congress of the International Society of Soil Science VIII. Hamburg, 828 (1986) 18. Landwirtschaft der UdSSR. Statistischer Sammelband. M. 1988, 528 (1988) 19. Low, A.J. and Armitage, E.R.: Plant and Soil, 33, 2, 393-411 (1970) 20. Makarov, M. L: Experiment. und mathematische Modellierung beim Studium von Bio- geozenosen von Waldern und Mooren. Unionssitzung, Westdwina, 4-6 August, 1987, Leitsttze zum Bericht, 36-39 (1987) 21. Marko%sky A.G. el al.: Sammelband Bodenbearbeitung und Diingungssystem in Fruchtfolgen an der Mittelwolga, Kuybyschev, 95-104 (1973) 22. Mengel, K. and Steffens, D.: Bio. Fert. Soils, 53-58 (1985) 23. Mineev, W. G.: Agrochemie und Biosphaire, 246 pp. (1984) 114 24. Moberg, L B and Nielsen, .D.: Acta Agricultura Scandinavica, 33, 1, 21-27 (1983) 25. Nemeth, K.: Landwirt. Forsch., 25, Sondern. 27/2, 184-196 (1972) 26. Niederbudde, E.A.: Z. Pflanzenern. Bodenk., Heft 2, 217-234 (1975) 27. Niederbudde, E.A. and Rlhlicke, G.: Z. Pflanzenern. Bodenkunde, 144, 2, 127-135 (1981) 28. Noordwijk, M. van and Willigen, P. de: Neth. J. Agr. Sci., 34, 3, 273-281 (1986) 29. Ohlendorf W. and Vomel, A.: Forsch. und Ber., 30, 29-37 (1976) 30. Orlova, N. Yu., Prokoshev V V and Sokolova T A.: Einfluss von Kalidongemittein auf einige Werte von dem Kaligehalt im Boden (in press) 31. Petrova, L: Bodenkenntnis und Agrochemie, 14, 5, 45-49 (1979) 32. Poonia, S.R., Mehra, S.C and Pal, R.: Soil Science, 141, 1, 74-83 (1986) 33. Rajkovie, Zivan: Radov-scientific yearbook, 11, 49-71 (1967) 34. Rhoads, EM.: Proc. Soil and Crop Science Soc. of Florida, 30, 298-304 (1971) 35. Riabzev, RM.: Sarumelband von wissenschaftlichen Werken, Weissrussland - land- wirtschaftliche Akademie, 19, 11-20 (1976) 36. Rubanov, W. S. and Markewitch, I. 7 Sammelband von wissenschaftlichen Werken, Weissrussland Nil-Bodenkunde, 101-110 (1967) 37. Russel, E.J.: Soil conditions and plant growth, 623 pp. (1950) 38. Schaimuchametov, M.Sch. and Knjaseva, N. W.: Bodenkunde, 12, 134-139 (1987) 39. Schewtchenko, L.N.: Ekonomische Probleme der Verwendung von Diungemitteln. Wissenschaft, 210 (1984) 40. Schlichting, E. and Clemens, J.: Z. Pflanzenernhr. und Bodenk., 147, 3, 361-370 (1984) 41. Schweiger, P and Amberger, A.: Landwirt. Forsch., 31, 4, 327-335 (1978) 42. Skarda, M.: Rost]. vyroba, 13, 3, 271-284 (1967) 43. Sokolov, A. W.: Verteilung von Nhrstoffen im Boden und Pflanzenernte, 328 pp. (1947) 44. Sokolova, TA.: Tonmineralien in Humusboden der UdSSR, Verlag 115 Soil Moisture and K Mobility H. Grimme* Summary The soil-plant system is a dynamic system in which movement of nutrient and water is an important component of its functioning. The degree of mobility determines nutrient availability, because plant roots get into direct contact only with a small proportion of the total quantity required for good growth and economic yield. The larger part of the nutrients taken up by plants has to move to the plant roots. This movement or transport takes place in (diffusion) or with (massflow) in the soil water (the soil solution). The extent of transport by diffusion and massflow to the plant roots depends mainly on plant demand (withdrawal of water and nutrients by the plant), nutrient concentration in the soil solution and soil water content. This applies especially to nutrients like K the larger part of which is transported by diffusion. The soil water content determines the cross sectional area available for diffusion, thereby affecting the magnitude of diffusive flux. In dry years there is a greater response to K dressing and the critical K level in the soil is higher under dry conditions. It could be established that there is a close correlation be- tween soil water content and K diffusive flux in a soil and K-uptake. Drying out of the arable layer reduces K availability while water supply from the subsoil may still be sufficient to meet plant demand. Laboratory and field methods have been developed to study the soil water-nutrient-plant response relationship. Among these are split-root techniques, controlled differential water supply to topsoil and subsoil and the use of soil solution probes and tensiometers. Leaching is a special case of massflow in downward direction out of the rooting zone caused by a downward hydraulic potential. Its magnitude depends on solute concentration and groundwater recharge. 1. Introduction Within the last three decades, knowledge of the factors of nutrient avail- ability has made significant progress. It was Bray's[1954] paper on a «nutrient mobility>> concept of soil plant relationship that stimulated a number of soil scientists to direct their research efforts to the problems of nutrient availabil- ity, to investigate which factors actually affect nutrient availability and to study the quantitative effects of the individual factors. Barber [1962] and Barber et al. [1963], identified diffusion and massflow (convection) as the mechanisms governing nutrient supply to the roots and estimated the contri- bution of each mechanism. Nye and coworkers elaborated this concept further and gave a consistent quantitative treatment (Nye and Tinker [1977]). * Dr. H.Grimme, Buntehof Agricultural Research Station, Bunteweg 8,3000 Hannover 71, Federal Republic of Germany 117 This newly developed theoretical concept has proved to be a powerful tool in soil fertility work. It is especially helpful, when it comes to the interpreta- tion of results of field and pot experiments, which apparently are incompat- ible with the soil nutrient status, as determined by established conventional soil testing methods. The important point arising from the work of the above-mentioned workers is the fact that the soil-plant system is a dynamic one and that a number of processes is involved in the nutrient supply to plants. This conclusion implies that soil test results give a rather poor correlation with plant response if a range of soils and climatic conditions is included, because only static parameters are measured, although nutrient mobility is of overriding import- ance with respect to nutrient availability. Let us assume that a root grows into a soil compartment in which the in- dividual components and phases are in equilibrium. The moment when the root begins to take up water and nutrients, this equilibrium will be disturbed and a number of interdependant processes will be initiated. The root absorbs potassium from the soil solution which is in direct contact with the root with the uptake rate being proportional to the initial K concentration in the soil solution. This will reduce the K concentration near the root and a concentra- tion gradient will be generated, resulting in a net diffusive flux towards the roots. This diffusive flux is kept going as long as the root is absorbing K. Besides nutrients the roots also absorb water of which only a small propor- tion of total plant demand is in proximity of the root so that most of the water taken up by a plant has to flow to the roots along a water potential gradient. The water then takes along the solutes dissolved in it. This type of transport is called mass flow. The solutes will be taken up by the roots if there is a demand or else they will accumulate around the root with a concomitant effect on exchange and solubility equilibria and on release processes. Soil water is an important component in the processes involved in affording the plants an adequate nutrient supply. Soil water content has two main effects on plant growth which operate independently. It is in the first place an import- ant constituent of the plant and as such is essential for plant life. On the other hand, it is the medium from which the plant takes up the mineral nutrients, and it also serves as the transport medium for nutrients to the roots. One can, therefore, imagine a situation where the water content in the soil is below field capacity but the water supply is still adequate for plant growth, and yet nutrient transport to the roots is already restricted. Leaching is another aspect of nutrient mobility related to nutrient availability, because nutrients leached out of the rooting zone are lost to the plants. The extent of leaching losses also depends on nutrient mobility and thereby on the soil water regime. There are observations from field and pot experiments demonstrating the interaction between soil water content and nutrient availability and confirm- ing thus the theoretical argument. During II years of field experiments Bruns [1935] found yield differences between control plots and K fertilized plots to be greatest in dry years. Barber [1959/ recorded a varying effectiveness of P and K fertilization from year to year, although P and K contents in the 118 soil of the individual treatment did not vary. And he found that the reduction of yield which was obtained in dry years was least on those plots with the highest nutrient contents in the soil. Van der Paauw [1958] published the results of 12 years of field experimentation, which showed that the yield response of potatoes to 400 kg K20/ha increased with the number of rainless days during the growing season (Figure 1). Similar results were obtained with sugar beet (Table 1). Comparing the K response in a wet (1986) and a dry year, in both years nearly the same sugar yield was attained with the highest K dressing, whereas in the control treatment sugar yield dropped off by nearly 50% in the dry year (1989). 20- 0) o C 0 0a10 1 ~0 40 50 60 70 Number of days without rain 1. Figure Relationship between the number of rainless days during the growing period (May through July) and yield response of potatoes to K fertilization (van der Pacuw [1958]). Table 1. Response of sugar beet to K fertilization in a longterm experiment (Niestetal) in a wet (May through Sept.: 221 mm) and dry (May through Sept.: 178 mm) season (Orlovius, unpublished) K dressings (kg/ha) Extractable sugar yield (t/ha) 1986 1989 ...... 0 5.9 (1000%0) 3.3 (100% 100 ...... ) 6.4 (10807o) 5.4 (164% ) 200 ...... 7.3 (124% ) 6.7 (203% ) 300 ...... 7.9 (134 o) 7.5 (227% ) Results to the same effect were also published by other authors (Russeland Watson [1938]; Ferrari[1952j; Richards and Wadleigh [1952; Danielson and Russet [19571; Brown et al. [1960]; Peters and Russel 11960]). All the authors assumed that reduced soil water content affected in some way nutrient avail- ability, whereas the plants' water requirement was largely met, especially in Van der Paauw's experiments, otherwise yield should have been reduced in 119 spite of K fertilization. But it was not before nutrient mobility and transport in the soil were recognized as important factors of nutrient availability (Barber [1962]) that this interpretation was substantiated, e.g. by split-root-experi- ments (von Braunschweig and Grimme [1973]). 2. Mechanisms of K movement in the soil K can move in the water (massflow or convection) or with the water (diffu- sion). a) Massflow When the plant absorbs water it generates a flow of water to the roots carry- ing along the solutes (e.g. K). This movement of nutrients through soil pores is the convective flow of water or massflow to plant roots. Its magnitude de- pends on the consumption of water by the plant and the nutrient concentra- tion in the soil solution. The effective contribution of massflow in transport- ing potassium to the plant is rather low. Till recently, this statement was based on rough approximations, taking into account total water used by a crop and mean K concentration (Barber et al. [19631; Mengel et al. [1969; Barber [19841). Field data obtained with several crops support the above assumptions (Strebel et at. [1980]; Renger et al [19811). The observations made with spring wheat showed that at anthesis, total K uptake had been 205 kg/ha out of which only 4 kg K/ha were provided by massflow (Figure 2). As after anthesis no further K uptake occurred, supply through massflow became negligible. Similar results were recorded for sugar beet and spring barley. S0-110 cm 0 0 30 20 30o 00 4 as w-seoi'"oh 30-60ca cm 0 lii i 1111i i1 10 20 30 10 20 30 0203 1020 30 1 May June July Aug. Sept. Date (day and month) 22 Figure 2. Cumulative K supply through massflow to the roots of spring wheat. Total massflow as well as the contribution from the 0-30 cm and 30-60 cm soil layers are given. Although t he roots grew down to 110 cm, K supply by massflow from chat layer was negligible (Renger et al. /1981]). 120 Leaching losses also occur through massflow along a vertical water poten- tial gradient caused by gravity. In the temperate zone such a situation prevails usually only in winter. b) Diffusion Movement by diffusion is the major supply mechanism for potassium, the driving force for K to diffuse towards plant root resulting from the depletion caused around an absorbing one. This depletion zone was first demonstrated by Farr et at [1969/ and subsequently in greater detail by Claassen and Jungk [1982]. Plants can, indeed, reduce the K concentration around their roots to very low levels generating a steep concentration gradient causing a net dif- fusive flux towards the roots. K diffusion takes place in the soil solution (Vaidyanathan et at [1968]). Therefore water content has an important influence on diffusive flux in the soil, which is the lower the lower the water content of the soil (Grimme et at 11971/). This difference in diffusive flux at different water contents is also reflected by the change of exchangeable K content in the soil with distance from the absorbing surface. Figure 3 shows that with low pF values (high water content) there is a larger decrease in K content near the soil root mimicked by an ion exchanger interface and that the distance to which K depletion occurs by diffusion is larger than at high pF values. This means that in a soil with high water content a larger soil volume can be exploited by the root within a given time than at low moisture supply, and consequently a larger amount of nutrients is available to the plant within a given time. Thus soil water content affects nutrient availability (Figure 1, Table 1). 0_°i20" 0 E A / *-. pF 3.2 (21% H,0) S10 --- pF 1.8 34% H20) Lu I t II 10 20 30 Diffusion path (mm) Figure 3. Depletion of exchangeable soil K through diffusion into a synthetic ion ex- changer serving as sink at two soil water contents. Depletion was greater at the higher soil water content. (Soil: loess, 22% clay) (Grimme, 119711). 121 The knowledge of the contribution of massflow to K supply to the plants from individual soil layers (Figure 2) allows us to make an estimate of the K supply through diffusion from various depths of the profile. If rooting den- sity and soil water content are known as a function of depth and time, one can also use diffusion measurements done in the laboratory on samples taken in the field from the individual soil layers, in order to calculate K diffusion to the plant roots for specified time and depth intervals (Grimme et aL [1981]; Fleige et a [1983]). Figure 4 compares the calculated total K supply to a crop of spring wheat through diffusion with the actual K uptake to the stage of maximum uptake, which was at anthesis. The agreement is quite good con- sidering the uncertainties involved in such a calculation. The K supply from each horizon changed during the course of the growing season (Figure 5). This change reflected mainly the effects of root growth and soil water content, which changed during the growing period. Rooting density and rooting depth increased whereas soil water content fluctuated. 300-• K uptake - K supplied 2 by diffusion 100 O, 10.520.531.510.620.630.6 Date Figure 4. K uptake of spring wheat and K supply through diffusion. Same experiment as in Figure 2 (Fleige et aL. [1983]). .-2_: 100" -- -- o 80-- 1 0 N 0 .> =.B -- o 20-.. . I...... S t- ...... -o,...... ,...... BvB 10.5 20.5 31.5 10.6 20.6 30.6 Date Figure 5. Relative contribution of individual soil horizons of a grey brown podzolic soil (Udalf) to K supply through diffusion to spring wheat (same experiment as in Figure 2) (Fleige el a. [19831). 122 3. Experimental approach There is a number of techniques available to study nutrient mobility in soils and the effect of soil water content on nutrient mobility. The methodology is well developed. A general outline will be given in the following chapter (see also: Barber et at [19631; Barber [1984]; Nye and Tinker [1977]). 3.1 Laboratory methods (Diffusion) There are extensive theoretical treatments of solute movement by diffusion in the soil (Nye and Tinker [1977]; Barber [1984/) and the interested reader is referred to these monographs for details. A thorough understanding of the theoretical background is necessary for a proper interpretation of experimen- tal results. As explained in the previous chapter, transport by diffusion is the major K supply mechanism to plant roots. It results from a disturbance of equilibrium by K uptake through the roots. A mathematical expression for the simplest case of one-dimensional flow and stationary conditions is given by Fick's First Law: de Diffusive flux=-D d. (1) where c is the concentration of the mobile species and x is the distance over which net diffusion occurs. D is the diffusion coefficient. For soil system equation (1) has to be modified, since the cross sectional area available for diffusion is restricted to the water filled pore space and a term for the tortuosity of the diffusion path has to be included (Nye [1972]). dc Diffusive flux=-D, vf- +F. (2) D, is the diffusion coefficient in the soil solution, v is the fractional volumetric soil water content, f is an impedance factor taking account of the tortuosity and discontinuity of the soil pore system. F, represents the contribution of the solid component which is negligibly small if at all existing. Since the cross sectional area and the tortuosity factor are functions of soil water content, ion concentration in the soil solution and soil water content emerge as the important factors governing K diffusion in soils. A further elaboration of the equation would also include the buffer power of the soil for K in solution dc b= T (3) The driving force for K to diffuse towards the roots results from the K deple- tion around an absorbing root. This depletion zone was demonstrated by Farr et at [1969] and more recently and in more detail by Claassen and Jungk 119821. 123 The measurement of diffusive flux does not require sophisticated equip- ment. The principle consists in placing a block of soil in intimate contact with a sink; either plant roots (Farre at [19691; Claassen and Jungk [1982J) or a synthetic ion exchanger (Grimme et at [1971]) or exchange resin paper (Vaidyanathan and Nye [19661). Soil water potential is adjusted to the same level on ceramic plates in a pressure pot in order to prevent any water flux. Water flux cannot completely be prevented when working with plants. An example of the effect of soil water content is presented in Figure 6, which shows that soil water content has a significant influence on K diffusive flux. Fractional volumetric water content - 2.0- (0,51) E (0.39) (0.27) 1.0- (0.20) 0<20 0 if I I I I 1 I 17i- I= 0 0.5 1.0 K concentration (me/I) Figure 6. Correlation between K concentration in the soil solution and K diffusive flux at 4 soil water contents (Chernozem) (Grimme and von Braunschweig (1974]). 3.2 Field methods Only one method, which, however, affords a very comprehensive study of nutrient-water interactions in the field will be described. This method employs ceramic soil solution probes (suction cups) (Renger and Strebel [1979]). Using soil solution probes and tensiometers is an alternative method allow- ing quasi-continuous monitoring of nutrient concentrations and of the soil water regime with a minimum disturbance of the soil. This method also allows gathering of data over unlimited time from the same field. This method relies on separate determination of nutrient concentration in the soil solution and water flow. Suction probes have been in use for quite some time (Strebel et at [1973]), but it is the simultaneous combination of soil solution sampling and determi- nation of water movement that makes the method attractive (Strebel et at. [1973]; Renger et at [1975]; Strebel et at [1980/). 124 Suction probes and tensiometers are installed at various depth according to the design of the experiment. Soil solution can be sampled at any time the soil contains sufficient water at a matric potential >-7x 104 Pa. The maxi- mum suction that can be applied to the probes is usually around -8x I0 Pa and depends largely on the quality of the porous cups. Whether a sufficient amount of water is extracted depends, of course, on the amount of water held by the soil in the working range of the soil solution probes, and on the hydraulic conductivity in this range of matric potentials. It is possible to get a good resolution in time and depth and thus a very detailed picture of what is going on in the soil. The method allows calculation of total water flow (w10 )= flow through roots (Vr)+capillary flow (Vcap) as a function of depth and time: + Wioi=Vcap Vr (1) By differentiation one obtains the water uptake rate of the roots. Written in the form of finite differences, it reads as follows (z=depth): Av r=- z (2) The capillary flow is determined using Darcey's law: Az (3) k = soil water conductivity as a function of 0 = matrix potential z = gravitational potential (or depth resp.) 0 = hydraulic potential (4'+z) The total flow is calculated using the continuity relationship. Z2 AO V0o,2 -v, 1 = AzA (4) Z' (8=water content, t=time) Ion massflow to the roots can then be calculated for each depth interval using the relationship ri..= v, • concio. (5) and the amount of an ion leaching from a given compartment or layer of a soil can be calculated as follows: 125 lion VCap - COnCion In order to be able to calculate vcap and vot, one needs to know volumetric soil water content and the soil water matric potential both as a function of time and depth. In addition for each soil layer the relationship between soil water conductivity and soil water matric potential needs to be known. Water conductivity can be either determined in the laboratory or in the field. A combination of this method with diffusion measurements in the labora- tory allows to obtain the data used in Figures 2, 4, 5. 3.3 Direct measurement of the effect of soil water content on K availability An unequivocal proof and a quantitative estimate of the effect of soil water content on K availability to plants can only be provided with an experimental design that rules out a reduced water supply to the plant but at the same time allows variations of K availability caused by variation in soil water content. This can be achieved by employing a split-root technique (von Braunschweig and Grimme [1973]) where half of the roots grow in sand nutrient solution medium and half of the roots in the soil (Figure 7). The water potential and therewith the water content in the soil is controlled and kept constant by means of ceramic filter candles which are linked to a water reservoir under the re- quired negative pressure. This technique allows the plants to take up nutrients and water liberally from the nutrient solution except the nutrient, the availabil- ity of which is to be investigated. In our case, it was K that had to be taken Plants = T cm Hg Nutrient solution I Pressure gauge to vacuum pump Tensiometer H O nutrient solution Ceramic filter candles Figure 7. Schematic representation of split root technique (von Braunschweig and Grim me /1973]). 126 up from the soil. The uptake rate of K can then be determined in relation to soil water content, whereas water supply to the plants is not growth limiting, because there is an ample supply through the roots growing in the nutrient solution. Growth differences are then only due to differences in K availability (mobility). A typical result of such a study is shown in Figure 8. At each of the four K levels tested, K uptake increased with increasing soil water content. As all growth factors except K availability were kept constant, the differences were caused by the effect of soil water content on K availability. The reduction of K availability caused by low water content could be compensated by raising the K concentration in the soil solution. The yield (not shown) behaved in the same way as uptake demonstrating that the experiment was carried out within a relevant range of K supply. The soil water contents applied were be- tween 20 and 40076 (by volume) corresponding to water tensions of 0.1-0.5 bar, which would be well within the limits of good water supply. But as these results show, significant reductions of yield may occur, by a decrease of soil water content long before the plants suffer from water shortage. These findings al- low an unequivocal interpretation of field trials demonstrating that under dry conditions soils should have a higher K status (Figure 1, Table 1). A split root experiment appears to deal with a rather academic situation. But, in fact, this situation may occur in the field quite frequently, the main difference being the vertical rather than the horizontal arrangement of com- partments (Figure 9). Exch. K K-conc. (mg/100 g) (me/I) 500- 46 1.07 0 o A31 0.47 o300- 17 0.17 0 100 .. c -.. 11 0.08 ID I I 20 30 40 Soil water content (Vol %) Figure 8. The effect of soil water content on K uptake of young maize plants at four soil K levels (Chernozem) (Grimme [1979]). 127 FL-I Vacuum pump Hg-manmeter_ Topsoil ==0- 5 0 0 cm Suboil13 -100 cm H20 Subsoil - - Porous ceramic filter candles Water reservoir (negative pressure) Figure 9. Schematic representation of experimental setup with controlled differential water potential in top- and subsoil. Total length of soil column 80 cm (Grimme, unpublished). Table 2. Effect of soil water content on dry matter yield, mineral composition and nutrient uptake of berseem (Trifolium alexandrinum) (Grimme, unpublished) Topsoil dry (0.7 bar) Topsoil wet (0.1 bar) Subsoil wet (0.1 bar) Subsoil dry (0.7 bar) Dry matter yield (relative) 100 130 Mineral composition (%) Relative difference (W7) K ...... 2.10 2.80 33 M g ...... 0.25 0.30 20 C a ...... 1.90 2.30 21 P ...... 0.26 0.47 67 N ...... 3.60 4.00 11 Nutrient uptake (g/soil column) K ...... 2.26 3.58 58 M g ...... 0.27 0.38 41 C a ...... 1.97 2.86 45 P ...... 0.26 0.61 135 N ...... 3.57 5.19 45 Plants were grown in 80 cm long and 25 cm wide soil colums consisting of a reconstituted chernozem profile. Water was supplied through porous cups (30cm long, 3 cm wide), three each in the top soil (0-40 cm) and the subsoil (40-80 cm), connected to water reservoirs maintained under a specified negative pressure. Total water supply was equal in all the treatments. 128 It often happens that the top soil dries out while the lower layer is still quite moist. The plant can get enough water from the subsoil but its nutrient supply is likely to be impaired because of the normally lower nutrient content in the subsoil than in the surface. In an experiment using soil columns in which only the top soil was fertilized and either the top layer or the subsoil was kept drier than the other, this effect could be demonstrated. Although total water supply was the same in both cases, yield and nutrient uptake were significantly reduced when the topsoil was kept dry (Table 2). The difference in yield was attributed to reduced nutrient availability. This finding underlines the conclu- sion from the split root experiment that under dry conditions which especially affects the top soil nutrient status should be higher and that attention should be paid to the nutrient status of the subsoil. 4. Leaching It was already mentioned that leaching is a special case of massflow. It is only of importance, when there is a substantial downward flow of water out of the root zone. This is usually the case in the winter season of the temperate humid regions and in the rainy season of the humid and subhumid tropics and even in large areas of the semiarid tropics (because of high rainfall inten- sity within the short rainy reason). Broadly speaking, three approaches are possible for studying leaching losses under field conditions (1)The auger method, (2) use of lysimeters, and (3) use of porous cups as soil solution probes in combination with tensio- meters and/or equipment for determining soil moisture at frequent intervals such as neutron probes (see chapter 3.2). Method (1) requires only minimal instrumentation and can still yield meaningful results in system with low buffer power. Lysimeters (2) have certain shortcomings if not properly designed and maintained. Method (3) (chapter 3.2) is a more elaborate approach and yields the most meaningful results be- cause of high spatial and temporal resolution. An estimate of the extent of leaching losses is obtained when K concentra- tions at the lower boundary of the rooting zone and when the groundwater recharge are known. Since both may vary widely K-leaching may also vary widely, in most cases, however, at rather low absolute levels, because the K concentration in the subsoil is rather low. The range of K-leaching encoun- tered in central Europe is given in Table 3. 129 Table 3. The range of annual leaching losses (kg K/ha) encountered under temperate cli- matic conditions at typical K concentrations in the soil solution and the range of groundwater recharge pertaining to that region Groundwater recharge (I/m2 ) K conc. in soil sol. (me/I) 50 100 150 200 300 (Clay) 0.01 ...... 0.2 0.4 0.6 0.8 1.2 0.05 ...... 1 2 3 4 6 (Loam) 0.10 ...... 2 4 6 8 12 (Sand) 0.50 ...... 10 20 30 40 60 5. Conclusions As soil water content decreases, K availability to plants is impeded because of restricted mobility. K availability is reduced before water availability becomes a limiting factor. This applies to all nutrients the major partion of which travels to the roots by diffusion, e.g. K and P. In dry years only soils with a high K status will attain maximum yield. 6. References Barber, S.A.: Relation of fertilizer placement to nutrient uptake and crop yield. 11.Effects of row potassium, potassium soil-level and precipitation. Agron. J. 51, 97-99 (1959) Barber, S. A.: A diffusion and massflow concept of nutrient availability. Soil Sci. 93, 39-49 (1962) Barber, S.A., Walker, J.M. and Vasey, E.M.: Mechanisms for the movement of plant nutrients from the soil and fertilizer to the plant root. J. Agric. Food Chem. IL 217-219 (1963) Barber, S.A.: Soil nutrient bioavailability. A mechanistic approach. Wiley and Sons, New York (1984) Braunschweig, L. C. von and Grimme, H.: Eine Split-root Technik zur Untersuchung der Nihrstoffverfflgbarkeit in Abhgtngigkeit vom Wassergehalt im Boden. Z. Pflanzener- nahr. Bodenk. 134, 246-256 (1973) Bray, R. H.: A nutrient mobility concept of soil plant relationship. Soil Sci. 78, 9-22 (1954) Brown, D.A., Place, G.A. and Petter, J V.: The effect of soil moisture upon cation ex- change in soils and nutrient uptake by plants. 7th Int. Congr. Soil Sci. Transact. 111, 443-449 (1960) Bruns, W.: Untersuchungen Uiber Nahrstoffaufnahme und Wasserhaushalt der Ackerbohne. Journ. f. Landwirtsch. 83, 285-325 (1935) Claassen, N. and Jungk, A.: Kaliumdynamik im wurzelnahen Boden in Beziehung zur Kaliumaufnahme von Maispflanzen. Z. Pflanzenernahr. Bodenk. 145, 513-525 (1982) Danielson, R. E. and Russel, M. B.: Ion absorption by corn roots as influenced by moisture and aeration. Soil Sci. Soc. Amer. Proc. 21, 3-6 (1957) Drew, M. C, Vaidyanathan, L. V. and Nye, PH.: Can soil diffusion limit the uptake of potassium by plants? Proc. Int. Soc. Soil Sci. Joint Meeting of Comm. I + IV, Aberdeen, 335-343 (1966) 130 Farr, E., Vaidyanathan, L. V and Nye, PH.: Measurement of concentration gradients in soil near roots. Soil Sci. 107, 385-391 (1969) Ferrari, Th.: An agronomic research with potatoes on the river ridge soils of the Bommeler- ward. Vers]. Landbouwk, Onderzoek 58.1, 132 pp. (1952) Fleige, H., Grimme, H.. Renger, M. and Strebel, 0: Zur Erfassung der Nahrstoffanliefe- rung dutch Diffusion im effektiven Wurzelraum. Mittelgn. Dtsch. Bodenkdl. Gesellsch. 30, 381-386 (1983) Grimme, H., Nemeth, K. and Braunschweig H. C von: Some factors controlling potassium availability in soils. Proc. Int. Symp. on Soil Fertility Eval., New Delhi Vol. 1, 33-43 (1971) Grimme, H. and Braunschweig, H. C. von: The interaction of K concentration in the soil solution and water content on K diffusion. Z. Pfianzenernghrung Bodenk. 137, 147-158 (1974) Grimme, H., Strebel, 0., Renger, M. and Feige. H.: Die potentielle K-Anlieferung an die Pflanzenwurzel dutch Diffusion. Mitt. Dtsch. Bodenk. Gesellsch. 32, 367-374 (1981) Mengel, K., Grimme, H. and Nemeth, K.: Potentielle und effektive Verfflgbarkeit von Pflanzenniihrstoffen im Boden. Landw. Forschung So. H. 23/1, 79-91 (1969) Nye, PH.: Localized movement of potassium ions in soils. Proc. 9th Coll. Intern. Potash Inst., 147-155 (1972) Nye, PH. and Tinker, PRB.: Solute movement in the soil root system. Blackwell Scientic Publications, Oxford (1977) Paauw, E v. d.: Relationship between the potash requirements of crops and meteorological conditions. Plant and Soil 9, 254-268 (1958) Peters, D.B. and Russel, M. B.: Ion uptake by corn seedlings as affected by temperature, ion concentration, moisture tension and moisture content. Transact. 7th Intern. Congr. Soil Sci. I1, 457-466 (1969) Renger, M., Giesel, W., Lorch, S and Strebel, 0.: Comparison of laboratory and field meas- urements of water conductivity in unsaturated soil. Transact. 10th Intern. Congr. Soil Sci. 1, 309-318 (1974) Renger, M. and Strebel, 0.: Water and nutrient transport to plant roots as a function of depth and time under field conditions. Proc. 14th Coll. Int. Potash Inst., Sevilla/Spain, 79-91 (1979) Renger, M., Strebel, 0, Grimme, H. and Fleige, H.: Nfihrstoffanlieferung an die Pflanzen- wurzel durch Massenfluss. Mitt. Dtsch. Bodenk. Gesellsch. 30, 63-70 (1981) Russel, E. J. and Watson, D. J.: The Rothamsted field experiments on barley 1852-1937. Part. II. Empire J. Exptl. Agr. 6, 293-310 (1938) Strebel, 0., Renger, M. and Giesel, W.: Bestimmung des vertikalen Transports von loslichen Stoffen im wassergeskttigten Boden. Wasser u. Boden 25, 252-253 (1973) Strebel, 0, Grimme, H., Renger, M. and Fleige, H.: A field study with Nitrogen-15 of soil and fertilizer nitrate uptake and of water withdrawal by spring wheat. Soil Sci. 130, 205-210 (1980) Vaidyanathan, L. V., Drew, M. C. and Nye, PH.: The measurement and mechanism of ion diffusion in soils. IV. The concentration dependence of diffusion coefficients of potassium ions in soils at a range of moisture levels and a method for the estimation of the differential diffusion coefficient at any concentration. J. Soil Sci. 19, 94-107 (1968) Vaidyanathan. L. V. and Nye, PH.: The measurement and mechanism of ion diffusion in soils. I1. An exchange resin paper method for measurement of the diffusive flux and diffusion coefficient of nutrient ions in soils. J. Soil Sci. 17, 175-183 (1966) 131 K-Uptake from Subsoil M. De Nobili, L. Vittori A ntisari and P Sequi Summary Little attention has been paid until now to the role of subsurface layers of soil in agricul- tural production and farming. The chemical as well as physical properties of subsoil, however, may be concerned in supplying K nutrition to plant roots. Particle-size distribution is important for the evaluation and utilization of subsoil K reserve to crops. K-uptake from deeper soil layers may be limited by unfavourable soil characteristics, e.g. acidity arising from leaching or presence of hard plough-sole layers. K-uptake from subsoil varies according to the growth stage of crops, and depends on root length density distribution, which is also influenced by soil moisture deficit. Deeply rooting species such as alfalfa (Medicago sativa L.), soybean (Glycine max. L. Merr) and cotton (Gossypium irsutum L.) possess a higher potential for subsoil K exploitation, but under certain conditions even cereals can absorb subsoil K up to about 50% of total K uptake. 1. Introduction In most soils the levels of exchangeable potassium in subsoil layers are very low and most plant available K accumulates within few centimeters from the soil surface. As a consequence, K fertilizer recommendations are commonly based on soil tests carried out on surface soil samples and the contribution of subsoil K to plant nutrition is believed to be small (Hannay and Johnson [19851). Very few studies can be found in literature on K-uptake from subsoil as compared to the enormous amount of work concerning K nutrition of crops in general. Experimental approaches are mainly based on determination of residual soil K contents, generally exchangeable and HNO3-extractable or diluted-double-acid (DDA)-K, of samples taken at different depths in the soil profile. Leaching of K from upper layers can therefore easily mask the contri- bution of deeper layers to plant nutrition in sandy soils (ParkeretaL [1989b]). In one case only (Haak [1980]) a tracer technique was used. This lack of interest is justified by the fact that subsoil K uptake is indeed likely to be negligible in a wide number of cases; the calculated contribution of subsoil to K-uptake for corn genotypes, for example, was only 10% of the overall observed uptake (Schenk and Barber [1980]). These experiments, however, were conducted on high fertility soils with adequate K-availability. * Dr. Maria De Nobili, Institute of Crop Production, University of Udine, P. le M. Kolbe 4, 1-33100 Udine; Italy and Dr. Livia VittoriAntisari, and Prof. Dr. Paolo Sequi, Institute of Agricultural Chemistry, University of Bologna, Viale C. Berti Pichat 10, 1-40127 Bo- logna, Italy 133 Recent reports (Parkeret al [1989b]; Gulick ef al. [1989]) on two different cases of unresponsiveness to K fertilization, have shown that under particular but not so uncommon circumstances, depending on both soil and crop charac- teristics, the role of subsoil in determining the nutritional levels of crops can- not always be neglected. The actual extent of utilization of subsoil K certainly depends in the first place on soil type. In a summary of results of experimental investigation car- ried out in Sweden from 1971 to 1976 on nutrient uptake from subsoil by cereals, Haak [19801 pointed out the superiority of a clay subsoil as compared to a sandy subsoil. The percentage uptake from subsoil, calculated after con- taminating the plough layer with a tracer, was much higher from the clay sub- soils (Table 1), which contributed from I/5 to 1/2 of total K-uptake during the whole growing period of cereals. Higher straw yields and grain yields were also obtained from cereals growing on the clay subsoils. The same author indicated several other factors which may influence K-uptake from subsoil: early sowing, rainfall pattern and a good P status in the plough layer. Table 1. Effect of plough layer/subsoil combination on the integrated K uptake from sub- soil by cereals. Adapted from Haak [19811 Plough layer Kind of subsoil Straw Grain Lanna silty clay Loamy sand ...... 6 16 C lay ...... 32 33 Ultena clay Loamy sand ...... 0 18 C lay ...... 10 37 Ulleri clay Loamy sand ...... 11 15 C lay ...... 44 43 Kopparsl. loam Loamy sand ...... 25 23 C lay ...... 40 44 All these factors, however, act through their effect on the abundance and distribution of roots, as a higher root development and activity of roots in the subsoil certainly means a higher uptake of nutrients from the deeper layers. We will therefore examine the problem of potassium uptake from subsoil, by considering, first of all, the potentialities of the different soil types to act as subsoil K reservoirs, and of the factors which affect distribution of the different forms of K along the soil profile. Next we will discuss the actual utilization of subsoil K by crops with different rooting patterns as well as factors which influence the distribution of roots between plough layer and subsoil. 2. Potassium reserve in the subsoil It is certainly true that, besides having important effects on moisture regimes and aeration capacities, subsoil layers may also supply certain plant nutrients and therefore affect the growth of crops, grasses and trees. The term «subsoil 134 has been used in a number of different ways: i.e. the 135 [19891). For each group, exchangeable K was found to be related to water 2 soluble K (r'=0.86-0.96) and HNO 3 extractable K (r -0.81-0.83), showing that the capacity to supply K under continuous cropping is greater for smec- titic than for kaolinitic soils. The use of these relationships enabled the author to predict water soluble, K, clay HNO 3 extractable and mineral K in 60 other soils from exchangeable content and clay mineralogy. Indications of the potential subsoil K pool can therefore be simply derived from data given in pedon descriptions by Soil Taxonomy. Examples of information obtained in this way for some typical soils from the clay content of the different horizons are reported in Table 2. Table 2. Some example of distributions of HNO,-extractable and mineral K in three different soil profiles derived according to Sharpley [1989] Depth Clay Exchangeable % of exchangeable Mineral K (cm) % K (me/100 g) K in HNO, (mg kg- ) extractable K Typic A rgiusloll Ap 0- 15 33.2 1.9 13.94 7290.1 7293.0 B211 15- 30 46.5 1.8 16.38 B221 30- 46 42.7 1.7 15.68 7292.2 7290.3 B3ca 46- 76 34.1 1.6 14.11 Cl- 76-115 29.2 1.7 13.21 7289.3 12.97 7289.0 C2ca 115-145 27.9 1.7 1.7 13.03 7289.1 C3ca 145-185 28.2 llAIb 185-220 29.5 1.4 13.26 7289.4 Typic Paleargid At 0- 5 12.6 0.8 13.57 1781.0 1782.9 A12 5- 15 11.7 0.7 15.63 19.71 1785.3 A3 15- 28 12.7 0.9 42.48 1798.5 B,21 28- 53 7.1 2.4 B22, 53- 89 12.7 2.6 36.76 1795.1 B3,,. 89-110 19.1 2.0 23.51 1787.5 C 110-180 32.0 1.8 16.56 1783.4 Typic Haplargid At 0- 3 13.8 2.0 18.13 1784.3 1785.2 A 12 3- 12 16.6 1.4 19.65 B, 12- 23 20.7 1.6 21.88 1786.5 1789.3 B21 23- 46 29.6 1.8 26.87 B22la 46- 66 29.3 1.3 26.54 1789.2 1791.4 B 3tca 66- 84 36.3 1.2 30.35 26.60 1789.2 B24 ,ca 84-110 29.4 0.6 24.75 1788.1 B3 1ica 110-130 26.0 0.3 B32, 130-160 23.1 0.3 23.18 1787.2 23.51 1787.4 B331ca 160-180 23.7 0.2 C 180-230 19.3 0.3 21.11 1786.0 136 3. Root development in acid subsoils A cause of limited root development is subsoil acidity: aluminium causes dis- ruption of mitosis in root tips (Horst et al. [1983]) and seriously damages top roots. An acid subsoil reduces root growth of an Al sensitive, deep rooting species like alfalfa by 69% and that of the less resistent cultivars of Al tolerant sericea lespedeza (Lespedeza cuneata G. Don) by 40% (Joost and Hoveland [19861). Acid subsoils are not uncommon in areas where rainfall exceeds evapotran- spiration and may represent a serious problem for agriculture in the wet tropics. As a result of the intense weathering process, these soils are low in 2:1 clay minerals and organic matter and possess a low cation exchange capac- ity in the upper horizons. In these highly leached soils, application of fertilizer salts can produce relatively high increases in ionic strength and depression of soil solution pH. Manson and Fey [1989] have recently investigated ionic strength effects on the solution composition of an acid B21 horizon of a Plin- + thic Paleudult, and found that K ions can displace quantities of Al3 + similar to thosedisplaced by divalent cations, by lowering pH and enhancing kaolinite solubility. These authors found that root penetration of maize planted in ameliorated topsoil placed above acidic subsoil was completely blocked if sub- soil was pretreated with 10mg/I KCI solution. In highly weathered soils signifi- cant leaching losses of applied K fertilizer are normal: as much as 53% of 108 Kg ha -' K applied to uncropped soil was found below 60 cm depth after 11 months (Vilela and Ritchey [1985). In the presence of an acid subsoil, excess K leached down the soil profile could cause increased Al toxicity damage to the roots of the same or a subsequent crop. Besides possible nutrient unbalance, this could explain the yield decreases sometimes observed in these soils at the highest K rates applied. 4. Effect of root development and growth stage on K-uptake from subsoil In a soil body with layered distribution of available nutrients, maximum nutrient utilization by a crop can only be achieved when root distribution matches nutrient availability. Even if undisturbed root growth results in a gradually decreasing amount of roots with increasing depth; under uniform favourable moisture conditions, root mass concentrates in soil zones where nutrient availability is highest. As a general rule, higher root densities and potassium reserves are both concentrated within the first 20 cm of topsoil. No definitive evidence exists of root growth stimulation by K + ions, as for H 2PO4 - or NH 4 + and NO 3 - ions, but, as nitrogen and phosphorus are also mostly concentrated in the surface layer where organic matter accumulates, potassium uptake is certainly highly enhanced by the higher root density de- veloped in this layer by most crops. The favourable effects of high root density can however be hampered by competition when the half distance between roots becomes smaller than the thickness of K diffusion gradient from the 137 root surface (Schenk and Barber [1980]). Not all crops display sufficient plasticity of the root system to match nutrient distribution in soil: a widely different behaviour was observed by Gulick et al. [1989] for cotton (Gos- sypium irsutum L.) and barley (Hordeum vulgare L.). In their experiment on exploitation of soil potassium in layered profiles, they observed a widely different root distribution in the two crops, with barley roots concentrated in the surface layer (0-18 cm) regardless of topsoil depth or irrigation treat- ment (Figure 1). On the contrary, root development of cotton was nearly uni- form throughout the container volume, expecially with a low frequency irriga- tion regime. This explains why cotton grown on vermiculitic alluvial soils in California displays late season K deficiency symptoms in spite of the fact that these soils are not considered K deficient. In the same soil, dry matter produc- tion of irrigated cotton responded to K fertilization whereas other crops such as barley, wheat and sorghum, which can absorb K more efficiently from the nutrient rich topsoil, did not. Although no other data are available in the literature on subsoil K uptake by cotton, the situation is very likely to be reversed on K deficient topsoils lying on K rich B horizons. In the case of soybean in fact, which under water stress conditions may develop high root densities in the deeper soil layers (Huck et al. [1987], Garay and Wilhelm [1983]), it was found that this crop is capable of exploiting K from argillic horizons deeper than 50 cm (Woodruff and Parks [1980]). In the same experiment multiple linear repression analysis on leaf and Ap and B2 horizons K concentrations showed that on the contrary corn grown in the same soils was unable to absorb K from B2 horizons if deeper than 50 cm. Corn leaf K was however significantly correlated (between 5 and 10% confidence) with the K content of B2 horizons situated within 50 cm of the surface. These results point out the importance of interspecific differences in root morphology in determining potential K uptake from sub- soil. Intraspecific differences can also be important in some cases: two varie- ties of barley were found to have a quite different root distribution, with Arabic Abiad having shorter length in the surface layer and longer length in layers below 30 cm with respect to root distribution of Beecher. The greater development of roots in subsoil layers resulted in more water extracted and greater dry matter and grain yield (Gregory [1987]). The rate and pattern of root growth in the soil vary with soil physical and chemical properties: the genetic potential of the root system of a crop is only one of the many para- meters that influence root distribution. Parker et al. [1989a] observed that corn grown on sandy Atlantic Coastal Plain soils, which showed the highest exchangeable and HNO 3 extractable K concentrations in Bt horizons, did not respond to K application and was actually able to absorb K from sandy fractions of K bearing minerals in subsoil layers situated at a depth between 36 and 113 cm from the soil surface. Utilization of subsoil K derived from B2 horizons by corn plants was testified by declines in DDA-K throughout the soil profiles on plots which did not receive fertilizer K for three years, although some leaching evidently occurred in the sandier soils (Parker et al 1989b]). 138 High irrigation Low irrigation 0 4 8 12 18 N ocotton -c 36 45 0 4 8 12 ?18 barley 0 V) 36 45 - 3 Root density (kmi m ED 0-20 E] 41-60 [81-100 E 21-40 1 61-80 U >100 Figure 1. Root density distribution of cotton and barley as affected by the irrigation regime and depth of more fertile top soil layer (bold black line). Adapted from Gulick et al 119891. 139 Forage grasses and legumes possess extensive root systems reaching even deeper soil layers (Evans [1977]; Lanyou and Smith [1985]) and certainly have high potential for exploiting subsoil K whenever available. Coastal Bermuda grass (Cynodon dactylon L.) grown on a Grossarenic Paleudult where no yield response to K fertilizer was observed during five consecutive years, was found to be able to deplete soil solution K concentration down to a depth of 160cm and exchangeable K to a depth of 235 cm (Hions et aL 119761). No quantitative evaluation of the contribution of subsoil K to the total K uptake by plants was given. Root distribution of alfalfa (Medicago sativa L.) suggests that sub- soil exploited K might constitute a considerable fraction of absorbed K, when this crop is grown on soils with high subsoil K content. Peterson and Smith 11973] actually found substantially lower recoveries from K2S0 4 placed at various depths in soil under established alfalfa than from surface placed fer- tilizer. By considering their experiment from another point of view, however, their data show that under high subsoil K availability conditions, a higher amount of K may be absorbed by alfalfa plants from subsoil than from top- soil. Moreover, as root density is certainly higher in the 0-30 cm layer, alfalfa appeared to absorb K more efficiently in the subsoil than in the topsoil. Com- petition caused by overlapping of depletion profiles is a likely explanation and represents a possible cause of enhanced contribution of subsoil K to plants. At the first harvest, K-uptake from surface and topsoil was higher than from subsoil (Table 3): obviously contribution of subsoil K to plant up- take increases with time because of increasing rooting depth (see also data by Grimme et al. [19861 on Table 4). This probably represents the most severe limitation to subsoil K-uptake in many crops: root length density in subsoil reaches a maximum only shortly before senescence (Figure 2). Schenk and Barber [19801 studying K-uptake by three corn genotypes grown in the field, observed that at 47 days after sowing (d.a.s.), more than two thirds of the total root surface was in the topsoil whereas at 68 d.a.s. more than 5007o of total root surface was in the subsoil. In this case the contribution of subsoil K to plants remained low, in spite of occurrence of topsoil root competition, because plants absorbed little K between the second and the third harvest. Retarded development of roots might have been caused by the particular water regime of the soil which was an aquic Argiudoll or, for the same reason, root functioning in the subsoil might have been hampered by insufficient aeration (Brouwer [1977]). Table 3. Recovery of K fertilizer applied at various depths as K2SO 4 in an established field. Adapted from Peterson and Smith [1973J K application K (kg/ha)* K (kg/ha)* depth 1st harvest 2nd harvest Surface ...... 62.8 91.5 0-30 cm ...... 54.6 108.0 30-90 cm ...... 26.9 116.6 * difference from control 140 Table 4. Relative contribution of individual horizons of an Alfisol to K supply to plant roots. From Grimme et al. [1986] Date 20. V. I. VI. 10. VI. 20. VI a) at field capacity Ap ( 0-30) ...... 68/ 79% 73% 66% A l (31-50) ...... 2807a 7% 1207o 13 To Bt (51-75) ...... 4076 9% 130oo 17176 Btv (76-90) ...... - 5% 2% 4% b) at actual soil water content Ap ( 0-30) ...... 6607o 720D 65% 500o% Al (31-50) ...... 29% 8010 11% 1301o Bu (51-75) ...... 5% 12% 21% 29% Btv (76-90) ...... - 8% 4% 807o Root 0 a a o Maximum tillering density O0 0 Panicle initiation (cm/cm 3) N N 8 Harvest 0 2 000 0 0 0 1 0 0 0 0000 0se* 20 40 60 Soil depth (cmn) Figure 2. Changes in root density distribution along the soil profile during growth of rice plants on a free draining soil. Adapted from Beyrauly [19871. Diffusion, which is the main mechanism supplying K from the soil to the root soil interface, depends on the volumetric moisture content, but soil moisture markedly affects K-uptake by also influencing root distribution. Ga- ray and Wilhelm [1983/ observed that after 30 days drought, root length den- sity of two soybean isolines was greatest in the 30-120 cm layer. Irrigation favours high root development in surface layers, but deeply irrigated plants produce more roots in the lower depth than shallow irrigated plants (Jodari- Karimi et at. [1982]). 141 The distribution of deep roots is influenced by water content in the surface soil (Figure 3). Deeper roots grow about 8 times faster when surface soil is maintained at 20% of saturation than when they are maintained at 50%0 satu- ration (Malik et al. [1979]). The effect of soil water potential is probably not very important per se, but exerts a strong effect through its influence on penetration resistance of soil (Kleppler [19871). In some soils, formation of hard plough-sole layers may hamper drainage as well as represent a mechanical impedance to root growth. Subsoiling, which consists of loosening the soil beneath the plough layer without turning the subsoil or bringing it to the soil surface, reduces soil compaction with benefi- cial effects lasting for about five years (Ide et at. [1987a]). Removal of com- pacted plough sole layer, characterized by penetration resistances >3 MPa, resulted in increased root quantity and activity below 50cm depth and increas- ing yields of sugar beet (Ide et al.[198 7b]). Potassium uptake was also particu- larly improved by subsoiling. Root growth of winter barley increased signifi- cantly after subsoiling of a hard plough layer: at flowering stage, total root weight between 50 and 75 cm depth was 325% higher in subsoiled plots than in control. Potassium concentration in roots and green mass was increased by 3007 by subsoiling. 0.2 ,0 & 27/6 .011/7 27/6 1.0o 11/7_ 3 11/8 11/8 1.6 non irrigated irrigated 4 8 12 16 4 8 12 16 active roots (r/rn) Figure 3. Influence of irrigation and plant growth on distribution of active roots of soy- bean. Adapted from Huck et aL [1987] 142 5. References Beyrouty, C.A., Wells, B. R., Norman, J R., Marvel, J N. and Pillow, Jr. J A.: Characteri- zation of rice roots using a minirhizotron technique. In: Minirhizotron observation tubes: Methods and Applications for Measuring Rhizosphere Dynamics, pp. 99-108. ASA Special Publication No. 50, Madison, Wis., USA (1987) Brouwer, R.: Root functioning. In: Environment Effects on Crop Physiology J. J. Lundberg and C. V. Cutting (Eds.), pp. 229-245. Academic Press, London (1977) Evans, P S.: Comparative root morphology of some pasture grasses and clovers. New Zeal. J. Agr. Res. 20, 331-335 (1977) Garay, A. F and Wilhelm, W W.: Root system characteristics of two soybean isolines under- going water stress conditions. Agron. J., 75, 973-977 (1983) Gregory, R .: Development and growth of root systems in plant communities. pp. 147-166. In: Root development and functions. (P J. Gregory, J. V. Lake and D. A. Rose Eds.) Cambridge Univ. Press, Cambridge (1987) Grimme, H, Strebel, 0., Renger, M. and Fleige, H.: An estimate of the potassium supply from subsoil. Potash Review, Subject 6, 54th suite, and personnal communication (1986) Gulick, S.H., Kasman, K. G. and Grattan, S. R.: Exploitation of soil potassium in layered profiles by root systems of cotton and barley. Soil Sci. Soc. Amer. J., 53, 146-153 (1989) Haak, E.: Nutrient uptake from subsoil by cereals. In: Agricultural yield potentials in con- tinental climates. Proc. 16th Coll. Int. Potash Inst., pp. 87-94 (1981) Hannay, Y Y and Johnson, J. i: Potassium nutrition of soybean. In: Potassium in Agriculture (R.D. Munson Ed.) pp. 754-764 ASA Madison, Wis., USA (1985) Hons, F M., Dixon, J. R and Matacha, J. E.: Potassium source and availability in a deep, sandy soil of East Texas. Soil Sci. Soc. Am. J., 40, 370-373 (1976) Horst, WJ, Wagner, A. and Marschner, H.: Effect of aluminium on root growth, cell- division rate and mineral element contents in roots of Vigna unguiculata genotypes. Z. Pflanzenphysiol., 109, 95-103 (1983) Huck, M. G., Hoogenboom, G. and Peterson, C. M.: Soybean root senescence under drought stress. In: Minirhizotron observation tubes: Methods and Applications for Measuring Rizhosphere Dynamics pp. 109-121. ASA Special Publication No. 50, Madi- son, Wis., USA (1987) Ide, G., Hofman, G., Ossemerct, C. and Van Ruymbeke, M.: Root-growth response of winter barley to subsoiling. Soil Tillage Res., 4, 419-431 (1987) ide, G., Hofman, G., Van Ruymbeke, M. and Ossemerct, C: Influence of subsoiling on the yield of sugar beets. Z. Pflanzenernihr. Bodenk., 150, , 151-155 (1987) Jodari-Karimi, F, Watson, V., Hodges, H. and Whisler, F: Root distribution and water use efficiency of alfalfa as influenced by depth of irrigation. Agron. 1., 74, 207-210 (1982) Joost, R. E. and Moveland. CS.: Root development of sericea lespedeza and alfalfa in acid soils. Agron. J., 78, 711-714 (1986) Klepper, B.: Origin, branching and distribution of roots. In: Root development and function (RI Gregory, J V Lake and D.A. Rose Eds.) pp. 103-124. Cambridge Univ. Press, Cambridge (1987) Lanyou, L. E. and Smith, S. W.: Potassium nutrition of alfalfa and other forage legumes temperate and tropical. In: Potassium in Agriculture (R. D. Munson Ed.) pp. 861-888. ASA Madison, Wis., USA (1985) Manson, A. D. and Fey, M. V: Cation type and ionic strength effects on the solution com- position of an acidic subsoil. J. Soil Sci., 1, 577-583 (1989) Malik, R. S., Dhankar, J L and Turner,M. C.: Influence of soil water deficits on root growth of cotton seedlings. Plant and Soil, 53, 109-115 (1979) Marita, C F: Soil of the United States, Atlas of American Agri. U.S. Government Printing Office, Washington, D.C. (1935) 143 Martin, H. W and Sparks, D. L.: Kinetics of nonexchangeable potassium released from two Coastal Plain soils. Soil Sci. Soc. Am. J., 47, 883-887 (1983) Parker, D. R., Sparks, D. L., Hendricks, C. J and Sadusky, M. C.: Potassium in Atlantic Coastal Plain soils: I. Soil characterization and distribution of potassium. Soil Sci. Soc. Am. J., 53, 392-396 (1989a) Parker, D. R., Hendricks, G. £ and Sparry, D. L.: Potassium in Atlantic Coastal Plain soils: 11.Crop responses and changes in soil potassium under intensive management. Soil Sci. Soc. Am. J., 53, 397-401 (1989b) Peterson, L.A. and Smith, D.: Recovery of K2SO4 by alfalfa after placement at different depth in a low fertility soil. Agron. J., 65, 769-772 (1973) Schenk. M. K. and Barber, S.A.: Potassium and phosphorus uptake by corn genotypes grown in the field as influenced by root characteristics. Plant and Soil, 54, 65-76 (1980) Sharpley, A. N.: Relationship between soil potassium forms and mineralogy. Soil Sci. Soc. Am. J., 53, 1023-1028 (1989) Smith, G. D., A I/away, WIH. and Riecken, F F: Prairie soils of the upper Mississippi valley. Adv. Agron., 2, 157-205 (1950) Temant, D.: Wheat root penetration and total available water on a range of soil types. Austr. J. Exp. Agric. Husb., 16, 570-577 (1976) Vilela, L. and Ritchen, R.: Potassium in intensive cropping systems on highly weathered soil. In: Potassium in Agriculture (R. D. Munson Ed.), pp. 1155-1175. ASA Madison, Wis., USA (1985) Winters, E. and Simonson, R. W.: The subsoil. Adv. Agron., 3, 2-92 (1951) Woodruff J R. and Parks, C. L.: Topsoil and subsoil potassium calibration with leaf potas- sium for fertility rating. Agron. J, 72, 392-396 (1980) 144 Dynamics of Soil Potassium in Multicationic Systems Sala Feigenbaum, A. Bar-Tal and D. L. Sparks* Summary The increasing use of brackish water for irrigation calls for more information on the effect of water salinity and ion composition on potassium reactions in soil and on the availability of K to plants. The objectives of this work were to: (i) study plant response to K as affected by salinity; (ii) study K dynamics in soil as affected by irrigation with saline-sodic water; and (iii) test the validity of exchange selectivity coefficients obtained in binary systems for prediction of ternary exchange of K in a multicationic system. The reactions and fate of native and fertilizer K under saline conditions were studied in two soils in a screenhouse pot experiment with corn (Zea mays (L.) cv. «Jubilee>). Potassium adsorption by the soils from multicationic solutions was studied using batch and miscible-displacement methods in the laboratory. Salinity level in the irrigation water had a significant effect on yield decrease in both soils and there was a significant yield response to K application in the Nordiya soil. In both soils K fertilization increased K concentration and reduced the ratio Na:K in plant tissue. Potassium application moderated exchangeable K depletion in both soils. Potassium was preferred over the divalent cations Ca and Mg in both soils, regardless of the experimental method and Na concentration. A simple mathematical solution of the Gapon binary equations for K-(Ca+Mg) and Na-(Ca+Mg), in combination with the assumption of a constant CEC, was used to predict the amount of exchangeable K as a function of solution composition. The ability to predict the exchange of K with Ca + Mg and Na in a miscible displacement system using information obtained in binary systems was evaluated. 1. Introduction Release of interlayer K from mica-like minerals in semiarid and arid regions is a source of plant available K, but the major source of K for plant nutrition in most agricultural soils is exchangeable K (McLean and Watson [1985]). Ex- changeable K is most commonly used as an index for estimating available K (Carson and Dixon [1972). However, this parameter does not always ade- quately predict response to applied K since non-exchangeable K forms can * Dr. Sala Feigenbaum and A. Bar-Tal, Institute of Soils and Water, Agricultural Research Organization, The Volcani Center, Bet Dagan 50250, Israel; Dr. D. L. Sparks, Dept. of Plant and Soil Science, University of Delaware, 147 Townsend Hall, Newark, DE 19717-1303, USA 145 play an important role in supplying K to plants over a period of time (Feigen- baum and Hagin [1967]; Sadusky et al [19871). Fertilizer recommendations under non-saline conditions are based on crop requirement and K availability indexes (Mengel and Kirkby [19801). Available K is expressed basically as a ratio between K concentration and (Ca + Mg) concentration in the soil solu- tion which is in equilibrium with the solid phase of the soil (Beckett [1971]; Woodruff[19551). This ratio is an intensity index, and exchangeable K a capac- ity index (McLean and Watson [1985]). Potassium, one of the major plant nutrients, may behave differently under saline and non-saline conditions, from both plant and soil aspects. Sewage effluents and brackish waters contain considerable quantities of cations such as Na , K , Ca + 2and Mg 2. These cations undergo simultaneous exchange reactions with each other on the soil surface. For example, the possible reac- tions involving K-Ca-Mg, K-Na-Ca and K-Na-Mg in soils, and the study of their effect on K distribution between the solution and the solid phases are extremely important for plant nutrition. Irrigation with brackish water, in which the concentrations of Ca, Mg and Na are higher than in good quality water might cause an increase in downward movement of K beyond the root zone. With increasing concentrations of competing cations, larger fractions of K from the soil or from fertilizer sources are found in the solution phase (Feigenbaum et al. [1988]). Thus K may be more available to plant roots, but is also easily leached from the root zone, especially where excess leaching is required to ameliorate soil salinity (Feigenbaum [1986]). Salinity and low K fertility are important growth limiting factors; decreas- ing or eliminating one factor may affect the crop's response to the other (Feigin [1985]). Research on the interaction between salinity and fertility as it affects crop response has been reviewed by Feigin [1985] and Kafkafi [19851, but most of the studies have been concerned with salt-plant nutrition relationships for applied N and P fertilizers and only a few have involved K. Exchange reactions involving K with cations such as Ca, Mg and Na in clay minerals and soils, have been investigated intensively and reviewed by Bolt [1979] and Sparks [1987]. Our theoretical understanding of the relation- ship between ternary and binary exchange systems in clays and soils has im- proved during the last decade. Chemical and geochemical models were adapted to soils for predicting cation exchange in a ternary system using binary exchange data (Elprinceet al [19801). Chu and Sposito [1981], using data obtained from careful experiments with pure clay minerals, found that the effect of the ternary systems on the binary exchange coefficient was no greater than the experimental error. Unfortunately, reports that have appeared in the soil chemistry literature on exchange phenomena involving binary and particulary ternary systems, do not involve K (Chu and Sposito [1981]; El- prince et al. [1980]). The objectives of the current research were to: (i) study plant response to K as affected by salinity; (ii) study K dynamics in soil as affected by irrigation with saline-sodic water; and (iii) test the validity of ex- change selectivity coefficients obtained in binary systems for prediction of ternary exchange of K in multicationic system. 146 2. Materials and Methods Two soils, representing high and low available K, were chosen for the study of the interactive effect of salinity and fertility on plant growth and the dy- namics of K reactions in soil. The soils were collected from control plots of fields where irrigation experiments with brackish water were being conducted: one from the south of Israel - at Gilat, and one from the Coastal Plain - at Nordiya. The basic chemical and physical properties of the soils were deter- mined using standard methods and their properties are given in Table 1. This study included a corn pot experiment in a screenhouse, and a laboratory study of K exchange reactions in a multicationic system. Table 1. Selected chemical and physical properties of the studied soils Soil Soil CaCO 3 Organic Particle size Tex- pH CEC Exch. Subgroup matter Sand Silt Clay ture paste K ------g kg I------cmolkg - Gilat Calcic 203.0 9 436 368 196 1 7.8 11.0 0.80 Haploxeralf Nordiya Typic 1.4 14 854 48 98 s 7.5 6.2 0.50 Rhodoxeralf 2.1 Corn pot experiment Sweet corn (Zea mays (L.) cf. «Jubilee ) was grown in pots containing 3.0 kg of air dry soil. The experiment included nine treatments in six replications in a factorial design. Potassium application rates per pot were: 0, 15 and 30mmol K in the Gilat soil and 7.5, 15 and 30 mmol K in the Nordiya soil. The salinity levels St, S2 and S3 of the irrigation water were 4, 20 and 40 mmol 1 charge L- , respectively, with SAR values of 10 for S2 and S3, using a salt mixture of NaCI, CaC 2 and Ca(N0 3)2 in both soils. The plants were irrigated daily to bring to volumetric water fraction at pot capacity, plus 10-20% leach- ing (for more details see Bar-Tal et al. [19901). Irrigation and drainage volumes were determined throughout the experiment period. After the last harvest, 58 days after seeding, the soil in each pot was divided into four horizontal layers and exchangeable K and Na were extracted from the soils using I M ammonium acetate. The leachate and saturated pastes of the soils were ana- lyzed for EC, Cl, K, Na, Ca and Mg. 2.2 Laboratory experiment Exchange isotherms for binary and ternary systems of K, Na, Ca and Mg were performed with each studied soil at room temperature, 286+2.0 K, to obtain a complete description of the exchange reactions. Binary exchange 147 reactions were conducted using a batch technique. Soil samples were equilibrated with two sets of solutions with a constant C10 4 concentration of 20±0.5 mol m-': the first with a K-Mg salt mixture and the second with Na-Mg. The initial Mg-equivalent fraction in the solution of the two sets varied from 0 to 1, using two replicates. Soil samples were equilibrated with the different salt mixture solutions, at a soil:solution ratio of 1:5. Ternary exchange reactions were conducted using a miscible-displacement technique (Jardineand Sparks [1984]). The soils were equilibrated with solu- tions with a varied K/Na/Ca/Mg ratio in a constant Cl background concen- tration of 20±0.5 mol m - 3. Exchangeable cations were extracted with 0.5 M LiCI solution in batch and miscible-displacement methods (for more details see Feigenbaum et at. [19901). Potassium and Na concentrations in the equilibrium and displaced solutions from the binary and ternary systems were determined by flame photometry. Calcium and Mg in the equilibrium solu- tions were determined by atomic absorption spectrometry. Calcareous soils pose difficulties in the determination of adsorbed Ca and Mg. Consequently, the adsorbed Ca plus Mg in each treatment was calculated from the difference between the total CEC (Table 1) and the displaced Na and K, assuming constant CEC. This assumption is based on published data showing constant CEC of montmorillonite and montmorillonitic soils (El- prince et aL [1980; Jensen [1973]). From the above batch and miscible dis- placement techniques, K-(Ca+Mg) isotherms were drawn to present the ad- sorbed cation ratio as a function of soluble cation and the Gapon model was used to calculate selectivity coefficients, k0 . 3. Results 3.1 Corn pot experiment Fresh and dry weights of corn, for a given K treatment, decreased significantly with increasing salinity level in both soils (Table 2). In the Nordiya soil there was a significant positive yield response to K application 46 days after seeding. There was a significant interaction in yield response between K application and salinity level of the irrigation water at the last harvest (Table 2). Hence, at this harvest, the effect of K was analyzed separately for each salinity level and significant yield increase was obtained in S 2 water, while in S, there was yield increase from K1 to K2 and an unexplained decrease from K2 to K3; there was no significant effect of K application on yield using S 3. Potassium concentration in the plant tissue was enhanced by K increasing the K application rate and the salinity level of the irrigation water. Sodium concentration of the plant tops was not affected by K application in the Gilat soil, while in the Nordiya soil there was an interaction between K application and salinity effects on the tissue Na content; K application significantly reduced the plant tissue concentration only in the high salinity irrigation. Potassium application reduced by a factor of 1.5-3.0 the Na:K plant molar 148 concentration ratio, which has been considered an important factor in the effect of sodic saline water on plant growth (Liiuchli and Stelter [1982]; Helal and Mengel [1979]). Table 2. Corn yield and plant K and Na concentration as affected by salinity and K appli- cation Fresh weight Nutrient content at last harvest (g/plant) (mol/kg) Treatment 46 day 58 day K Na Na/K Gilat KS...... 91.27 126.80 0.37 0.02 0.04 K2S1 ...... 91.43 138.90 0.53 0.01 0.02 K3 St ...... 108.00 135.49 0.60 0.01 0.02 KIS, ...... 66.07 89.49 0.49 0.28 0.57 KZS 2 ...... 62.08 99.32 0.63 0.23 0.36 KS 2 ...... 61.48 92.80 0.74 0.29 0.39 KS ...... 50.37 63.69 0.52 0.47 0.91 K2S3 ...... 50.28 67.43 0.69 0.40 0.58 K3S 3 ...... 48.83 81.84 0.79 0.47 0.60 Source df ...... Probability>F Model 10 ...... 0.000 0.000 0.000 0.000 0.000 Block 2 ...... 0.442 0.123 0.150 0.273 0.160 S 2 ...... 0.000 0.000 0.000 0.000 0.000 K 2 ...... 0.487 0.184 0.000 0.185 0.001 S • K 4 ...... 0.200 0.615 0.925 0.821 0.160 Nordiya KSj ...... 97.27 174.39 0.12 0.01 0.08 K2S ...... 101.50 204.68 0.18 0.01 0.06 K3S...... 122.97 173.27 0.31 0.01 0.03 KIS, ...... 88.35 129.11 0.17 0.37 2.21 K2 S2 ...... 95.90 139.10 0.24 0.47 1.94 K3S2 ...... 96.32 159.26 0.43 0.43 1.01 KS 3 ...... 69.45 110.75 0.19 0.62 3.30 K2S ...... 77.38 104.84 0.34 0.58 1.73 KJS3 ...... 76.33 121.54 0.50 0.48 0.95 Source df ...... Probability>F Model 10 ...... 0.000 0.000 0.000 0.000 0.000 Block 2 ...... 0.198 0.785 0.164 0.440 0.760 S 2 ...... 0.000 0.000 0.000 0.000 0.000 K 2 ...... 0.003 0.035 0.000 0.196 0.000 S - K 4 ...... 0.087 0.001 0.000 0.029 0.000 149 Soil analyses after the harvest show that Ksolub, Kexch and PAR, which have been widely used as availability parameters, were increased significantly by K application, indicating that the K application treatment was indeed effective (Table 3). After the last harvest, the exchangeable K levels in the Nordiya and Gilat soils under low and medium K application dropped considerably below the initial exchangeable K values found in the soils. Even though there was no corn yield response to K application in the Gilat soil, the highest K applica- tion eliminated exchangeable K depletion (Table 3). Large differences in ex- changeable K between the Nordiya and Gilat soils were found in identical treatments. The highest capacity parameter for available K (exchangeable K) in the Nordiya soil was 0.28 cmol kg -I which was lower than the lowest value in the Gilat soil, 0.32 cmol K kg- I. In both soils, soluble and exchangeable Na were a function of the salinity of the irrigation water, but were unaffected by K applications. Table 3. Concentration of soluble and exchangeable ions in Gilat and Nordiya soils as affected by K and salinity treatments Soluble (meq L-') Exchangeable (cmol kg i) Treatment Cl K Na Ca Mg PAR SAR EXK EXNa EPP Gilat K,S, 1.8- 0.05 1.66 4.52 1.03 0.032 1.4 0.32 1.32 2.91 K2S, 2.0 0.09 1.56 4.32 1.19 0.056 1.3 0.39 1.45 3.55 KS, 2.4 0.19 1.16 3.71 1.27 0.122 1.0 0.55 1.12 5.00 K,S, 62.9 0.11 47.94 9.39 2.87 0.043 27.6 0.32 2.67 2.71 K2 S2 50.5 0.18 40.73 5.76 1.70 0.095 30.3 0.45 3.33 4.09 K3S2 57.5 0.54 45.11 7.94 2.50 0.234 28.1 0.71 2.86 6.45 KS 3 101.8 0.18 60.51 19.93 5.19 0.050 24.5 0.37 3.18 3.36 K2S3 102.1 0.43 62.57 21.09 4.75 0.119 25.5 0.57 3.14 5.18 K3S, 103.3 0.95 60.94 18.90 4.68 0.277 26.4 0.78 3.36 7.09 Nordiya KS, 1.4 0.08 1.47 6.59 0.28 0.042 0.8 0.18 1.26 2.88 K2S, 2.2 0.10 1.73 8.09 0.34 0.047 0.8 0.19 1.45 3-04 K,,S 2.2 0.11 1.68 7.00 0.34 0.055 0.9 0.20 1.31 3.20 KS2 25.8 0.06 27.26 4.15 0.24 0.037 18.7 0.14 2.51 2.24 K2S2 22.4 0.07 24.15 4.60 0.26 0.045 15.8 0.17 2.37 2.72 K3S2 24.9 0.09 27.38 4.59 0.24 0.058 18.1 0.20 2.53 3.20 KS3 53.3 0.08 41.47 15.93 0.34 0.026 14.9 0.15 2.42 2.40 K2S, 49.7 0.11 39.17 15.15 0.40 0.039 14.3 0.16 2.39 2.56 K3S 52.8 0.28 42.83 17.37 0.45 0.092 14.7 0.28 2.44 4.48 Represents average concentrations for the four soil layers A linear relation between the exchangeable potassium percentage (EPP= 1 100 qK/CEC) and the soluble cation ratio (PAR = CK/(CCa+ CMg) 2) was ob- tained in both soils (Figure 1). In the current study this relationship was found 150 10 a 8- 8 Gilat 4- 2- R'=0.96 0 4 0 5 0.6 00 1 0'.2 013 0 PAR 4-S b 4 Nordyia 1.62 + 26.87X 2-Y= R2 =0.88 0.16 0.20 0.08 0.12 0. 0 04 PAR Figure 1. Exchangeable potassium percentage (EPP) as a function of the potassium ad- sorption ratio (PAR) in the solution of the soil pastes obtained in the pot experi- ment under three salinity levels. 151 to be independent of soil salinity and for a wide range of Na concentrations: from 2 to 60 mmol L - 'for the Gilat soil, and from 2 to 43 mmol L - ' for the Nordiya soil (Table 3). The SAR values were similar over a wide range and neither one directly affected the relation between adsorbed and soluble K fractions. All the results can be described by one linear regression line for each soil (Figure 1). The slopes and intercepts were related to the soil type. In order to understand the effect of Na on K exchange with (Ca+Mg) in these soils, binary and ternary exchange experiments were conducted. 3.2 Binary and ternary exchange experiments Potassium exchange with the divalent cations, Ca and Mg, was analyzed, treat- ing Ca and Mg as identical cations, following the approach of the US Salinity Laboratory Staff [19541. This approach was justified by the non-preference of Ca over Mg for the Nordiya (Seyfried et al. [1989]) and Gilat soils (Feigen- baum et al. [19901). To determine which cation is preferred over the other in mono-divalent exchange systems, the renormalized K equivalent fractions in solution, EK', and on the solid phase, PK', were calculated according to Thellier and Sposito [1988], excluding Na concentration in both phases: EK' - CK (1) CK + 2(Cca + CMg) EK qK (2) qK + 2(qca + qM.) where C is the concentration in the soluble phase (mol L- '), and q is the quantity of adsorbed cations (mol kg- '). The isotherms of tK' vs EK' for the Nordiya and Gilat soils are presented with calculated lines (after Jensen [1973]) for nonpreference and ideal solution conditions for the liquid and solid phases (Figures 2a, 2b). The experimentally observed points for the two soils lie well above the theoretical curve, indicating a preference for K over (Ca+ Mg). Similar results were found for clay minerals and for soils (Jensen [19731; Schwertmann [1962). To examine whether sodium concentration affected K exchange with (Ca+Mg), isotherms of PK' vs EK', obtained from the ternary system (Table 4) were compared with the binary isotherms. It can be seen that the presence of Na in the ternary system did not alter the curve of aK' vs EK', in either the Nordiya or the Gilat soil, respectively (Figures 2a, b). The selectivity coefficients (kG) for the exchange reaction in the binary and ternary systems were determined using the following modified Gapon model (Sposito and Mattigod[1987]), in which ion activity ratio and mol L -' units are used, rather than concentration and mmol L - units, as used by Gapon [19331. 152 EPR = kGK PARi (3) where PARi = and EPR = / 2 2 f(Cca + CM,)" ( qca + 2qm,) and ESR = kGNa SARi (4) where 3 Na SARi = N. and ESR = 2 f(Cca + CM,)"1 (2qca + 2qMS) Table 4. Miscible displacement exchange data for K with Na, Ca and Mg in 20 mol m- 3 C) background CK CNa C(ca +M) PARI SARI qK qNa qca + qMg* kG 0 0 mol m-, (mol/L) ' cmol kg -' (mol/L)o Gilat 2.2 10.3 4.12 0.034 0.161 1.28 0.52 3.90 4.9 2.2 10.1 4.00 0.034 0.159 1.64 0.35 3.81 6.3 4.3 5.0 5.72 0.057 0.066 1.97 0.35 3.64 4.7 4.3 9.8 3.19 0.077 0.173 2.30 0.35 3.47 4.3 4.2 12.1 2.13 0.091 0.259 2.65 0.38 3.28 4.4 6.1 10.2 2.08 0.134 0.223 3.31 0.32 2.98 4.1 5.4 12.6 1.18 0.157 0.366 3.32 0.61 2.83 3.7 7.2 9.2 2.08 0.158 0.202 3.32 0.52 2.88 3.6 7.7 8.7 2.13 0.165 0.187 3.02 0.26 3.16 2.9 8.6 7.9 2.08 0.187 0.174 3.92 0.13 2.71 3.9 9.6 7.1 2.08 0.210 0.154 3.78 0.22 2.81 3.2 10.5 6.0 2.13 0.226 0.129 4.38 0.21 2.51 3.9 14.7 5.4 1.38 0.434 0.160 5.00 0.70 1.95 2.9 Nordiya 2.2 10.1 4.21 0.033 0.156 0.69 0.22 2.10 4.7 2.2 10.3 3.95 0.035 0.164 0.64 0.22 2.22 4.2 4.3 4.9 5.67 0.058 0.066 0.95 0.17 2.09 4.0 4.3 9.8 2.81 0.082 0.185 1.12 0.22 1.98 3.5 4.2 12.0 2.13 0.090 0.259 1.35 0.23 1.86 4.0 6.1 9.8 2.08 0.134 0.214 1.53 0.18 1.79 3.2 7.1 3.9 4.55 0.147 0.058 1.76 0.01 1.77 3.4 7.3 9.2 2.15 0.157 0.199 1.41 0.18 1.86 2.4 7.7 8.7 2.06 0.169 0.191 1.66 0.43 1.61 3.1 5.4 12.6 0.94 0.177 0.410 1.60 0.27 1.71 2.6 8.6 7.9 2.08 0.187 0.174 1.81 0.13 1.68 2.9 9.8 7.0 1.94 0.223 0.158 2.24 0.27 1.39 3.6 10.5 6.0 2.13 0.227 0.129 2.12 0.11 1.53 3.0 14.3 5.0 0.97 0.459 0.160 2.65 0.16 1.24 2.3 * Calculated as the difference between total CEC and qK+qNa. 153 1.0- a Gilat 0.8- . Binary 0.6- * Ternary 0.4 A * A 0.2- A Nonpreference 0.01 0.0 0.2 0.4 0.6 0.8 1.0 1.0- b Nordiya 0.8- * Binary 0.6- * Ternary • A 0.4- 0.2-A E: • Nonpreference 0.0-j 0.0 0.2 0.4 0.6 0.8 1.0 EK Figure 2. Renormalized isotherim for K exchange with (Ca +Mg) in binary.and ternary systems. 154 where a and fare ion activity (mol L - ')and activity coefficient in the solution, respectively. The activities of the cations in the solutions were calculated using the Davies equation (Sposito [19811). The observed exchangeable potassium ratio (EPR) as a function of solution K ratio (PARi) of the two soils with the regression lines is presented in Figures 3a, b. The kGK value in the Nordiya and Gilat soils was 2.4±0.05 and 2.9±0.04 (mol L 1)0.5 with r' values of 0.989 and 0.996, respectively. Sodium exchange with (Ca + Mg) in the Nordiya and Gilat soils is presented in Figures 4a, b. The values of koNa for the Nordiya 2 and Gilat soils were 0.28±0.01 and 0.42±0.02 (mol L- 1).5, with r values of 0.98 and 0.99, respectively. The values of kG' for K-(Ca+Mg) and Na-(Ca+ Mg) exchange in both soils are of the same magnitude as published values for montmorillonite and montmorillonitic soils (Bolt [1979/; U.S. Salin- ity Lab Staff [1954]). Despite the high correlation of EPR with PAR in both soils, there are large deviations of the observed data from the model for low EPR values. To improve the prediction of EPR as a function of activity ionic ratio in the solutions (PAR), empirical functions relating kGK values to cor- responding EPR were determined using the SAS STEPWISE PROCEDURE (SAS [1985]) and are presented in Appendix A. The data for K exchange with (Ca+Mg) in both soils in the presence of Na and using the miscible displacement method are presented in Table 4. This was used to calculate P:x as a function of EK as presented in Fig- ures2a, b together with the binary data. Preference of K over (Ca+Mg) was found regardless of Na presence. The calculated EPR values as a function of PARi in the ternary system are presented together with the binary data (Figures 3a, b); the ternary experimental data are reasonably described by the calculated line from the binary system. To predict exchangeable K in ternary systems all the cations participating in the exchange reactions have to be taken into account. Assuming a constant CEC comprised of K, Ca, Mg and Na, the following equation can be introduced: 2 q= CEC-[qN. + (qca + q~M)J (5) If the Gapon selectivity coefficients which were obtained in the binary system hold in the ternary system, the following equation is obtained: qi' = CEC kGK PARI (6) PARi kGK + kGNa SARi-I A complete description of the mathematical steps to obtain Eq. (6) is given in Appendix B. The predicted values of exchangeable K (qK) using Eq. (6) were compared with the measured qK values and linear relationships were obtained for both soils (Figures 5a, b). Most of the predicted points lie near the 1:1 line but there is a slight underestimation of qK by the prediction method. The underes- timation of qK by the model is due to the underestimation of EPR and probably to the assumption of a constant CEC. The standard deviation in 155 1.6- - a 1.4- Gilat 1.2- 1.0- * Binary CC o Ternary 0 M0.8 0 0 0.6- • R=0.996 0.4- 0 0e 0.2- 0 0.0 0 0.2 0.4 0.6 PAR i 1.2- b - Nordiya 1.0- 0 Binary o Ternary 0.8- o0.6 - 0 0o 0 " Y=2.43"X 0.4- 0 0. 0.2 - 0.0, 0 0.2 0.4 PARi Figure 3. Exchangeable potassium ratio (EPR) as a function of solution K ionic ratio (PARj) in binary and ternary systems. 156 0.20 a Gilat 0.15- " YR= =0.42*X 0.99 o 0.10 0.05 0.00 0.0 0.2 0.4 SAR i 0.20 b Nordyia 0.15 Y=0.28*X - R2=0.98 Cr (n 0.10- 0.05. 0.00 0.0 0 .2 0.4 SARi Figure 4. Exchangeable sodium ratio (ESR) as a function of solution ionic ratio (SARi). 157 6 a O Gilat 0 E 04 00 0 246 Actual Exch. K (cmol/kg) 4 ' 2.8 . .4 - b 2 . - N o rd iy a1 :1 2.0 00 2. . 4. . . 6.. .s, 1.6- 7 ,x1.2- J 0.0 Actual Exch. K (cmol/kg) Figure 5. Predicted exchangeable K in oternary)) system using the Gapon model and coefficients obtained from the obinary> systems, versus the measured exchange- able K in the oternary)) system. 158 determining the CEC of the Gilat and Nordiya soils was 8.3-9.5% of their values, respectively. An overestimation of the CEC value in the ternary system could have led to the underestimation of qK by the model, since the model is based on an assumption of a constant CEC. 4. Conclusions From these studies the following conclusions can be drawn: (i) K fertilization increased corn yield in all salinity levels only in Nordiya soil and reduced the ratio of Na:K in plant tissue in both soils. (ii) A linear relationship between exchangeable potassium percent (EPP) and potassium adsorption ratio (PAR) was obtained in both soils, regardless of salinity level or SAR value. (iii) Potassium preference over the divalent cations (Ca+Mg), was found in both soils, regardless of the experimental method and Na concentration. (iv) A simple mathematical solution of the Gapon binary equations for K-(Ca+ Mg) and Na-(Ca + Mg), in combination with an assumption of a cons- tant CEC, was used successfully to predict the amount of exchangeable K as a function of solution composition. 5. References Bar-Tat, A., Feigenbaum, S. and Sparks, D. L.: Potassium-salinity interactions. Irrig. Sci., in press (1990) Beckett, PH. T: Potassium potential, a review. Potash Rev. 5/30. 1-41 (1971) Bolt, G. H.: Soil chemistry. B. Physico-chemical models. In: G. H. Bolt (Ed.). Development in Soil Science Series. Elsevier PubL., Amsterdam-Oxford-New York (1971) Carson, CD. and Dixon,. 1.: Potassium selectivity in certain montmorillonite soil clays. Soil Sci. Soc. Am. Proc. 36, 838-843 (1972) Chu, S. Y and Sposito, G.: The thermodynamics of ternary cation exchange systems and the subregular model. Soil Sci. Soc. Am. J. 45, 1084-1089 (1981) Elprince, A. M., Vanselow, A. P and Sposilo G.: Heterovalent, ternary cation exchange equilibria: NH4-Ba-La exchange on montmorillonite. Soil Sci. Soc. Am. J. 44, 964-969 (1980) Feigenbaum, S.: Potassium distribution in a sandy soil exposed to leaching with saline water. Proc. 13th Congr. Int. Potash Institute, Bern, 155-162 (1986) Feigenbaum, S. and Meiri A.: The effect of potassium fertilization on cotton response and potassium distribution under irrigation with saline water. BARD report 1-630-83, pp. 88-110 (1988) Feigenbaum, S.. Bar-Tal, A., Portnoy, R. and Sparks, D. L.: Binary and ternary exchange of potassium on calcareous montmorillonite soils. Soil Sci. Soc. Am. J. 54, in press (1990) Feigin, A.: Fertilization management of crops irrigated with saline water. Plant and Soil 89, 285-299 (1985) Gapon, E. N.: On the theory of exchange adsorption in soil. U.S.S.R. J. Gen. Chem. 3, 144-163 (1933) 159 Helal, H.M. and Mengel, K.: Nitrogen metabolism of young barley plants as affected by NaCI salinity and potassium. Pl. and Soil 51, 547-562 (1979) Jardine, PM. and Sparks, D. L.: Potassium-calcium exchange in a multireactive soil system: 11. Thermodynamics. Soil Sci. Soc. Am. J. 48, 45-50 (1984) Jensen, H.E.: Potassium-calcium exchange on a montmorillonite and a kaolinite clay: 1. A test on the Argersinger thermodynamic approach. Agrochimica 17, 181-189 (1973) Lduchli, A. and Stelter, W: Salt tolerance of cotton genotypes in relation to K/Na selec- tivity. In: San Pietro, (Ed.). Biosaline Research: A look to the future. Plenum Press, New York, pp. 511-514 (1982) McLean, E. D. and Watson, M. E.: Soil measurements of plant available potassium. In: R. D. Munson (Ed.). Potassium in Agriculture. ASA-CSSA-SSSA, Madison, WI., USA, 277-308 (1985) Mengel, K. and Kirkby, E.A.: Potassium in crop production. Adv. Agron. 59-110 (1980) Sadusky, M. C, Sparks, D. L., Noll, M. R. and Hendricks, G. J.: Kinetics and mechanisms of potassium release from sandy Middle Atlantic Coastal Plain Soils. Soil Sci. Soc. Am. J.51, 1460-1465 (1987) SAS Institute: SAS users' guide: Statistics. Version 5. Raleigh, North Carolina (1985) Schwertmann, U.: Die selektive Kationensorption der Tonfraktion einiger B6den aus Sedimenten. Z. Pfl. Erntihr. Ding. Bodenk. 97, 9-25 (1962) Seyfried, M. S., Sparks, D. L., Bar-Tal, A. and Feigenbaum, S.: Kinetics of Ca-Mg exchange on soil using a stirred-flow reaction chamber. Soil Sci. Soc. Am. J. 53, 406-410 (1989) Sparks, D. L.: Potassium dynamics in soil. Adv. Soil. Sci. 6, 1-63 (1987) Sposito, G.: The thermodynamics of the soil solution. Oxford University Press, New York (1981) Sposito, G. and Matligod, S. V.: On the chemical foundation of the sodium adsorption ratio. Soil Sci. Soc. Am. 1. 41, 323-329 (1987) Thellier, C. and Sposilo, G.: Quaternary cation exchange on Silver Hill illite. Soil Sci. Soc. Am. J. 52, 979-985 (1988) US Salinity Lab. Staff- Diagnosis and improvement of saline and alkali soils. In: L.A. Richards (Ed.). Handbook U.S. Dept. Agric. No. 60 (1954) Woodruff C. M.: The energy of replacement of calcium by potassium in soils. Proc. Soil Sci. Soc. Am. 19, 30-40 (1955) Appendix A The EPR power values were added to the model until the level of significance did not exceed 0.15. The following equations were obtained: Nordiya soil: koK=5.668-15.673EPR+33.278EPR 2-33.353EPR'+ 12.222EPR 4 0 Gilat soil: 4 koK=6.577-16.755EPR+27.153EPR2-18.656EPR'+4.609EPR 0< EPR <1.5 (A 2 ) k0 K=2.89 EPR= 1.5 160 Appendix B If the Gapon selectivity coefficients which were obtained in the binary system hold in the ternary system, the unknown values qNa and qca+qMg in Eq. (5) can be replaced by the known solution composition values using Eqs. (3) and (4) (the modified Gapon model), and the following equation is ob- tained: qK = CEC - qK kGNa SARi qK (B1) kGK PARt kGK PARS where the kG values are coefficients that are functions of EPR and ESR ac- cording to Eqs. (3-4). Equation (5) was rearranged to calculate qK: = CEC kG PARi qK PARi kGK+ kGN. SARI - I (B 2 ) 161 K-status and Dynamics in Salt Affected Soils I. Szabolcs* Summary Practical fertilizer recommendations must take account of soil type in the interest of fer- tilizer efficiency and economy. Difficulties in connexion with the treatment of salt-affected soils are well-known but the present classification of salt-affected soils is not very helpful and may lead to difficulty and confusion in formulating fertilizer recommendations. For practical purposes there are five groups of such soils: Saline soils: Salt concentration often >207a in surface soil. Usually no K deficiency but K fertilizer needed for high yields. Alkalisoils: High pH but lower salt content than the preceding. Poor physical condition and moisture relationships. The question of K-Na antagonism is sometimes important. Magnesium soils: High in soluble magnesium salts, or adsorbed Mg ions or Mg minerals. K fertilizer requirement variable. Gypsiferous soils: Dominated by calcium sulphate sometimes up to over 500o of total soil. No great agricultural potential. Uptake of K (and other nutrients) affected by acidity, toxicity and Ca-K antagonism. Acidsulphatesoils: In tidal marshes and lagoons, coastal sediments. Dominant minerals pyrite and jarosite, the latter being high in K, hence no K deficiency. 1. Introduction When dealing with the practical questions of fertilization these days it is increasingly important to take into consideration the properties of the soil so that we can elaborate the optimum quality and quantity as well as the best methods of application for each of the different fertilizers. In places where the rates of fertilizers are low, which is the situation in most of the underdeveloped regions (and was in the past in the greater part of the world) soil properties play a lesser role as factors of fertilization efficiency than they do in the developed countries today. With the rates of fertilizers increasing, the soil attributes have more and more influence on the nutrient uptake in the interaction between fertilizers and soils. This is valid not only in respect of complex soil processes but also in respect of economy. The tillers of some soils often receive not only a poor response to fertilizers in yields but are at the same time also guilty of squandering by applying high fertilizer rates which are rendered inefficient or of reduced efficiency by certain soil properties. From the above it can be concluded that the thorough study and * Prof. Dr. . Szabolcs, Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences, Herman Otto ut, H-1022 Budapest, Hungary 163 knowledge of soil properties are indispensable in parallel with the develop- ment of modern fertilizer application techniques. Such considerations are fully valid in respect of salt-affected soils (Szabolcs [1989]). Those soils have been well known for a long period and the obstacles of their utilization are also familiar in both research and agricultural practice. In spite of this, the grouping, classification and characterization of differ- ent salt affected soils are, for the purposes of fertilizer use, poorly elaborated and this results not only in difficulties but often also confusions when it comes to the utilization and improvement of such soils. While it is true on the one hand that the numerous and complicated systems of contemporary soil classification are hardly decipherable and the jargon of several soil taxonomy systems is strange for everybody except professional pedologists, it is also true, on the other hand, that many substantial soil properties are neglected or underestimated by the people in the field when the plans of fertilizer application are elaborated. This is the reason why the 2. A general characterization of salt affected soils Salt affected soils are widely distributed and occur on all the continents, cover- ing about 10% of their total territory. In Table I the areas of salt affected soils on the continents (except Antarctica) are demonstrated. Figure I shows a schematic map of the global distribution of salt-affected soils (Szabolcs [1989]). Table I. World extension of salt affected soils according to continents Continent Area (1000 ha) N orth A m erica ...... 15755 Mexico and Central America ...... 1965 South A merica ...... 129 163 A frica ...... 80 538 South A sia ...... 87 608 North and Central Asia ...... 211686 South East A sia ...... 19983 Australia ...... 357 330 Europe ...... 50 804 Total ...... 954 832 164 8t~ 0 a .,. Salt affected soils Figure 1. Global distribution of salt affectedsoils. In spite of the fact that the term <(salt-affected soils>> is unambiguous, it covers different soil formations. The soil types belonging to this big family have diverse physical, chemical and biological properties and consequently different degrees of fertility with different possibilities for fertilizer applica- tion. Evidently, they are also different in respect of their potassium content and its behaviour. The only common feature of all salt-affected soils is that they have higher electrolyte content in the soil solution than any of the other soil types and this high electrolyte content influences directly or indirectly all the major soil properties, limiting their fertility. At the same time the concentration of elec- trolytes in the soil solution varies over a wide range and the chemical composi- tion of electrolytes can also be very different in the different groups of salt affected soils. Consequently the effect of electrolytes on soil fertility varies and the plant nutrients, including potassium, behave differently in the differ- ent salt-affected soils. The more or less high electrolyte content results in different physical, chemi- cal and biochemical processes, forms different morphological features and different patterns of nutrient dynamics in soils. 3. Grouping of salt affected soils with particular regard to the potassium status In Table 2 the practical grouping of salt affected soils is presented with their potassium content and the main constraints on its uptake by the crops indi- cated. From the practical point of view there are five groups including all the salt- affected soils of our globe according to our present knowledge. From Table 2 one can see the considerable diversity of different types of salt-affected soils. Such diversity exists for instance in respect of their pH values which may vary between 2 and 12 as shown in Figure 2 (Szabolcs[1989]). In the following the different types of salt-affected soil will be briefly described according to Table 2 with particular regard to their K status and dynamics. pH 1 S 2I 3 I 4 I 5I 6 7 8I 9 I 10i 11I 12i 13 I Saline soils Alkali soils Magnesium soils Gypsiferous soils Acid sulphate soils Figure 2. pH range of different types of salt-affected soils. 166 Table 2. Grouping of salt affected soils Type of salt Electrolyte(s) causing Environment Content of soluble Main constraints of K up- Method for affected soils salinity and/or K compounds take by crops reclamation alkalinity Saline soils Sodium chloride and sul- Arid and semi- High in clay and High osmotic pressure of Removal of excess salt phate (in extreme cases - arid loam soils soil solution (toxic effect) (leaching) nitrate) Low in coarse Na-K antagonism sands Alkali soils Sodium ions capable of Semi-arid High Alkaline pH Lowering or neutraliz- alkaline hydrolysis Semi-humid Effect of water physical ing the high pH by Humid soil properties chemical amendments Na-K antagonism Magnesium Magnesium ions Semi-arid Diverse Toxic effect, high osmotic Chemical amend- soils Semi-humid pressure ments Mg-K antagonism Leaching Gypsiferous Calcium ions Semi-arid Low Acidic pH Alkaline amendments soils (mainly CaSO 4) Arid Toxic effect Ca-K antagonism Acid sulphate Ferric and aluminium Seashores, High Strongly acidic pH, Liming soils ions (mainly sulphates) lagoons with toxic effect leavy, sulphate- containing sedi- ments 3.1 Saline soils This group is characterized by high salt concentration which often exceeds 2% of the weight of the soil at the surface. Saline soils are common in arid and semi-arid regions where agriculture is only possible by the introduction of irrigation and drainage. In consequence of the high salt concentration the soil solution is practically saturated with salts and as a result its high osmotic pressure hinders the life of crops. Only poor natural vegetation exists on these soils represented mainly by halophytes. Besides the high osmotic pressure, toxic effects also occur. Chemically, the dominant types of salts are sodium chloride, and sodium sulphate. In extreme cases even sodium nitrate occurs (Kovda 11980]). The dominance of these salts entails a nearly neutral pH and in consequence of the high salt concentration the soil colloids coagulate resulting in a porous soil structure which makes leaching possible. On saline soils the addition of water (irrigation) and the removal of excess salts (drainage) are the preconditions for agriculture as has been confirmed by the practice of many thousand years of plant production in desert and semi-desert areas (Szabolcs [1986]). As a rule the mechanical composition of such soils can be: (a) saline soil with clay or loam texture with high mobile potassium contents (b) saline soil on blown sands with low mobile potassium content. From these soils leaching removes many sodium salts and as a result the Na-K ratio will shift in favour of potassium. If this process does not go on steadily sodium-potassium antagonism may occur which prevents potassium uptake by the crops. In spite of the low fertility of saline soils vast territories of them are, after reclamation, under irrigated agriculture and valuable crops like cotton, sugar cane and others are grown on them. That is the reason why we must deal with fertilization problems even in the case of high salinity. It is evident that proper drainage and sometimes reclamation methods are the preconditions of fertilization. Generally no potassium deficiency occurs in saline soils except on coarse sands. Still, in case of intensive production when high yields are expected, potassium application is necessary. 3.2 Alkali soils In contrast with saline soils, the pH value of alkali soils is, in most cases high and their salt content is much lower than that of the former type. The alkaline pH is mainly responsible in this case of salt-affected soils for the formation of poor physical and water-holding soil properties. As a result of high pH and comparatively low salt concentration the soil colloids are dispersed, the porosity and structure of the soil are very poor, and the water availability 168 for plants drops, drastically reducing nutrient uptake. Free sodium carbonate, which is toxic for plants, often occurs in alkali soils, and together with their also frequently occurring boron contents (besides the poor physical composi- tion of the soil) chemically prevent their agricultural utilization. Alkali soils are more (cosmopolitan than saline soils and can be found in practically all climatic belts from the equator to above the polar circle. In Europe alkali soils dominate among the salt-affected soils, in which respect it is different from all the other continents where, in the family of salt-affected soils, saline soils prevail. Alkali soils develop mainly on heavy parent materials which can be sub- divided, according to their clay mineral contents, into two groups: (1) Illite vermiculite type, (2) Montmorillonite type. In both groups comparatively high mobile potassium contents can be meas- ured by all the conventional methods and, as a rule, no potassium deficiency can be observed. It is well known that the so-called exchangeable sodium content is high in alkali soils; this is one of the diagnostic features of this group. Besides exchangeable sodium the values of exchangeable magnesium are also often high as mentioned in the description of the next group of salt-affected soils. In spite of their high exchangeable sodium content we must consider some- times a sodium-potassium antagonism in these soils, particularly when crops of high potassium demand are grown or in the cases of intensive production. Another difference between alkali soils and saline soils is that, in conse- quence of the lower salt concentration in the former, crop production is often possible without irrigation and even without chemical reclamation. In such cases the proper selection of fertilizers is essential because evidently, in alkali soils fertilizers of acidic reaction are much more effective than those of alkaline reaction. It is evident from the above that plans for the application of potassium fertilizers to alkali soils can be more diverse than those dealing with their use on saline soils. In this respect the following soil properties should be taken into consideration: (a) the pH value of the alkali soil which can be strongly alkaline but some- times can be neutral or on the surface even slightly acidic; (b) the depth of the so-called oB> horizon with maximum exchangeable so- dium content and maximum compactness; (c) the occurrence and concentration of soluble boron compounds; (d) water relationships of the soil. The main type of alkali soils is Solonetz, which is widespread in many coun- tries of Europe, particularly on the Iberian Peninsula and in the south-east of the continent. Before finishing this short review on saline and alkali soils, a few remarks should be made. 169 Saline and alkali soils together represent the majority of salt-affected soils, and in many places and numerous systems only these two are considered when discussing this family. Much as all this is true, other groups of salt-affected soils must not be neglected in spite of the fact that it is also true that saline and alkali soils have some important common features making them different from the other groups of salt-affected soils. The most important of these features of saline and alkali soils is a conse- quence of the geochemical processes of the continent, namely the behaviour of sodium compounds. It is well known that the rocks forming the mantle of the earth are rich in both sodium and potassium. Each of them account for a few percent of the rocks and total soil matter. The levels of their mobile or soluble compounds are evidently lower by I to 3 orders of magnitude (Fers- man [1934]). The general tendency during the weathering processes is the release of sodium and potassium from minerals, the enrichment of soils and waters with these elements and their erosion toward the ocean. There is a great difference in this respect between the behaviour of sodium and potassium in that soil colloids, even clay minerals, retain the potassium ions more inten- sively than the sodium ions. This is why during leaching the ratio of these two elements shifts, as a rule, in favour of potassium in lattice layers and soil minerals whereas a shift in the opposite direction takes place in soil solu- tions, among absorbed ions and in most of the waters. In consequence of the nature of dynamics of sodium and potassium during soil formation, in spite of the prevalence of sodium in saline soils, during the removal of salts the sodium can be leached out completely while part of the potassium will be retained by minerals and colloids. These considerations are not only theoretical issues, but have practical significance in the evaluation of available potassium content of improved salt-affected soils when elaborat- ing the strategy of fertilizer application. 3.3 Magnesium soils Among all the types of salt-affected soils magnesium soils are the most dis- cussed. Our knowledge and experience of this type are insufficient. What is known is only that several formations of magnesium soils occur and in some of them soluble magnesium salts, in others absorbed magnesium ions, and often clay minerals, rich in magnesium, result in the particular properties of this type of salt-affected soils (Darab[1980). Magnesium soils occur mainly in semi-arid and semi-humid areas and they often occur in association with alkali soils. The main attribute of magnesium soils from the point of view of plant nutrition is the competition of magnesium ions with potassium ions, often even their antagonism. Evidently, in case of high electrolyte concentration, high osmotic pressure and magnesium toxicity may occur, hindering nutrient uptake. 170 Without going into detail, it should be mentioned that a very great diversity exists in the properties and evaluation of magnesium soils where the applica- tion of potassium fertilizers is concerned. 3.4 Gypsiferous soils This group represents yet another insufficiently studied type of salt-affected soils, although it covers large areas in arid and semi-arid regions, practically on all continents, but particularly in North Africa, West of the US, in Central Asia, in the Middle East and other places. The salt content of gypsiferous soils is dominated by calcium sulphate, which constitutes sometimes more than half of the total soil material. It is evident that any agricultural utilization of such soils is practically impossible. However a great part of gypsiferous soils with lower calcium sulphate content have been included in the practice of irrigated agriculture. The main con- straints of nutrient uptake from such soils, with particular regard to potas- sium, are as follows: (a) acidic pH, (b) toxic effect, (c) Ca-K antagonism. The total and mobile calcium content of gypsiferous soils can vary depending on the parent materials of the soils, on their particle size distribution, environ- mental conditions, etc. 3.5 Acid sulphate soils Acid sulphate soils constitute a specific group of salt-affected soils and can be found in tidal marshes, lagoons and sulphuric seashore sediments along the coastlines of all continents from Finland to the Gulf of Guinea from Madagascar to Vietnam. In some countries, e.g. in Thailand and in the south of India, vast teritories are covered by this type of salt-affected soils (Szabolcs [1989]). Due to the nature of acid sulphate soils and in consequence of the high aluminium and iron sulphate content in their upper layer when they emerge from the sea, the pH of these soils can be as low as 2 to 3, sometimes even free sulphuric acid occurs. Evidently in this state they are unfit for growing anything but, after liming they will be very good for the production of many crops, particularly rice. The two dominating minerals in acid sulphate soils are pyrites and jarosite (van Breemen [J973). The latter, as is well-known, is rich in potassium and in consequence, in practice potassium deficiency is very rare in acid sulphate soils because of the release of potassium ions from lattice layers. However the ratio between pyrites and jarosite can be very differ- ent in the different acid sulphate soils, consequently the potassium supply 171 cannot in all cases satisfy the potassium requirement of production especially when high yields are demanded. A few considerations have been reviewed in the foregoing within the broad limits of potassium status of salt affected soils and as an example on correla- tions between pedological and soil chemical aspects. The aim was to make a modest contribution, with particular regard to potassium, to the world-wide efforts used for the solution of contemporary soil fertility problems. 4. References Breemen, N. van: Soil forming processes in acid sulphate soils. International Institution for Land Reclamation and Improvement, Publication 18, Vol. 1. 66-130 (1973) Darab, K.: Magnesium in solonetz soils. International Symposium on salt affected soils, Karnal 92-101 (1980) Fersman, A. .: Geochemistry, Leningrad, (in Russian), 1934 Kovda, VA.: Problems of combating salinization of irrigated soils. UNEP, Nairobi, 1980 Szabolcs, L: Agronomical and ecological impact of irrigation on soil and water salinity. Advances of soil science, Vol. 4, Springer, 189-218 (1986) Szabolcs, L: Salt affected soils. CRC, Boca Raton, Florida, USA, 1989 172 Coordinator's Report on the 2nd Working Session Dr. I. M. Bogdevitch, Byelorussian Institute for Soil Science and Agrochemis- try, Minsk, USSR There were five papers in this session. Prokoshev and Sokolova discussed the potassium supplying power of USSR soils against the background of the results of surveys, long-term field experiments with potassium fertilizers and work in lysimeters. There are great differences in the distribution of potas- sium-bearing minerals, especially illite, within the soil profile. The ratio between illite content of the clay fraction in the upper horizons and that in the lower layers is shown to increase steadily from north to south, from podzol- gley soils to chernozems and solonetzs. Hence, soil potassium status increases towards the south and the effectiveness of potash fertilizers, on the contrary, decreases. The effect of anthropogenic factors (organic manures and mineral fertilizers, liming etc.) on the potash regime in soils often predominates over the effect of natural factors. The increase in potassium fixing capacity of soils and in the potential buffering capacity (PBCK) under the influence of liming is shown. The thermodynamics of potassium behaviour in soils are important in estimating the rate of potash fertilizer needed for high crop yields. Long- term lysimeter trials show that potassium removal depends on soil texture, rates of N, P, K, forms of potash fertilizer, exchange dynamics and migration of chlorides within the soil profile. The dependance of potassium mobility on soil moisture is reviewed by Grimme. This is important as there is often a very poor relationship between soil K content and the response to K fertilizer on a given soil series under varying climatic conditions. In long-term field trials it is found that crop response to K fertilizer is sig- nificantly greater in dry years because soil K availability is reduced. There is a possibility of yield loss due to drying out of the plough-layer and reduced K availability, though the moisture content in the lower layers is still enough for plants. In the drought period the soil must receive the larger part of potash fertilizers. The author gives a detailed description of the mechanism of K movement in soil, diffusion and massflow, laboratory and field methods, and of investigations concerning the soil moisture - nutrient - plant response rela- tionship. Split-root techniques, controlling differential water consumption from subsoil and topsoil, is worthy of attention. The significance of K uptake by plants from subsoil is shown in the paper by De Nobili, Vittori A ntisari and Sequi. There has been little work on this 173 topic. The proportion of K absorbed by plants from subsoil may vary between 10 and 50% . Deeply rooting species such as alfalfa, soybean and cotton pos- sess a high potential for subsoil K exploitation. But under certain conditions, especially lack of moisture in topsoil, even cereals can absorb much K from subsoil. Unfavourable soil characteristics, e.g. presence of a plough pan, high acidity etc. may limit K-uptake severely; that is why subsoil tillage has a favourable effect on the yield of many farm crops. On very acid soils, very high rates of potash fertilizer may have a negative effect. It is convincingly demonstrated that investigation of the physico-chemical characteristics of subsoils is needed in developing a fertilizer policy. Two papers dealt with the behaviour of potassium in saline soils. Feigen- baum, Bar-Tal and Sparks investigated the behaviour of potassium in two saline soils in pot experiments with maize. Applying potassium fertilizer reduced the Na/K ratio in the plant on both soils and at all three levels of salinity but yield was increased only on soil with a low potassium content. Potassium uptake was large in relation to that of divalent cations (Ca+ Mg) regardless of Na concentration. Szabolcs discussed the potash fertilization of salt-affected soils in which the conventional classification is not very useful. He has evolved a practical classification placing such soils in five groups which have differing behaviour as regards their potassium fertilizer requirements. 174 Chairman of the 3rd Session Prof. Dr. A. van Diest, Department of Soil Science and Plant Nutrition, Agricultural University, Dreijenplein 10, 6703 HB Wagenin- gen, Netherlands; member of the Scientific Board of the International Potash Institute 3rd Session Soil and Plant Test Methods and their Calibration for Long-term Sustainabiiity of Soil Fertility 175 The Use of Plant and Soil Analyses to Predict the Potassium Supplying Capacity of Soil A. E. Johnston and K. W T Goulding* Summary Rational K manuring policies are essential for farmers faced with ever increasing variable costs of food production in many parts of the world. Yet long-term soil fertility must not be jeopardised because of short term financial constraints. The need to apply K fertilizer is complicated because many soils can retain K and subsequently release it. Inputs and losses of K from soil are discussed in relation to its transfer between the various analytically defined categories of soil K. Policies for potassium manuring are outlined. For light tex- tured soils with little buffer capacity annual K manuring is suggested to minimise possible leaching losses. For heavier textured soils rotational K manuring based on K balance and with periodic checks on changes in soil exchangeable K is likely to be adequate. The effec- tiveness of such manuring policies can now be checked by crop analysis. Critical K concentrations in field grown crops can now be defined when K concentrations are expressed on a tissue water basis and this is discussed. Various methods for determining water soluble, exchangeable, fixed and matrix K are discussed in relation to either improving our understanding of potassium exchange processes in soil or to the more practical farming need of deciding whether supplementary K as fertilizers or manures should be applied to improve growth. For many soils, the determination of exchangeable K by extraction with ammonium salts is still the quickest and most reliable way to predict the need for sup- plementary K. 1. Introduction The fertility of a soil can be defined as its capacity to produce the crops desired by the cultivator within the constraints of local climate. Such a definition encompasses various factors which control plant growth: i) soil structure which affects not only the workability of soil and its capacity to provide an- chorage for roots but also suitable air/water relationships for active root growth and function; ii) the incidence of pests, diseases and weeds; iii) the availability of nutrients. In well developed systems of food production many of the constraints to yield can be wholly or partly controlled. The fertilizer industry has developed to supply farmers with plant nutrients so that they can produce food to feed * Dr. A. E. Johnston, Lawes Trust Senior Fellow and Dr. K. W T Goulding, Soils and Agronomy Department, AFRC Institute of Arable Crops Research, Rothamsted Experimental Station, Harpenden, Herts AL5 2JQ/United Kingdom 177 an ever increasing population. As yields per hectare of agricultural crops in- crease and more nitrogen (N) is used the need for potassium (K) increases and attention to the role of potassium in crop quality and in maintaining soil fertility becomes ever more important. Whilst the application of potassium to agricultural soils rarely leads to en- vironmental problems, improving the cost effectiveness of its use depends on our ability to correctly identify those situations where potassium is needed. This paper summarizes some current ideas about the usefulness of crop and soil analysis for estimating the potassium supplying power of soils, and seeks to develop rational K manuring policies to achieve optimum yield and the maintenance of soil fertility. 2. The potassium cycle in agricultural soils Figure I shows a simplified potassium cycle for agricultural soils. Following the early work of Hoagland and Martin [1933] soil K is divided into four categories. Their precise definition is less important than the realisation that K can transfer in both directions from one category to another. The rate at which this transfer occurs and the direction of movement (i.e. the balance of the equilibrium) has important implications for the usefulness of soil ana- lysis in predicting soil K status and crop requirement for potassium. Rain Crop Manures TV Mineral Solution Exchangeable Fixed Matrix K K NativeK Drainage Figure I. A diagramatic representation of the potassium cycle for agricultural soils. 2.1 Inputs Inputs in rain are usually as soil dust. In Britain they are often less than 5 kg K/ha annually and have little effect on the total K balance. Inorganic fertilizers and organic manures invariably supply K in water- soluble forms and are therefore interchangeable as sources of K when the farmer plans the manuring for each field. 178 2.2 Losses in drainage A not insignificant proportion of the exchangeable K in soil is water soluble and at risk to loss by leaching. Losses are generally small and K poses no risk to health. Recent annual mean concentrations of K in water draining from a sandy loam soil were 3.4 mg/litre and from a sandy clay loam 1.3 mg/litre under arable and 0.5 mg/litre under herbage crops. Approximate annual leaching losses ranged from 0.2 to 5.0 kg K/ha from the clay loam and 3 to 20 kg K/ha from the sandy loam depending on cropping, rainfall and evapotranspiration. The amount and type of clay in both the surface and subsoil affects the retention of K within the soil profile (Johnston [1986]). Potassium retained in the subsoil is available to deep rooted crops like sugar beet, winter wheat and lucerne. Benefits from subsoil enrichment with both K and P have been shown (Johnston& McEwen [19841). Kuhlman et al. [19851 discussed methods for determining the amount of K taken up by crops from subsoils. Rarely, however, is the readily extractable K content of subsoils determined. On many soils there is the difficulty of getting a representative sample and the additional cost of sampling and analysis to improve the fertilizer recom- mendation is probably not justified. However, poor correlations between crop response to K fertilizer and soil analysis values for surface soils could be ex- plained in some cases by K uptake from the subsoil (see Section 6.1). Sugar beet, a deep rooted crop, invariably removes more K than other crops from soils which have been without K manuring for many years. In one experiment at Rothamsted on a silty clay loam with 80 mg/kg exchangeable K, cereals, potatoes and sugar beet contained 40, 30 and 120 kg K/ha in the harvested produce. Mengel [1989/ discussed the efficiency with which different crop species use potassium and this is a topic of considerable importance. When various crop species are compared in the field it will be necessary to distinguish be- tween effects due to biochemical differences and those due to soil variation and crop rooting patterns in the soil profile. 2.3 Effect of K balance on exchangeable K in soil The K balance can be defined as the difference between the amount of K applied to a crop and that removed from the field in the harvested produce. A positive balance enhances soil fertility, a negative balance depletes soil K reserves. Johnston [1986] showed how positive and negative K balances in long-term experiments related to increases and decreases in exchangeable K. Some examples are in Table 1. On soils which were cropped for many years without K manuring, large total negative K balances, 910 and 1350 kg/ha, caused little change in exchangeable K, only 8 and 6% of the K balance respec- tively. Much of the K had come from initially non-exchangeable reserves over the long period of cropping. Large positive K balances exceeding 3000 kg/ha 179 increased exchangeable K by less than 450; much of the excess K had become non-exchangeable. Such results might not be typical of the lightest textured sandy soils low in organic matter because they have little buffer capacity. However the results help explain why exchangeable K in many soils changes little over a number of years, a fact which often puzzles many farmers. Table I. Changes in exchangeable K related to positive and negative K balances when barley, clover and grass were grown continuously at Rothamsted Crop Period Treatment K balance Change in Change in exch. K (kg/ha) exch, K as a To of K (kg/ha) balance Barley 1856-1903 None - 530 + 10 - K +3760 +1100 29 1903-1974 None - 910 - 80 8 (K)' - 1840 - 1040 56 Clover 1956-1966 K + 620 + 260 42 1968-1978 K + 1670 + 690 41 1979-1983 K - 1490 - 560 38 Grass 1965-19762 None - 500 - 80 16 (K) - 3460 - 1470 42 1970-1980' None - 1350 - 80 6 (K) -2110 - 420 20 K residues from applications between 1856 and 1903 2 None: no K since 1856, (K): K residues from applications between 1856- 1964 None: no K since 1899, (K): K residues from applications between 1899- 1969 A valuable laboratory screening technique for the assessment of the K fixing capacity of soil was discussed by Johnston [19861. His results showed the im- portance of soil pH and past manuring on the amount of added K which remained exchangeable in soil after a cycle of wetting and drying episodes to simulate conditions in the field. (See also York et aL [1953]; Karim & Malek [/19571.) 3. Potassium in plants 3.1 Amounts of potassium in crops At harvest many crops contain similar amounts of potassium and nitrogen (Table 2). Although Table 2 shows that there is less K than N in grain plus straw of cereals at harvest, these crops often contain more K than N at anthe- sis. For example, between 1968 and 1978 winter wheat crops, given 144 kg N/ha and grown after a two-year break from cereals on Broadbalk at Rothamsted, yielded on average 5.84 t grain/ha each year. At anthesis the 180 above ground part of the crop contained 132 kg N and 203 kg K/ha whilst the grain plus straw at harvest contained 145 kg N and 135 kg K/ha. Table 2 also shows that if the straw of cereals or the tops of sugar beet are incorporated into soil then the amount of K removed from the field will be considerably less than that in the crop at harvest. Most of the K in plants and plant residues is water soluble and readily available to succeeding crops. Table 2. Yields (t/ha), and nitrogen and potassium contents, N and K (kg/ha), of some important arable crops at harvest and for comparison grass/clover leys Winter wheat Winter barley Spring barley grain straw total grain straw total grain straw total Yield 6.70 7.92 8.48 6.85 5.90 5.37 N 114 47 161 170 33 203 81 25 106 K 29 89 118 39 135 174 24 54 78 Potatoes Kale Grass clover ley tubers (dry matter) Yield 48.0 75.0 11.7 N 161 211 231 K 252 246 290 Field beans Oilseed rape grain straw total seed straw total Yield 3.01 3.51 2.80 6.67 N 124 31 155 90 56 146 K 32 56 88 26 103 129 Sugar beet Carrots tops roots total tops roots total Yield 46.5 44.7 27.1 64.5 N 155 76 231 72 129 201 K 272 78 350 141 227 368 Thus it is important for farmers to ensure that in relation to other inputs, the maximum amount of K required by the crop will be readily available to it in the soil. The amount in fertilizers and manures should augment that which is supplied by the soil. In deciding on the amount of K to apply both the K status of the soil and the K balance must be considered. These concepts are discussed in Section 7 in relation to developing manuring policies. 3.2 Role of potassium in plants Potassium has two main functions in plants: i) it has a vital and irreplaceable role in certain metabolic processes including protein synthesis and the translo- 181 cation of the products of photosynthesis; ii) it appears to be the preferred cation for the generation of osmotic pressure to maintain cell turgor. Much larger quantities of K are needed for the second role than the first which ex- plains why plants take up so much K. A cellular explanation for critical K concentrations in plants was recently summarized by Leigh.[19891 based on a more detailed discussion by Leigh & Wyn Jones [1984 and the references therein]. 3.3 Critical levels of K in plants It has been accepted piactice in agronomic studies to express the concentration of K in plants as a % in dry matter. This is analytically convenient and if dry matter yield is known K uptakes and balances can be calculated. However, attempting to relate % K in dry matter to crop response to either soil or fer- tilizer K invariably meets with little success for arable crops, because 07o K in dry matter declines appreciably during growth. For example, Figure 2a shows 07oK in dry matter of two crops of spring barley (Hordeum vulgare cv. Georgie) both grown on soil well supplied with K (330 mg/kg exchangeable K) but one given 96 kg N/ha, the other no N. Lack of nitrogen rather than K affected % K in spring barley early in the growing season and grain yield at harvest was 4.8 and 1.6 t/ha with and without N respectively. Because most of the potassium in plants is used to generate osmotic pres- sure, which is a function related to the aqueous concentration of K within cells, Leigh & Johnston [1983a & b/examined whether expressing K concentra- tions on a tissue water basis had any benefits over % K in dry matter. Figure 2b shows the K concentrations for the crops in Figure 2a recalculated on the basis of tissue water. When expressed in this way both crops maintained the same concentration of K throughout the growing season until the onset of water loss as ripening commenced. The K concentrations were not affected by nitrogen manuring. Later studies (Leigh & Johnston [1983b]) showed that neither drought nor phosphorus nutrition affected K concentrations in tissue water, but they did affect % K in dry matter. Only when barley was grown on soils with different levels of exchangeable K (325 and 55 mg. K/kg) were the levels of K in the tissue water different (Figure 2c), about 200 mM and 50 mM in crops grown on K sufficient and K deficient soils respectively. Both crops received 144 kg N/ha and yields on the low and high K soil were 2.6 and 4.8.t grain/ha respectively. There was no further increase in tissue K concentration in barley grown on soils contain- ing even higher levels of exchangeable K (Table 3) (Leigh [19891). On this very K deficient soil 07o K in dry matter was appreciably less than that in the crop grown on the K enriched soil, but for both crops the concentration declined throughout growth. Thus it would be difficult to use T K in dry matter to diagnose K deficiency unless the optimum concentration could be defined accurately for each growth stage for barley grown on a range of soils in different seasons. 182 a 8 b 600 a 144 kg N/ha A144kgNha 0 0 kg N/ha 500 00 kg N/ha af 6 03 400 4 300- 0-0 r 200 1DO0 60 80 100 120 140 160 180 60 80 100 120 140 160 180 Days after sowing Days after sowing C 600 A 90 kg K/ha ~500 A 0 noK since 1852 400 4 0 200 00 60 80 100 120 140 160 180 Days after sowing Figure 2. Concentration of K in field-grown spring barley (total above ground plant) from emergence to just prior to harvest. a. percentage K in dry matter, crop given 0 or 144 kg N/ha b. K in tissue water, crop given 0 or 144 kg N/ha c. K in tissue water, crop given 144 kg N/ha and grown on soil given no K since 1852 or 90 kg K/ha each year. Table 3. The effect of soil exchangeable K on the concentration of K in tissue water of spring barley grown on Hoosfield, Rothamsted in 1981 Soil exchangeable K K in tissue water (mg/kg) (mM/kg) 55 57±4* 325 206 ± 7 827 219+9 Mean ± s.e. over 75 days 183 Similar values for K in tissue water to those in barley, about 200 mM, have been found for well fertilized grass, but oilseed rape (Brassica napus) and field beans (Vicia faba) apparently maintain K concentrations only about 100mM under conditions of ample K supply (Leigh [19891). 3.4 Use of plant analysis in diagnosing K deficiency The results above show clearly that barley, grass, oilseed rape and beans can be sampled at any time throughout growth during spring and summer and K deficiency can be readily diagnosed if the K concentration is expressed on a tissue water basis. Oilseed rape and beans apparently maintained lower K concentrations in tissue water than did cereals and grass when grown on soils on which there was no response to K fertilizer. Similar values for other crops must be determined experimentally under field conditions. We now have a method of expressing K concentrations in crops which allows us to tell whether the crop is K sufficient or K deficient. Although the method is excellent for diagnosing K deficiency which will allow corrective K applications for future crops, it is usually too late in the growing season to apply K to the crop in which deficiency is observed. 4. Soil analysis 4.1 Early work In the early 1870s von Liebig [1872] used dilute acetic acid to determine both readily soluble K and P in the 0-23 cm depth of soil from five plots on the Broadbalk experiment at Rothamsted. He readily differentiated between soils which had or had not received K fertilizer for 20 years. During the next few years several workers used various dilute acids as extractants. Dehdrain [1891] used dilute acetic acid to analyse the soils from the long-term plots at Grignon near Paris. Dyer [1894, 1901, 1902] used soils from both Broadbalk and Hoos- field at Rothamsted to check the usefulness of 1076 citric acid as an extractant for K and P. 107o citric acid was chosen because the acidity of the solution approximated to that of plant sap, and it was widely held at that time that roots exuded solutions which dissolved soil constituents thus making plant nutrients available. For potassium, Dyer compared 107o citric acid with constant boiling HCI for 48 hours and with dissolution in HE He showed that it was only with citric acid that there was a sufficiently wide range of values, 8 or 10:1 between K manured and unmanured soils, to allow it to be used diagnostically. Dyer thus highlighted an important criterion for soil analytical reagents intended to estimate «plant available> nutrients: there must be a wide range of values between very responsive and non-responsive soils so that soils of intermediate response can be readily identified. A second criterion is that the reagent 184 should not change its strength during the period of extraction. This invariably rules out the use of acid reagents on calcareous soils. Dyer showed that the 48 hour boiling HCI extractant failed to distinguish between soils with and without K manuring for 50 years because it extracted too much K. For many clayey and loamy soils total K content rarely differenti- ates between soils with different K manuring even after this has continued for more than 100 years (Johnston [1986]). However, for sandy soils from the tropics total K is a good indicator of past K manuring (Johnston, unpub- lished). 4.2 Exchangeable K The idea that the exchangeable bases in soil are the source of cations for plants was put forward by Knop [1871], who used a solution of ammonium chloride to extract them. Prianishnikov[1913] was probably the first to suggest the use of neutral ammonium acetate (2 M). An increasing awareness of the im- portance of base exchange reactions in chemical and physical properties of soil led to many investigations of analytical methods in the 1920s and 1930s. The usefulness of any analytical method for assessing soil fertility must be tested in field experiments. Schollenberger and Dreibelbis[1930aJ related K response by crops in the field to exchangeable K in soil estimated by extrac- tion with ammonium acetate, as in the method outlined by Schollenberger and Dreibelbis [1930b]. Today the use of I M ammonium acetate or I M am- monium nitrate has been widely adopted; ammonium nitrate has analytical convenience where the acetate ion interferes with the determination of K in some analytical instruments. Both reagents are often used for exchangeable cations in those soils where the Olsen bicarbonate method (Olsen et al. [19591) is good for determining readily soluble P. Many other reagents have been suggested and a number are widely used, e.g. ammonium acetate and acetic acid at pH 4.8 (Morgan [1935]), calcium lactate (Egner and Riehm [1955]), ammonium acetate-lactate (AL) (Egner, Riehm and Domingo [19601) and calcium acetate-lactate (CAL). A number of these reagents have the advantage that they extract both K and P in amounts which can be related to crop responses to these nutrients. In general the amounts of K extracted by such reagents are strongly correlated with those extracted by ammonium acetate. 4.3 Fixed K Many research workers consider that the rate and amount of K transferring between exchangeable and fixed or non-exchangeable categories (Figure 1) appreciably affects the response of crops to K fertilizer. Much effort has been expended in attempting to find methods to determine the quantity of non- exchangeable K in soil. Martin and Sparks [1985] gave a comprehensive list 185 and discussed sixteen of the methods. As yet there is no rapid, reliable method. However, it is necessary to clearly differentiate between their usefulness and applicability in predicting the availability of soil potassium to crops and their use in assisting our understanding, at a fundamental level, of K exchange processes in soil. 4.3.1 Mineralogical analysis Arnold [1962] showed that the ability of many British soils to release K is correlated with the K content of the fine clay (<0.1 gm) and the percentage of fine clay in soil. Increasing amounts of fine clay could make up for decreas- ing K concentrations in the fine clay and vice versa. There was no similar correlation for the coarse clay (0.3-2.0 pm) fraction, and Arnold concluded that neither its amount nor T K were important in determining the K releasing power of the soils he studied. In these soils the clays were mainly micas and hydrous micas (illites) so in general the smaller an illite particle the faster it releases K. Unfortunately, estimates of the illite content of soil, based on the intensity of the 10 A peak of the <2 pm clay cannot be used to predict K releasing capacity, because the micas and hydrous micas in the coarse clay dominate the spectrogram. Arnold also found that soils with a long history of K manuring gave anomalous results. 4.3.2 Quantity/potential and quantity/intensity measurements The quantity/potential relationship (Woodruff 1955; Barrow et al. [19651) relates change in the exchangeable K content of a soil to K potential defined as the free energy associated with replacing one equivalent of K by one equiva- lent of Ca (Schofield [1947]; Woodruff [1955]; Arnold [1962]). K potential, AG=RT In aK/a'/Aca+Mg From the same measurements the relationship between change in exchange- able K and activity ratio aK/al/ca+Mg can also be drawn. This is usually called the quantity/intensity (Q/I) curve (Schofield & Taylor [19551; Beckett [1964]). The analytical procedures are time consuming but not difficult. For exam- ple, Addiscott [1970a-cJ shook soils for I h with 0.01 M CaC12 containing KCI from 0.00025 to 0.006 M (according to exchangeable K content) at a 1: 10 soil to solution ratio. Usually six to eight K concentrations in solution were used for each soil. Soils were also shaken with 0.01 M CaCI2 only, at soil:solu- tion ratios 1:10 to 1:250. The suspensions were centrifuged and the K, Ca and Mg determined in the supernatant liquid. Potassium potentials and activity ratios were calculated thus: K potential, AG=2.303 RT logio AR where AR=activity ratio (intensity) 186 AR = aK = CK f. + V aa + M, C'ACa. M,.t f where fU and f are the activity coefficients of the monovalent and divalent 4 ions. f /f " was taken as 1.18 after Beckett [19651, who found that the value varied little from this within the range of K and (Ca+Mg) concentra- tions in soil. The change in exchangeable K content of the soil (± AK) was calculated from the gain or loss of K by the solution. Equilibrium K potentials (A~o) and equilibrium activity ratios (ARo) at which the soils neither gained nor lost K were interpolated from the quantity/potential and quantity/ intensity relationships. Both relationships are independent of the plant. The quantity/potential curve is often more useful in relating to measurements of K removed from the soil because the (logarithmic) potential scale is larger in the region of K removal. The quantity/potential relationship also gives a means of measuring K initially available to different plant species (Arnold [1962]; Addiscot [1970b, c]). The Q/Irelationship has been used more widely than the quantity/potential relationship (e.g. Nair & Grimme [19791; Sparks & Liebhardt [1981]; Evan- gelou et al. [1986]). The curves have been used to estimate the extent to which available K is buffered by fixed K. Opinions are still divided about the useful- ness of Q/I curves especially for practical advisory purposes (see e.g. Sparks & Huang [19851 and Bertsch & Thomas [1985]). Addiscott [1970c] measured QII curves in soils with contrasted cropping and manuring over many years, taken from long-term experiments on differ- ent soil series at Rothamsted and Woburn. Within each experiment, Q/I curves for differently treated soils could be superimposed by appropriate horizontal and vertical shifts. The vertical shifts, on the Q axis, needed to bring the curves into coincidence were equal to the differences in exchangeable K. The QI curves from different experiments on the same phase of one soil series could also be superimposed but not those from experiments on different phases of the same soil series. There were marked differences in the shapes of the Q/I curves from different soil series. Figure 3 (adapted from Addiscott [1970c]) shows the Q/I curves for soils from contrasted treatments on the Barnfield experiment at Rothamsted. Curves for treatments 5/0 and 7/0 are different, although neither has received K fertilizer since 1903. Before then treatment 5/0 had been without K manur- ing since 1843, whereas treatment 7/0 had K from 1843 to 1902, and this was replaced by Na and Mg from 1903. Although both soils had been without K manuring for some 60 years when sampled for these analyses the difference in K treatment in the previous sixty years, 1843-1902, was still detectable in the exchangeable K contents, 350 and 170 mg K/kg for 7/0 and 5/0 respec- tively, in the shape of the Q/Icurves (Figure 3) and in the ARo values 6.4X 103 and 2.4x 10' (m/I)'/ respectively. The persistent effects of K manuring be- tween 1843 and 1902 were partly the result of not applying nitrogen to the crops grown between 1902 and 1967 so that the soil was not stressed to supply K. 187 1.6- 1/0 1.2 08 5/>07 o 02/ 8 -16 24 40 48 --- 56 64 72 80 0- . 1.6- " 2,4-/ 28- Figure 3. Effect of contrasted long-continued treatments with K on the QiI curve for a silty clay loam soil from Barnfield, Rothamsted. No nitrogen was applied to fertilizer only plots. Key: 5/0 P only since 1843; 7/0 PK 1843-1902, P Na Mg since 1903; 1/0 P only 1845-1852, FYM since 1856; 4/0 PK Na Mg since 1843 (except 1861-70 no K); 2/0 P only 1845-1852, FYM plus P 1856-1902, FYM plus PK since 1903. Annual rates P 33 kg/ha, K 230 kg/ha, Na 88 kg/ha, Mg 22 kg/ha, FYM 35 t/ha containing on average 210 kg K/ha. 4.3.3 Buffer capacity The buffer capacity, dQ/d, has been variously defined. Beckett et al. [19661 and Beckett & Nafady [1967] measured the potential buffering capacity of the linear section of the Q/I curve at large activity ratios. Addiscott [1970c] found that for many of the soils he analysed the Q/I curves were often not linear. Following Talibudeen & Dey [1968], Addiscott took the buffering ca- pacity as the slope of the tangent to the Q/I curve when AK=O, i.e. (dQ/dI)aK=o. Buffer capacity, whichever way it is calculated, is a measure of the ability of a soil to maintain the concentration of K in solution. Some have found this a useful concept (Beegle and Baker [1987]) but others have found it does not relate well to nutrient uptake (KtchI [1987]; Novozarnsky and Houba [1987]) nor is it suitable for routine analysis (Villemin [1987]). This last point seems to be the chief reason for the very limited use of the QiI concept in soil analysis. Although it gives more information than that given by simple extractants, the extra work required often outweighs the use- fulness of the additional information. This is true in general of all methods involving ion exchange equations. Whilst they have increased our understand- ing of cation exchange processes they are unlikely to provide a rapid soil test for K availability. 4.3.4 Strong acids The extraction of fixed K from soil using boiling HNO 3 (Haylock [1956]; MacLean [1961]) or HCI has been widely used. Attempts have been made to 188 relate the results to categories of soil K called ((available , ostep> and ((cons- tant rate>) K (JiaXian and Jackson [19851; Sailakshmiswariet al. [1985]) but the amounts of K extracted often relate poorly to the response of annual field- grown crops to K fertilization. However, in exhaustion studies in pots in the glasshouse the amounts of K taken up in the latter stages of growth often relate quite well to K extracted by strong acids, presumably because initially non-exchangeable K is by then the main source of K supply to the crop. Doll & Lucas [19731 considered strong acids to be useful for research purposes but not for practical advisory work. 4.3.5 Electro-ultra filtration (EUF) The EUF methodology and its application have been described by Nemeth [1982], [1985. Cumulative K in repeated water extracts is plotted against time and certain fractions correlated with K availability as measured in the field and laboratory. A number of research workers (e.g. Sinclair [1982]) found little advantage over other, less expensive procedures. Table 4 shows that for a number of soils with contrasted manuring history from long-term experi- ments at Rothamsted the total amount of K extracted by the EUF procedure was about the same as that extracted by ammonium acetate. Thus, the EUF procedure predicted K reserves in these soils no better than did exchangeable K. However, current EUF machines extract rather more K using higher tem- peratures and voltages and some organisations have adopted EUF multi- analysis techniques. Table 4. Comparison of the EUF-K and exchangeable K (mg/kg) in soils from long-term experiments at Rothamsted EUF-K' Treatment Experiment I I1 1+l Exchangeable K None ...... Barnfield 78 28 106 144 Broadbalk 62 21 83 100 PK ...... Barnfield 475 134 609 680 Broadbalk 246 76 322 356 NPK ...... Broadbalk 2 170 74 244 260 FYM ...... Barnfield 471 210 681 802 Broadbalk 559 160 719 686 FYM+PK ...... Barnfield 661 157 818 1086 ' EUF 1 0-30 min, 200V, 20°C; 1130-35 min, 400V, 80"C 2 FYM farmyard manure 4.3.6 Sodium tetraphenylboron (NaTPB) When the Na in NaTPB is exchanged by K the potassium salt precipitates out. NaTPB has been used in research especially on separated clays, less widely on total soils (Qudmener [1979], [1986]). The salt itself is costly and though the analytical procedures are not difficult, the potassium salt must be dis- 189 solved from the soil residue before K can be determined. Both factors may explain why the method is not widely used in advisory work even though the amounts of K extracted often correlate well with K uptake by grass in pots. Jackson [1985] has recently proposed a scheme for use in the analysis of pasture soils in New Zealand: the current rapid advisory test, based on a short extraction with ammonium acetate, is used to screen out soils that are either unresponsive or very probably responsive to added K, and the NaTPB test is then used as a supplementary screen for those soils on which there was doubt about the need to apply K fertilizer immediately. The principle of this approach is similar to that of Arnold [1962], who showed that K potential is a better predictor of K release to ryegrass from soils containing between 100 and 200 mg/kg exchangeable K than is exchangeable K itself. 4.3.7 Ion exchange resins Ion exchange resins have been used for many years to extract K and other ions from soil (see Qudmener [1979]). Early workers used only single extrac- tions and often the results were little better than those with salts such as ammo- nium acetate (Arnold[1958;Haagsma& Miller[1963]). More recently multi- ple extractions of the same soil sample with Ca-resins, which act as an infinite sink for K diffusing into the solution, have allowed cumulative K release curves to be constructed (Talibudeen et al [19781; Havlin & Westfall [1985]). The curves, in the form EK: t /, often have the same shape and magnitude as those obtained using cumulative K uptakes by ryegrass grown in pots in the glasshouse (c.f. Figure 4 and Figure 5). The curves invariably differentiate soils with different clay content, clay mineralogy, manuring and cropping history. 1500 Sooo. Denchworth 1000 (mg kg - 1) Mrl 500 ". - M , 0 15 30 45 60 t0 (hoursY) Figure 4. Relationship between the cumulative K release to Ca resin and time for two soils with differing clay content, Denchworth Series 49% clay, Newport Series 8%7oclay. 190 300 Ko grass soils 4 2 6 2003 5 0 E 3- -v 4 02 00 5 1 0 c (66 , E700 Koaal/alo4ol 033 010 0 5605 20200 5 4 sol 22 4f 5nwan 6rppnmaurn 4i1 6 7cut Manuring once every four years during 1848-1951: plots 5 and 6 unmanured, plots 3 and 4 PK Na Mg, plots I and 2 NPK Na Mg. Cropping plots 2 4 6 turnips, barley, clover, wheat plots 1 3 5 turnips, barley, fallow, wheat. From 1958-67 no K was applied and half of each plot grew grass the other half arable crops or fallow. On the arable/fallow soils the greater uptake of K on plots 6 and 4 compared to 5 and 3 respectively is related to the larger clay content of these soils. 191 Like K uptake curves, resin curves have been interpreted as separating two, three or four main categories of soil K depending on clay content (e.g. Figure 4 and Goulding [1984]). This observation can be tested by fitting both a linear spline and the best smooth curve to each set of data using the Rothamsted Maximum Likelihood Program (Ross [19801). The linear spline almost always has the least residual mean square. From the EK: t curves amounts (M) and rates (R) of K release can be calculated. Amounts are obtained by extrapolat- ing each linear segment back to t " =0 and taking the difference between the intercepts, whilst rates of release are the slopes of each segment. Figure 4 shows that a sandy soil (80 clay, Newport Series) had a two-part curve, whereas a much heavier textured soil (49% clay, Denchworth Series) had a four-part curve. Unfortunately a full analysis of a soil can take 6 to 9 months and this and the complexity of the procedure makes it unsuitable for routine analysis. As a research tool the method has much to commend it; it is less expensive than pot experiments for determining the K releasing capacity of soil. Goulding and Loveland [1986] have suggested that the technique could be used to map K reserves in soils with different clay mineralogies and clay contents, especially soils without a history of long continued K manuring. The technique has been used by Goulding & Stevens [1988] to measure K reserves in a soil under forestry as affected by felling practice and the ability of the reserves to meet the nutrient requirement of the tree crop. Goulding & Johnston[unpublished] have attempted to produce curves simi- lar to Ca-resin curves but much more quickly by sequential extraction of soils with 0.5 M- or I M-HCI during a period of 10 hours. Unfortunately the total amount of K extracted was often little different from the exchangeable K, but sometimes linear splines fitted the data better than smooth curves, sug- gesting that the method separated different fractions within the total ex- changeable K; it also proved to be a better predictor of K and Mg uptake and yield in pot and field experiments. Other acids and longer extraction periods may remove more fixed K and prove better predictors of nutrient avail- ability. 4.4 Water soluble and matrix K Potassium ions (K ) , present in the soil solution - or more accurately their activity - and potassium as a structural element in soil minerals - matrix or mineral K in Figure 1 - represent the extremes in plant availability and are rarely measured in soils. 4.4.1 Matrix K Both the rate and amount of matrix K released to plants depend on the quan- tity of clay, especially the smaller clay particles, and its mineralogy. For example, Johnston [1986 and the references therein] showed that only small amounts of K are now taken up each year by cereals grown on soils which 192 have had no K additions for 80 or more years. On a sandy loam (10% clay) at Woburn, the amount is only 10 kg K/ha; on a silty clay loam (20% clay) at Rothamsted, 20-35 kg/ha; on a sandy clay loam (25% clay) at Saxmund- ham, 35 kg/ha. The clay fractions of all three soils are mainly interstratified expanding minerals with some mica and kaolin and the different amounts of K released relate principally to the amount of <2 pm clay in each soil. For most cultivated soils the release of matrix K is of little practical conse- quence, because such soils have received applications of K in manures and fertilizers; reserves of exchangeable and fixed K rather than matrix K dominate the amount of soil K available to crops each year (Goulding [1984]; Goulding & Loveland [1986). 4.4.2 Water soluble K Mengel and Kirkby [1982] pointed out that K + concentration in the soil solu- tion largely controls the rate of K diffusion towards roots and therefore the uptake of K by plants. But amounts of K in solution are too small to supply crop needs. Warren and Johnston[1962] showed a strong relationship between water soluble and exchangeable K; about 15% of the exchangeable K above 170 mg K/kg was water soluble. Johnston [1986] suggested that it is worth- while considering relating variation in the response of annual crops to K fer- tilizer on different soils to variation in the amounts of water soluble K. A simple measurement of water soluble K can, however, give no indication of the rate of replenishment. 4.5 Comparison of methods of soil analysis Many comparisons have been made between different methods of soil analysis for K. Johnston & Addiscott [19711 used 52 soils varying in pH (water) from 5.4 to 8.0, with exchangeable K contents from 50 to 1100 mg/kg and with a range of organic matter contents, in an exhaustive cropping experiment with ryegrass in the glasshouse. Soil textures were silty clay loams and sandy loams. The soils were analysed for exchangeable K, Ke; equilibrium activity ratio, ARo; equilibrium K potential, A(o; K buffer capacity, BCo, (see Section 4.3.3) and the K removed from soil before the potential fell to -5600 cal/ equivalent, the uptake potential of ryegrass, K5600 . Nine harvests of ryegrass were taken and the experiment lasted 608 days. Regressions on the quantity measurements, Ke and K5600 , accounted for more variation in K uptakes in either the first three, or all nine cuts, than did regressions on the other K meas- urements. There was little to choose between Ke and K560o because they were strongly correlated (r=0.989). The good relationship between K uptake and Ke suggested that differences between soils in continuous grassland or arable cropping or ley/arable cropping, or with and without K fertilizers or farmyard manure, were attributable solely to differences in the quantity of exchangeable 193 K, not its rate of release. Currently it seems that the K status of soils can be classified as well by exchangeable K as by any other rapid analytical procedure. 5. Glasshouse experiments Pot experiments have been widely used to measure the K releasing capacity of soils and to relate the data to laboratory soil tests. Pot experiments have also been used to help understand the processes governing the uptake of fixed K. Qudmener [1979] reviewed techniques for pot experiments. They are expen- sive and can last many months, possibly years, even when only small amounts (e.g. 400 g) of soil are used per pot. Often K taken up by the test crop during the first 8 to 10 months estimates that which is exchangeable and fixed (Figure 1); the period of cropping need only be extended if the release of matrix K is being studied. 5.1 Comparison of glasshouse and field experiments Pot experiments are attractive because a direct simple relationship between K uptake by plants in pots and in the field might well be expected. This rarely happens for a number of reasons. Pot experiments are often conducted in the glasshouse with supplementary lighting and heating so that plants con- tinue to grow throughout the year and there is a constant demand for K. Field- grown plants in temperate climates rarely grow continuously and when there is no plant demand for K non-exchangeable K can diffuse to exchange sites (see Johnston [19861). The roots of plants grown in pots invariably exploit the soil more thoroughly whilst those of field-grown crops can explore deeper soil horizons than those usually sampled for pot experiments. When K uptake from the same soil was compared in the glasshouse and the field, ryegrass in pots took up as much K in 3.5 years as did the field-grown crop in 9 years (Johnston & Mitchell [1974]). Moberg and Nelson [1983] found that several years of exhaustive cropping in the glasshouse was the same as 60 years cropping in the field. Thus whilst pot experiments may appear to be a useful model for K release in the field, the relationship between K release data from pot and field experiments varies with many factors and the relationship is often poor (see Ogunkunle and Beckett [19881). 5.2 Understanding the processes of K release Addiscolt and Johnston [1974/ discussed reasons for examining the relation- ship between cumulative K uptake and ltime, t'A, in glasshouse experiments. Figure 5 gives examples taken from Johnston & Mitchell [1974]. They grew 194 ryegrass in pots in the glasshouse for two cropping periods, the first of 540days, the second 734 days, and in each period seven harvests (cuts) of grass were taken. Between the first and second periods the soils were air dried before being resown with grass. They used 48 soils taken from a field experi- ment at Rothamsted where the cropping and manuring histories were known since 1848. For soils which were in an arable/fallow rotation between 1958-1967, the amount of K taken up by the first cut of grass was larger than the decrease in exchangeable K during the first cropping period. This suggests that K taken up by the second to seventh cuts came from initially non-exchangeable sources. The EK: t /v relationship for cuts 2 to 4 was almost linear and much steeper than the linear relationship for cuts 5 to 7. Such linear relationships suggest that a diffusion process most probably determined the rate of K uptake and that K diffused either from a larger «pool or at a much faster rate for cuts 2 to 4 than for cuts 5 to 7. The amount of K taken up during the first cropping period from the ara- ble/fallow soils I to 4 (Figure 5) reflected the build-up of K residues during the period 1848-1958. The much smaller amounts of K taken up from «grass soils I to 4 reflects the fact that during 1958-1966 those soils in the field grew grass which obviously removed much of the readily available K residues. Air drying the soil between the first and second cropping periods hastened the release of some K, which was removed by the first and second cuts of grass in the second cropping period. For all soils the amounts of K taken up by cuts 3 to 7 were similar, suggesting release from the same source. 5.3 Cumulative K uptake related to initially exchangeable K Using other data from the experiment described above, Johnston & Mitchell [19741 also showed that K uptake by ryegrass in the first cropping period and the exchangeable K in the soils at the start of the experiment were strongly correlated (r= + 0.94) (Figure 6). The amount of non-exchangeable K released from each soil was about twice the decrease in exchangeable K during the first cropping period. Weaker correlations are often found when soils from a wide range of parent materials are included in the same experiment. This could be related to the size of the non-exchangeable K pool or to the rate of diffusion of K from the pool. However, results from Ca-resin extraction (Goulding [1984]) show that including the rate of release of fixed K improves regressions of available K on plant yield and K offtake. This suggests that the rate of diffusion of non-exchangeable K is more important than its quantity. 195 800 ao X 600 " 400 2 se .,.E 01 200 ~48.6," 100 200 300 370 Initial exchangeable K in air-dry soil (ppm) Figure 6. Relationship between K uptake during the first cropping period with ryegrass in the pot experiment and the initially exchangeable K in the soils from the Agdell experiment Rothamsted. Soils from grass 0 and arable/fallow 0 plots in 1958-67 (slope y=2.856x-13.88; r= +0.94). 6. Response of crops to soil and fertilizer K in the field Many recent experiments have shown that on soils enriched with K residues (high K status) yields of arable crops often exceed those on impoverished soils (low K status) irrespective of how much K is applied in fertilizer or manures (Johnston el al [19701; Johnston [1986]). To offer farmers sound advice on K manuring requires a knowledge of two factors, the response of Crops to soil K and the probable effect of K fertilizer at each level of soil K. 6.1 Crop response to exchangeable K Figure 7 shows a linear relationship between the yield of potatoes and field beans (Viciafaba) and exchangeable K in an experiment at Rothamsted. The range of exchangeable K values was not wide enough to determine the level above which there would be no further increase in yield, but it was larger than 200 mg K/kg. Figure 8 shows responses of barley grain and sugar from sugar beet, again in experiments at Rothamsted, but where there were more plots. Spring barley yielding about 6 t/ha grain did not need more than 80 kg/ha exchangeable 196 K, but sugar yield was still increasing up to 200 mg/kg. For both crops the general relationship was clear but there was some scatter in the values, more for the sugar beet than the barley. Sugar beet roots explore subsoils more efficiently than do the roots of barley, and subsoil K could have supplied part of the K demand for the sugar beet, the amount varying from plot to plot with the amount of root in the subsoil (see also Section 2.2). Beans grain (t/ha) K in grain (kg/ha) 0 2.5 - 20 0 1.25 10 0 Potatoes tubers It/ha) K in tubers (kg/ha) 50 125 - o5 100 0 oOo0 0 25 0 100 200 250 100 200 250 K soluble in 1 M ammonium acetate (mg/kg) Figure 7. Relationship between yield of potatoes and field beans (Vicia faba) and ex- changeable K in soil. 197 6- 0 010 'Do 5 .0 " Ov Obt 00 .A . A A 4 3 .E 2 0 0 50 100 150 200 Exchangeable K (mg/kg) 6 b3 . So 0 34 0 A A a A0 A a 0 I I I p 0 50 100 150 200 Exchangeable K (mg/kg) Figure 8. Relationship between yields of spring barley (a) and sugar from sugar beet (b) and exchangeable K in soil. Crops grown on soils with 1.5 and 2.4% organic matter (not shown) and manured from 1848-1951 with: NPK, circles; PIK, squares; unmanured, triangles and when cropping was a 4-course rotation: turnips, barley, fallow, wheat, open symbols; turnips, barley, clover, wheat, closed symbols. 198 6.2 Crop response to fertilizer K Crop response to freshly applied K fertilizer often depends not only on the exchangeable K content of the soil but also on soil type which affects the amount of K released during the growing season. Table 5 shows the response of four crops to freshly applied fertilizer on soils of different texture, each with two amounts of exchangeable K (poor and good). On the sandy loam and silty clay loam an application of potassium fertilizer increased the yield of spring barley on the low K soil to that on the high K soil. However, on the sandy clay loam so much K was released even on the low K soil that there was no response to potassium fertilizer. Also on this sandy clay loam winter wheat and winter barley yielded 8.5 and 7.9 t grain/ha respectively on both the low and high K soil, and applying K fertilizer gave no increase in yield. However, potassium fertilizer failed to increase the yields of potatoes, sugar (except on the sandy loam) and beans grown on all the low K soils to those on the high K soils. Yields of these three crops benefited from the presence of K residues in soil and fresh K fertilizer could not match these benefits. These results were all obtained on long-term field experiment sites where the various categories of soil K were in equilibrium. Such experimental sites are the only way of getting reliable information on crop response to both soil and fertilizer K. Table 5. Effect of soil texture and potassium status on the yield (t/ha) of spring barley, potatoes, sugar, from sugar beet and field beans* and the response to fresh potas- sium fertilizer Soil texture Light Medium Heavy Crop K fertilizer Soil potassium status applied Poor Good Poor Good Poor Good Barley No 3.12 3.32 3.34 3.54 5.67 5.71 grain Yes 3.38 3.31 3.56 3.56 5.67 5.86 Response 0.26 - 0.01 0.22 0.02 0 0.15 Potatoes No 32.9 41.2 17.1 27.6 28.8 43.1 tubers Yes 44.2 47.2 31.1 36.7 39.6 44.0 Response 11.3 6.0 14.0 9.1 10.8 0.9 Sugar No 3.59 4.60 3.76 4.94 6.76 - Yes 5.81 5.83 4.53 5.55 - 6.74 Response 2.22 1.23 0.77 0.61 - Beans No 2.18 2.96 2.52 4.42 grain Yes 2.67 2.97 3.60 4.38 Response 0.49 0.01 1.08 - 0.04 * Field beans: Viciafaba 199 7. Developing K manuring policies The simplest approach to K fertilization is a balance sheet approach, i.e. replace all the K taken off in the crop. Such quantities can be calculated from known yields and tables of average composition of crops such as those used by Kali and Salz A.G. in Germany and the Agricultural Development and Advisory Service in England and Wales. Such a system is best on soils which contain sufficient exchangeable K that most crops are unlikely to respond to fresh K fertilizer. However, it does not allow for any added potassium mov- ing rapidly into non-exchangeable and not immediately available forms, nor for any potassium which may be released from non-exchangeable reserves. If K is rapidly fixed by, or released from, soil then any manuring policy should allow for this but there is no rapid analytical method to measure such fixation and release. Some advisory systems do attempt to allow for soil type but only in a very unspecific way and yet soil type is clearly an important factor in the need to supply K fertilizer. If soil is so light-textured and contains so little clay and organic matter that there is almost no ability to retain potassium in exchangeable and non- exchangeable forms, then K manuring should be on an annual basis. The amount applied should be sufficient to at least meet the expected removal in the crop and timing the application should aim to minimise leaching loss. It may well be that on such soils N and K could be supplied together. On heavier textured soils, where appreciable amounts of potassium can be held in both exchangeable and non-exchangeable pools, soil fertility will be enhanced if K manuring exceeds K offtake in the crop. Such reserves should be accumulated to the extent that crops don't respond to added K fertilizer, and the K applied should maintain the soil K level. Such an approach has the advantage that the soil is enriched with K throughout the cultivated layer where roots are most active in taking up nutrients. On such soils there may be no need to apply K every year. A sufficiently enriched exchangeable pool, supported by residues and reserves will meet annual need. The pools of ex- changeable and non-exchangeable K can be maintained by rotational manur- ing. The amount of K applied every few years should at least equal the amount of K expected to be taken off in the crops to be grown before the next applica- tion. There are now two methods to check the validity of such manuring policies. The exchangeable K content of the soil can be monitored over time paying especial regard to always taking a truly representative sample from each field and maintaining a constant depth of sampling. If there is any doubt that the amount of exchangeable K is sufficient to meet the needs of the crop, then the crop can be sampled and the K concentration in tissue water determined. If the concentration is below that accepted as being sufficient for that crop then additional K should be applied. 200 8. References Addiscott, TM.: Use of the quantity/potential relationship to provide a scale of the ability of extractants to remove soil potassium. J.Agric. Sci., Camb. 74, 119-121 (1970a) Addiscott, T M.: The uptake of initially available soil potassium by ryegrass. J. Agric. Sci., Camb. 74, 123-129 (1970b) Addiscott, TM.: The potassium QII relationships of soils given different K manuring. J.Agric. Sci., Camb. 74, 131-137 (1970c) Arnold, P W.: Potassium uptake by cation-exchange resins from soils and minerals. Nature 182, 1594-1595 (1958) Arnold, P W.: The potassium status of some English soils considered as a problem of energy relationships. Proc. Fert. Soc. No. 72, 25-43 (1962) Barrow,N.J., Ozanne, P G. and Shaw, T C.: Nutrient potential and capacity. I. The con- cepts of nutrient potential and capacity and their application to soil potassium and phosphorus. Aust. J. Agric. Res. 16, 61-76 (1965) Beckett, PH. T: Studies on soil potassium. II. The immediate Q/I relations of labile potas- sium in the soil. J. Soil Sci. 15 (1), 9-23 (1964) Beckett, PH. T: Activity coefficients for studies on soil potassium. Agrochimica 9 (2), 150-152 (1965) Beckett, PH. T, Craig, J N., Nafady, M. H. M. and Watson,]. P: Studies on soil potassium. V. The stability of Q/I relations. P1. Soil 25 (3), 435-455 (1966) Beckett, PH. T and Nafady, M. H. M.: Studies on soil potassium. VI. The effect of K fixa- tion and release on the form of the K: Ca + Mg exchange isotherm. J. Soil Sci. 18 (2), 244-262 (1967) Beegle, D.B. and Baker, D.E.: Differential potassium buffer behaviour of individual soils related to potassium corrective treatments. Comm. Soil Sci. Plant Anal. 18, 371-386 (1987) Bertsch, PM. and Thomas, G. W.: Potassium status of temperate region soils. In: R. D. Munson (Ed.) Potassium in Agriculture, American Society of Agronomy, Madison, USA, pp. 131-162 (1985) Dehdrain, P2P: L'acide phosphorique du soil. Annls. Agron. 17, 445-454 (1891) Doll, E. C. and Lucas, R. E.: Testing soils for potassium, calcium and magnesium. In: L. M. Walsh and J.D. Beaton (eds.) Soil Testing and Plant Analysis. Soil Sci. Soc. Am., Madi- son, USA, pp. 133-151 (1973) Dyer, R: On the analytical determination of probably available (mineral) plant food in soil. J. Chem. Soc. 65, 115-167 (1894) Dyer, B.: A chemical study of the phosphoric acid and potash contents of the Wheat Soils of Broadbalk Field, Rothamsted. Phil. Trans. R. Soc. B 194, 235-290 (1901) Dyer, B.: Results of investigations on the Rothamsted soils. Bull. Off. Exp. Stns. US Dept. Agric. No. 106, pp. 180 (1902) Egner, H. and Riehm, H.: Lactatmethode zur Bestimmung der Bodenphosphorsiure. In: Die Untersuchung von Boden Neumann Radebeul pp. 177 (1955) Egner, H., Riehm. H. and Domingo W R.: Untersuchungen fiber die chemische Boden- analyse als Grundlage for die Beurteilung des Nahrstoffbestands der Bodenkunde. Landbruk-hogsk An. 26, 199-215 (1960) Evangelou, V P, Karanthanasis, A. D. and Blevins, R. L.: Effect of soil organic matter ac- cumulation on potassium and ammonium Q/I relationships. Soil Sci. Soc. Am. J. 50, 378-382 (1986) Goulding, K. W T: The availability of potassium in soils to crops as measured by its release to a calcium-saturated cation exchange resin. J. Agric. Sci., Camb. 103, 265-275 (1984) Goulding, K. W T and Loveland, ?J]: The classification and mapping of potassium reserves in soils of England and Wales. J. Soil Sci. 37, 555-565 (1986) 201 Goulding, K. W T and Stevens, PA.: Potassium reserves in a forested, acid upland soil and the effect on them of clear-felling versus whole-tree harvesting. Soil Use Manag. 4, 45-51 (1988) Haagsma, T and Miller, M. H.: The release of non-exchangeable soil potassium to cation- exchange resins as influenced by temperature, moisture and exchanging ion. Soil Sci. Soc. Am. Proc. 27, 153-156 (1963) Havlin, J. L. and Wesifall, D. G.: Potassium release kinetics and plant response in calcareous soils. Soil Sci. Soc. Am. J. 49, 366-370 (1985) Haylock, 0. F: A method for estimating the availability of non-exchangeable potassium. Trans. 6th Int. Congr. Soil Sci. BI, , I, 403-408 (1956) Hoagland,D. R. andMartin, J C: Absorption of potassium by plants in relation to replace- able, non replaceable and soil solution potassium. Soil Sci. 36, 1-32 (1933) Jackson, B. L.: A modified sodium tetraphenylboron method for the routine determination of reserve potassium status of soil. NZ J. Exp. Agr. 13, 253-262 (1985) Jia Xian, L. and Jackson, M. L.: Potassium release on drying of soil samples from a variety of weathering regimes and clay minerals in China. Geoderma 35, 197-208 (1985) Johnston, A. E.: Potassium fertilization to maintain a K balance under various farming systems. Proc. 13th Congr. Int. Potash Inst. Bern, pp. 177-204 (1986) Johnston, A. E. andAddiscoft, TM.: Potassium in soils under different cropping systems. I. Behaviour of K remaining in soils from classical and rotation experiments at Rothamsted and Woburn and evaluation of methods of measuring soil potassium. J. Agric. Sci., Camb. 76, 539-552 (1971) Johnston, A. E. and McEwen, J: The special value for crop production of reserves of nutrients in the subsoil and the use of special methods of deep placement in raising yields. Proc. 18th Coll. Int. Potash Inst., Bern. pp. 157-176 (1984) Johnston, A. E. and Mitchell, J. . D.: The behaviour of K remaining in soils from the Agdell experiment at Rothamsted, the results of intensive cropping in pot experiments and their relation to soil analysis and the results of field experiments. Rep. Rothamsted Exp. Stn. for 1973, Pt. 2, 74-97 (1974) Johnston, A. E., Warren, R. G. and Penny, A.: The value of residues from long period manuring at Rothamsted and Woburn. V. The value to arable crops of residues accumu- lated from potassium fertilizers. Rep. Rothamsted Exp. Sm. for 1969, Pt. 2,69-90 (1970) Karim, A. Q. M. B. and Malek, M. A.: Potassium fixation in East Pakistan soils under differ- ent conditions. Soil. Sci. 83, 229-238 (1957) Knop: Lehrbuch der Agrikultur Chemie, Leipzig, 1868. Die Bonitierung der Ackererde, (1871) Kochl, A.: Interpretation of long-term experiments with K manuring. Proc. 20th Coll. Int. Potash Inst., Bern. pp. 381-393 (1987) Kuhlmann, H., Claassen, N. and Wehrmann, J.: A method for determining the K uptake from subsoil by plants. Plant Soil 83, 449-452 (1985) Leigh, R. A.: Potassium concentrations in whole plants and cells in relation to growth. Proc. 21st Coll. Int. Potash Inst., Bern. pp. 117-126 (1989) Leigh, R. A. and Johnston, A. E.: Potassium concentrations in the dry matter and tissue water of field-grown spring barley and their relationships to grain yield. J. Agric. Sci., Camb. 101, 675-685 (1983a) Leigh, R. A. and Johnston, A. E.: Effects of fertilizers and drought on potassium concentra- tions in the dry matter and tissue water of field-grown spring barley. J. Agric. Sci., Camb. 101, 741-748 (1983b) Leigh, R. A. and Wyn Jones, R. G.: A hypothesis relating critical potassium concentrations for growth to the distribution and functions of this ion in the plant cell. New Phytol. 97, 1-13 (1984) MacLean, A. J.: Potassium-supplying power of some Canadian soils. Can. J. Soil Sci. 41, 196-206 (1961) 202 Martin, H. W. and Sparks, D. L.: On the behaviour of nonexchangeable potassium in soils. Comm. Soil Sci. Plant Anal. 16, 133-162 (1985) Mengel, K.: Experimental approaches on K + efficiency in different crop species. Proc. 21st Coll. Int. Potash Inst., Bern. pp. 47-56 (1989) Mengel, K. and Kirkby E. A.: Principles of Plant Nutrition- 3rd Edition, 655 pp. Bern: Int. Potash Institute (1982) Moberg, J. P andNielsen. J D.: Mineralogical changes in soils used for K-depletion experi- ments for some years in pots and in the field. Acta Agric. Scand. 33, 21-27 (1983) Morgan, M. F: Universal soils testing system. Connecticut Agr. Exp. Stn. Bull. 392, 129-159 (1935) Nair, P K. R. and Grimme, H.: Q1I relations and electroultrafiltration of soils as measures of potassium availability to plants. Z. Pflanz. Bod. 142, 87-94 (1979) Ndmeth, K.: Electro-ultrafiltration of aqueous soil suspension with simultaneously varying temperature and voltage. Plant Soil 64, 7-23 (1982) Nemeth, K.: Recent advances in EUF research (1980-1983). Plant Soil 83, 1-19 (1985) Novozamsky, L andHouba, V J G.: Critical evaluation of soil testing methods for K. Proc. 20th Coll. Int. Potash Inst., Bern. pp. 177-197 (1987) Ogunkunle, A. 0. and Beckett, P H. K: The efficiency of pot trials, or trials on undisturbed soil cores, as predictors of crop behaviour in the field. Plant Soil 107, 85-93 (1988) Olsen, S.R., Cole, C. V., Watanabe, E S. and Dean, L. A.: Estimation of available phospho- rus in soils by extraction with sodium bicarbonate. Circ. US Dept. Agric. No. 393, pp. 19 (1959) Prianishnikov, D.N.: Quantitative Bestimmung der in Boden vorhandenen absorptiv gebundenen Basen. Landw. Vers. Stat. 79-80, 667-680 (1913) Qudmener, J.: The measurement of soil potassium, IPI Res. Top. No. 4 Int. Potash Inst., Bern (1979) Quemener, J: Important factors in potassium balance sheets. Proc. 13th Congr. Int. Potash Inst., Bern, pp. 33-63 (1986) Ross, G.J.S.: MLP: Maximum Likelihood Programme. Harpenden: Rothamsted Ex- perimental Station, 256 pp. (1980) Sailakshmiswari, Y, Rao, A. S. and Pillai, R. N.: An assessment of potassium release characteristics of some deltaic alluvial soils. J. Pot. Res- , 109-116 (1985) Schofield, R. K.: A ratio law governing the equilibrium of cations in the soil solution. Proc. Ilth Int. Congr. pure appl. Chem., London 3, 257-261 (1947) Schofield, R. K. and Taylor, A. W.: Measurements of the activities of bases in soils. J. Soil Sci. 6, 137-146 (1955) Schollenberger, C.J.: Exchangeable hydrogen and soil reaction. Science 65 (1692 ns) 552-553 (1927) Schollenberger, C. J. and Dreibelbis, E R.: The effect of cropping with various fertilizer, manure and lime treatments upon exchangeable bases of plot soils. Soil Sci. 29, 371-394 (1930a) Schollenberger, C. J. and Dreibelbis, F R.: Analytical methods in base exchange investiga- tions on soil. Soil Sci. 30, 161-173 (1930b) Sinclair, A. H.: A comparison of EUF and Q/1 measurements of soil K with its uptake by ryegrass in Scottish soils. Plant Soil 64, 85-94 (1982) Sparks, D. L. and Huang, P M.: Physical chemistry of soil potassium. In: R. D. Munson (ed-). Potassium in Agriculture, Amer. Soc. Agron., Madison, USA, pp. 201-276 (1985) Sparks, D. L. and Liebhardt, W. C.: Effect of long-term lime and potassium applications on quantity-intensity (Q/1) relationships in sandy soil. Soil Sci. Soc. Am. 1. 45, 786-790 (1981) Talibudeen, 0., Beasley, J.D., Lane, P and Rajendran, N.: Assessment of soil potassium reserves available to plant roots. J. Soil Sci. 29. 207-218 (1978) 203 Talibudeen, 0. and Dey, S. K.: Potassium reserves in British soils. I. The Rothamsted classi- cal experiments. J. Agric. Sci., Camb. 71, 95-104 (1968) Villenmin, R: Translation of laboratory K-data into K fertilizer recommendations. Proc. 20th Coll. Int. Potash Inst., Bern, pp. 199-210 (1987) von Liebig, H.: Soil statistics and soil analysis. Z. landw. Ver. (Abstract in: I. Chem. Soc. (Abstr.) 25, 318 and 837 (1872)) Warren, R. G. and Johnston, A. E.: The accumulation and loss of soil potassium in long term experiments at Rothamsted and Woburn. Proc. Fertil. Soc. 72, 1-24 (1962) Woodruff C. M.: The energies of replacement of calcium by potassium in soils. Proc. Soil Sci. Soc. Am. 19, 167-171 (1955) York, E. J, Bradfield, R. and Peech, M.: Calcium-potassium interactions in soils and plants. I. Lime-induced potassium fixation in Mardin silt loam. Soil Sci. 76, 379-387 (1953) 204 Estimation of Root Density in Modelling Nutrient Requirements G. Wessolek and S. Gdth* Summary Potassium (K) delivery to the roots of cereal plants in soils is determined by the following factors: chemical and physical properties of the soil climatological conditions (water stress) and plant factors such as root growth and root density. The phrase 1. Objectives For the simulation of the supply of potassium and phosphate to the roots through diffusion, data for several root parameters are needed (Jungk and Claassen [1986]). Depending on the mathematical basis of the model, root radius, root length and/or density Lv (cm • cm - 3), and the root surface area are the relevant factors (Wessolek and Gdith [1989]; Claassen et al [1986]; Grimme et al. [1971/). However, measurements of root length are tedious and time-consuming. For this reason, an evaluative framework was established on the basis of the results from numerous root studies, to determine the root length densities (Meuser et al [1987]). This evaluative framework should enable users of the model to rely on average root length data in the absence • Dr. G. Wessolek, Technical University of Berlin, Department of Ecology, Institute for Soil Science, Salzufer 11-12, 1000 Berlin 10, Federal Republic of Germany and Dr. S. GOath, Justus Liebig University, Institute for Microbiology and Land Improvement, Senckenbergstrasse 3, 6300 Giessen, Federal Republic of Germany 205 of their own data on roots. Moreover, it is conceivable that the method could also be applied to other research areas. In the second part of this paper, simulation results for the potassium deliv- ery are presented. In these studies, the influence of different climatic condi- tions and root densities on K-delivery from the subsoil of a Luvisol (loess) are discussed. 2. Study sites The root data for cereal plants were collected from 25 different sites, some of which had been investigated continuously for many years. The data are listed in Table I according to the root zone depth of the soil. Table 1. List of the study sites sampled for the root evaluation framework Soil type Parent material Effective Available mois- Crop (FAO [1988]) rooting zone ture at field (cm) capacity (nFK) (mm H 20) I. Root zone depth <60 cm: Cambisol graywacke-schist ...... 40 62 rye Cambisol shale ...... 40 36 rye Gleysol pleistocene mat ...... 50 131 oats Leptosol pleistocene mat ...... 50 89 wheat Cambisol loess ...... 50 92 wheat Leptosol shale ...... 60 91 wheat Cambisol sand ...... 60 74 wheat Luvisol glacial marl ...... 60 90 wheat 2. Root zone depth 60-100 cm: Vertisol glacial clay ...... 80 83 wheat Cambisol loess ...... 90 208 wheat Luvisol loess ...... 90 166 wheat Fluvisol flood sediments ...... 90 156 wheat Solonchak tidal mud deposits ...... 90 185 wheat Gleyic- loess ...... 100 240 wheat Luvisol Gleyic- loess ...... 100 236 oats Luvisol Luvisol loess ...... 100 236 wheat Phaeozem loess ...... 100 170 wheat Luvisol loess ...... 100 193 wheat Luvisol loess ...... 100 188 wheat 3. Root zone depth>100 cm: Luvisol loess ...... 110 251 oats Regosol loess ...... 110 180 barley Fluvisol flood sediments ...... 110 155 barley Regosol loess ...... 110 189 barley Solonchak tidal mud deposits ...... 120 210 wheat Anthrosol loess ...... 120 315 wheat 206 3. Methods Two methods were applied to determine the root lengths: - Soil core-/monolith method With the aid of soil cores or monoliths, volume samples were taken from the soil, washed over a >0.6 mm sieve width and the length determined by the intersection method of Newman (cited in: BOhm [19791) or the modified method of Tennant (cited in: Bohm [1979]). - Profile wall method At the profile wall, the root system was directly transferred to a foil; before- hand, a defined soil layer was rinsed off with a hand sprayer so that the root lengths could be related to a specific soil volume. The methods can be cor- related by applying conversion factors (Kopke [19791). An attempt was made to relate the existing root densities observed and pheno- logical stages of development with specific soil and climatic data. In this manner, information was obtained on parameters which influence root length growth. Multiple regression analysis was applied for the calculations. The soil parameters included variables that could easily be determined or estimated in the field: - physical parameters of the soil: air capacity (vol. %), available moisture at field capacity (nFK) (vol. %), particle size distribution (weight %/a) and/or soil type, bulk density (g • cot '), - chemical parameters of the soil: cation exchange capacity (me/100 g soil), organic matter content (weight 076), CaCQ 3-content (weight 0), pH value. The climatological parameters included in the calculations for the vegetation period were: precipitation (mm), air temperature (°C), climatic water balance (mm). Two approaches were followed in calculating, the correlation: In the first, an attempt was made to find explicit relationships between each soil depth and the soil and climate data. In the other, depth itself was taken as a variable in the calculations. Both linear and non-linear relationships were incorporated in the calcula- tions. The soil and climate data were correlated individually as well as in different combinations with the root length densities. In order to take into account the heterogeneity of the data, the different approaches for the calculations also included grouping into classes (for example, according to available moisture at field capacity or soil type). 207 4. Results and conclusions of the calculations In all calculation runs, a strong relationship was found between root density (Lv) and soil depth. Other more specific parameters taken individually or in combination exhibited either no correlation or correlations deviating from logical theoretical principles. From the pedologic data, the closest relationships were identified between Lv and the soil type (Figure 1) and/or the available water capacity. In regard to soil type, the smallest Lv values of the loamy and silty soils were frequently still greater than the greatest values of the sandy sites. When available water capacity was taken as input variable and soil depth was included, a close corre- lation (r2=0.86) was foundwith root density in relation to 10 cm depth steps for the stage of ear emergence (Equation 1): Eq. (1) Lv=-9.53 log depth+0.10 nFK/dm+16.13 Root density Lv (cm. cm -3) S Loamy and silty soils 10- 0 Sandy soils 9- TTMax. 8- j) Mean J- Min. 7- 64 6- -T 5 4- I 2- I -20 -30 -40 -50 -60 -70 Depth (cm) Figure 1. Root density (cm cm- 3) of loamy-silty and sandy soils as a function of soil depth. 208 The calculated values from this equation are given on the nomogram (Fig- ure 2). Root density Lv -3 (cm - cm ) Depth (cm) 0-10 8.0 7.0 10-20 5.0 5'0- 20-30 4.0 30-40 3.0- 40-50 2.0- !50-60 60-70 1.0- 70-80 5 10 15 20 25 30 35 Available water capacity - pF 1.8-4.2 (mm . dm ) Figure 2. Root density (Lv) of cereal plants at the stage of ear emergence as a function of available water capacity of the soil (nFK) and depth. Shown example: nFK=20 mm/dm, depth=30-40 cm, Lv=2.7 cm •cm- 3. 209 Depth (cm) -.. 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-SO 19 I- - 8 6 4 2 0 2 4 6 8 Root density (cm cm - 3 ) Figure 3. Root densities at the stage of flowering of winter wheat in a vegetation period with (1988) and without (1979) water stress (Gdth et aL [19891). 210 If the variables influencing available water capacity, soil type and soil depth are combined, the Lv can also be determined for other phenological stages before and after ear emergence because there are close relationships between the periods of investigation. The Lv of other time periods can be calculated if the root density at the onset of ear emergence is known (Equations 2-7): - sandy soils (sand >30 weight 076, clay <25 weight 0): (2) Lv(stem elongation) =0.751 Lv(ear emergence) - 0.06 (r = 0.97) (3) Lv(floweifs) = 1.232 Lv(.r emergence) - 0.06 (r = 0.99) (4) Lvceow ripeness) = 1.175 Lv(ear emergence) +0.04 (r = 0.96) - loamy and silty soils (silt>50 weight /o, clay <25 weight 07t): (5) LV(stem elongation) = 0.972 LV(ear emergence) -0.06 (r = 0.99) (6) Lvnowering) =0.969 • Lv(ear emergence) +0.16 (r =0.98) (7) Lv(yeliow ripernvs) =0.882 LV(ear emergence)+0.35 (r=0.93) Apart from the soil properties, climatological conditions also have an in- fluence on the root growth. In Figure 3 measured root densities for winter wheat are presented for two years with different amounts of rainfall during the vegetation periods. The root growth measurements were carried out on a Luvisol (loess) near Hannover at the stage of ear emergence. For the plants in the 1979 growing season no real water stress occurred (pF <2.5), while in 1988 a high water deficiency (> 200 mm) was registered in the months April and May. As a reaction to the droughtiness the root densities in the 0-55 cm soil layer decreased, while the root densities in the deeper layers increased. From these results it could be concluded, that the calculated Lv values must be adjusted by deductions or additions as a function of the available water capacity and climatic water balance (precipitation minus potential evapotran- spiration). Corresponding correction values as a function of the climatic water balance are shown in Table 2 for the phenological period flowering/ear emer- gence for sandy and silty soils. Table 2. Correction values of the root density Lv (cm - cm 3) for the phenological period flowering/ear emergence as a function of the climatic water balance and of the soil texture (I=sandy soils, II = loamy and silty soils) Climatic water balance Positive Negative -100 mm -200 mm Root density (Lv) I II 1 1i I 11 7 - +2.5 -0.7 -0.5 - 1.5 -1.0 211 The method described is suited for determining the root densities of cereal plants in simulation models for water and nutrient transport. Clayey soils and/or horizons (clay content > 25 weight %) had to be excluded due to their particular swelling and shrinkage properties which influence root growth. 5. Principles of the transport process Molecular diffusion and convection play major roles in nutrient transport processes in soils. Because convection is only dependent on the pore water movement and on the concentration of dissolved substances in the soil water, its importance in nutrient delivery to plant roots is easily estimated from the transpiration and average concentrations. The process of diffusion in the soil can be described by Fick's First Law (Equation 1): Eq. (1) l= -Da - dC .dx- Id: diffusion flow rate (pg cm2 sec-l) 2 Da: apparent diffusion coefficient (cm sec -) dC1 dx-I : concentration gradient (pg • cm 3 cm- 1) Nutrient transport by diffusion is found in the area near the roots, if in the course of nutrient uptake a concentration gradient dCI/dx arises between the root surface and the surrounding soil. The rate at which the nutrient is delivered by diffusion is dependent on the so-called diffusion coefficient. In contrast to free water, the apparent diffusion coefficient in the soil Da is affected by physical (water content, tortuosity factor (= ((inhibiting fac- tor))) and chemical factors (buffering) (Eq. 2): Eq. (2) Da=D, • ® - f. b -1 Da: apparent diffusion coefficient (cm' •sec- 1) 2 DI: diffusion coefficient in free water (cm sec ) 3 e: water content (cm cm- 3) f: tortuosity factor (Beese [1986]) b: nutrient buffering capacity The nutrient buffering capacity is a term for the relationship between the total nutrient content of the soil involved in diffusion and the nutrient concen- tration of the soil solution (Eq. 3): Cexch Eq. (3) b- CL b: nutrient buffering capacity Cmxch: exchangeable nutrient amounts (pg. cm 3) - 3 Cl: nutrient concentration in the solution (pg . cm ) 212 The potassium flux into the roots depends on K concentration in the soil solution at the root surface. The net influx (In) can be described by the Michaelis-Menten equation (Eq. 4) as modified by Nielsen [1976]: Eq. (4) In = Imax (C-Cmin) 2 ) Km+C-Cmin In is the net influx of potassium. lmax the maximum influx, is the parameter for the capacity of a root to absorb a nutrient. Kin, the Michaelis constant, describes the affinity of the uptake system for a nutrient and Cm is the mini- mum concentration to which plants can deplete a solution. A detailed descrip- tion of this model is given by Claassen et al. [1986j. 6. Model calculations for potassium uptake The following calculations demonstrate the influence of the root density on potassium uptake from the subsoil for different climatological conditions. The model simulations have been done for a vegetation period with high, medium and low rainfall on a Luvisol (loess) near Hannover. For the calcula- tions, the root densities of the topsoil (0-30 cm) have been varied from 4 to 8 cm . cm -. For these conditions, the potassium uptake from the subsoil (in percentage) is presented in Figure 4. The results show that the K-uptake from the subsoil varies from 37% in dry vegetation periods to 24% in wet periods. On average the subsoil delivery amounts to 25-30% for Luvisols in the northern parts of West-Germany. Furthermore, the results demonstrate, that with decreasing root densities in the top soil, the K-uptake from the subsoil increases rapidly to 40-50%. On the other hand, the K-delivery of deeper soil layers decreases with increas- ing root densities in the Ah-horizon. Investigations of Kuhlmann [1987] show that high root densities ( > 9 cm • cm 3) result in decreasing potassium uptake per unit root area. The above mentioned model results have to be interpreted carefully, because in nature plant-physiological processes might affect these interactions in a not yet known way. Finally, the results of Figure 5 demonstrate the accuracy of the simulated K-uptake by a comparison with measured field data. The comparison for two vegetation periods shows that the calculated uptake of potassium differs by 20-30 kg/ha from the measured field data. It can be concluded that the model can quite well be used for K-delivery quantifications. The restrictions of the model are strongly related to limited knowledge of specific interactions between soil, soil solution and root surface (root-hairs) as a function of time, climate and species. 213 55" Luvisol loess deposit 50. Hannover-Ohlendorf summer wheat 45- 40- 35- 0 30- 0 25- 0, E weather conditions 0- 215 - "during the vegetation period = (O in Q-30 cm) U dry (pF >3.01 0 10 - norm al(pF'2-3) 5wet (pF < 2.2) 0.0 4 8 12 Mean root density in0-30 cm depth (cm cm - ') Figure 4. Potassium uptake from the subsoil, as a function of different root densities in the topsoil and climatological conditions for summer wheat. 214 Cumulative K-uptake (kg K/ha) 350 Luvisol winter wheat summer wheat 1979 * loess deposit 1988 250-300 Hannover-Ohlendorf , 250- */ 200 • 150 U * 100 smeasured 50 - * calculated 50- 1 100 110 120 130 140 150 160 170 180 days April I May I June month Figure 5. Comparison of measured and calculated potassium uptake for Iwo years with different climate conditions (Gth et aL 11989]). 7. References Beese, F: Parameter des Stickstoffumsatzes in Okosystemen mit B6den unterschiedlicher Aciditt. Gbttinger Bodenkundl. Ber. 90, 1-344 (1986) Bohm, W.: Methods of studying root systems. Springer-Verlag (1979) Claassen, N., Syring, K.M. and Jungk, A.: Verification of a mathematical model by simulating potassium uptake from soil. Plant and Soil 95, 209-220 (1986) Gth, S., Wessolek, G. and Renger M.: Modellrechnungen zur Bedeutung des Unterbodens bei der K-Versorgung von Getreide auf LOIbOden. VDLUFA, Schriftenreihe, Kongrei- band 1989 (in press) Grimme, H., Nemeth, K. and von Braunschweig, L. C.: Some factors controlling potassium availability in soils. Proc. Int. Symp. Soil Fert. Evaluation, N. Debli, Vol. 1.33-43 (1971) Jungk, A. and Claassen, N.: Availability of phosphate and potassium as the result of inter- actions between root and soil in the rhizosphere. Pflanzenern. Bodenk. 149, 411-427 (1986) Kopke. U.: Ein Vergleich von Feldmethoden zur Bestimmung des Wurzelwachstums land- wirtschaftlicher Kulturpflanzen. Dissertation Universitfit Gottingen (1979) Kuhlmann, H.: Ursachen und Ausmafi der N-, P-, K- und Mg-Ernlihrung der Pflanzen aus dem Unterboden. Habilitationsschrift am Fachber. Gartenbau, Universitkit Han- nover (1987) Meuser H., Wessolek, G. and Renger, M.: Ein Verfahren zur Abschitzung der Wurzellan- gendichte von Getreide. Mitt. Dr. Bodenkdl. Ges. 55, 637-642 (1987) Nielsen, N. E.: A transport kinetic concept for ion uptake by plants. Plant and Soil 45, 659-677 (1976) Wessolek, G. and Gth, S.: Integration der Wurzellllngendichte in Wasserhaushalts- und Kaliumanlieferungsmodellen. Kali-Briefe (BOntehof) 19 (7), 491-503 (1989) 215 Modelling K Uptake by Plants from Soil PB. Barraclough* Summary Potassium uptake depends on the supply from the soil and the demand by the plant and over a dozen plant and soil properties are currently recognized as being important for K uptake. Nearly all K reaches roots by diffusion which has the potential to supply K at rates considerably in excess of the requirements of high-yielding crops. Many plant and soil properties feature in mechanistic models which simulate the dynamic interaction between supply and demand and the complexity of the models makes it difficult to validate and apply them, especially in the field, although they greatly aid our understanding of K uptake. Particular difficulties in modelling K uptake are caused by the measurement of soil buffer power, and the measurement and prediction of root growth and morphology. More meas- urements on concentration at the root surface are required to establish how plant demand should be characterized. The potassium concentration in the soil solution that is required for high crop yields can be calculated from a simplified diffusion model and for a winter wheat crop with an active rooting density of 3 cm/cm 3, ranges from 50-200 ItM K, depending on the soil conditions. 1. Introduction The practical approach to ensuring adequate K nutrition for crops has been to identify a minimum level of soil K above which there is no yield response to fertilizer. The most convenient and reproducible measure of plant-available K in the soil is that which exchanges with ammonium ions (K,,) and the critical soil K,, level by this method is currently 120 ppm in the UK (MAFF[1986]). This procedure takes no account of several processes that can influence a crop's response to K fertilizer and a universal critical K,, value does not exist; rather it varies with the soil, season and crop species. To explain why variations in fertilizer response occur and to improve fertilizer recommendations, it is necessary to understand the underlying mechanisms that contribute to K uptake. The prediction of K uptake by plants growing in soil, from a knowledge of plant and soil properties, is a formidable problem because of the many processes that are involved. Potassium uptake depends on both the supply * Dr. PB. Barraclough, Biochemistry and Physiology Department, AFRC Institute of Arable Crops Research, Rothamsted Experimental Station, Harpenden Herts AL5 2JQ, United Kingdom 217 from the soil and the demand by the plant. These two processes are linked via the growth and activity of the root system. Components of supply are the total amount of available K and its equilibrium with the soil solution, the growth and morphology of roots, and the transport of K to roots by mass flow and diffusion. Demand is largely determined by the growth rate of the plant, but amounts in excess of immediate growth requirements may be taken up and stored in the plant. Our understanding of K uptake has been greatly aided by the development of mechanistic mathematical models which simulate the dynamic interaction between supply and demand and enable uptake to be predicted from soil and plant properties. Description of such modelling approaches can be found in Nye & Tinker [19771 and Barber [19841. The objectives of this paper are to discuss some of the difficulties in validating and applying these models and in our current understanding of the principles governing K uptake. A simpli- fied approach for estimating critical soil K levels for uptake, which bridges the gap between the complexity of the dynamic models and the simplicity of practical soil tests, is presented. 2. Nutrient supply-uptake models Over a dozen plant and soil properties are currently recognized as being im- portant for K uptake and many of these are incorporated in published simula- tion models (Baldwin et al. [1973]; Barber & Cushman [1981]; Claassen et al. [1986]). In these models, a particular nutrient is transported radially to an absorbing root by mass flow and diffusion which is characterized by four soil factors: soil solution concentration (CL), buffer power (b), diffusion coefficient (D) and transpiration rate (V,), and by three root parameters: rate of growth (k), radius (ro) and half-distance between roots (ri). Roots are as- sumed to be uniformly distributed in a given soil volume and, to account for root competition, each root is only allowed to exploit a concentric cylinder of soil. As rooting density increases, the radius of these cylinders (ri) decreases. At the root surface, the ion is absorbed at a rate which depends on external concentration and this is characterized by three root absorption factors: maximal flux (Fmax), concentration for half maximal flux (Kin) and minimum concentration for net uptake (Cmin). Silberbush & Barber [1983] have performed sensitivity analyses for K uptake by soyabeans and established that the order of importance of the parameters was: k, ro, CL, b, D, Fiax, rl, Vo, Cmin, Km. The models have successfully predicted K concentration profiles close to roots under artificial conditions (Claassen el al [1986]). They have also predicted the uptake of K reasonably well for young plants growing in pots of soil (Barber [1984]) and even for soyabeans growing in the field (Silberbush & Barber [1984]). However, more recent tests showed that they over-predicted K uptake by field-grown wheat by upto 4 times (Seward et al [1990]). 218 3. Soil supply 3.1 The transport processes: mass flow and diffusion Mass flow and diffusion in the soil solution are now accepted as the means by which nutrients reach roots. Direct contact exchange between roots and soil solids is physically impossible and nutrients are not ingested by an advanc- ing root tip (so called ((root interception>). Mass flow is conventionally as- sumed to be the «faster of the two processes. It is calculated as the product of transpiration rate and nutrient concentration in solution. The amount of nutrient supplied by diffusion is calculated as the difference between meas- ured plant uptake and that supplied by mass flow. However, diffusive supply can be readily calculated in simple situations. For radial diffusion to a cylindrical sink (the root) in an infinite volume, the change in concentration (C) around the root with time (t) is given by Fick's second law; dC = I d (rDdC) (I) dt r dr dr where r is the radial distance from the root axis and D the diffusion coefficient of the nutrient. This is readily solved under steady-state conditions, where concentrations are constant at the root surface and at some fixed distance from the root. The solution has been given by Baldwin et al. [1973] and has been re-arranged and transformed (Barraclough [19861) to give: I = 2nDLOfAC1 (2) In( x/1.65r0 ) where I (the inflow) is the diffusive flux per unit root length, DL is the diffusion coefficient of the nutrient in water, 0 the volumetric soil water content, f the ionic impedance factor, ACL the difference in concentration of the nutrient in the soil solution between the root surface and the bulk soil, x the mean half-distance between roots (=l/iiLU, where L, is root length density 3 (cm/cm )) and r0 is the root radius. Equation (2) can be used to assess the potential of diffusion for supplying K compared with that by mass flow. Under the temperate conditions of the UK, a typical evapotranspiration rate for a rapidly growing wheat crop in a moist soil is 3 L/m 2/day (Barraclough et al. [1989]). Thus, given a K concen- tration in the soil solution of 100 pM, the potential supply rate by mass flow is 0.12 kg/ha/day. Under the same conditions with 3 cm/cm3 of active root, $ 2 with DL = 1.24x 10- cm /s, 0= f= 0.4 and r0 =0.015 cm (Barraclough [1989]), the potential rate of supply of K by diffusion (from equation (2)) is 9.8 kg/ha/day; nearly 100 times greater than that by mass flow. Actual uptake rates are of the order of 5 kg/ha/day (Table I) so diffusion is the major 219 mechanism for supplying K in most cases and mass flow can be ignored. Diffusion in fact has the potential to more than meet the K requirements of rapidly growing crops. Table I. Potassium uptake and root growth of some high-yielding crops Crop Total DM Max. K Max. K Max. L. Reference (t/ha) uptake uptake in topsoil (kg/ha) rates (cm/cm') (kg/ha/d) Potatoes 13 325 7.3 2.0 Asfary et al. [19831 W. Rape 16 308 3.5 9.7 Barraclough 119891 W. Wheat 18 285 5.2 10.1 Barraclough [1986] Maize 23 329 11.6 4.0 Mengel and Barber [1974] S. Beet 28 367 6.8 2.5 Armstrong et al. [1986] 3.2 The transport parameters: CL, b, D Potassium in soil is present in four forms (1) solution (K,,,,), (2) exchangeable (K¢ ), (3) fixed (Knx) and (4) matrix (Kma,). The total K content of temperate topsoils is of the order of 30 t/ha, but only 107a of this may be immediately available to plants from K,01. and Kx, and less than 0.307a of this may become additionally available during a growing season from fixed sources (Wild [1988]). Roots absorb K from solution where it is present in very low concentra- tions, although the maintenance of K levels in solution is critical for yield (see section 5). K,01 is in rapid equilibrium with Kex and the ratio K~s/K,01 , the slope of the sorption isotherm, is the buffer power. This ratio is rarely constant but increases with decreasing concentration in solution. Kx is in slow equilibrium with Kf 1,, whilst release of Kmar can be considered negligible on an agricultural time-scale. Crop yields are usually well correlated with K,, (see for example, Arnold [1962]) and this is the basis of most fertilizer recommendation systems throughout the world, although it has been shown in pot studies (Sinclair [1979]) and surmised from field studies (Johnston & Poulton [19761) that Kri is released and taken up during growth. It is not easy to establish the contribu- tion that Kfix makes to the uptake of a field crop. It is necessary to accurately measure changes in Kx in the topsoil during growth and also to know how much K comes from the subsoil. Detailed changes in K,, during crop growth have rarely, if ever, been measured. The fact that in most soils Kx changes on drying is an added complication. In K-rich soils some K,, is likely to be fixed, and in K-poor soils some fixed K is likely to be released to Kx on drying. Sinclair [1979] found that K,, was increased by 20016 (from a low value of 40 ppm) when K-depleted sandy soils were air-dried, and by 40% for a heavier soil. 220 The potential K uptake from the subsoil can be calculated from equation (2). Using values for winter rape at Rothamsted of DL (100C) = 1.28 x 10 5cm 2 /s, 0=0.35, f=0.18, ro=0.010 cm, A CL= 30 iaM and L, (20-100 cm)= 1.3-0.2 cm /cm' (Barraclough[1989]), the maximum potential uptake rate from the sub- soil is 0.67 kg/ha/day. The main uptake period lasts about 100 days for this crop and 200 kg K/ha can be taken up in this time, so about one third of the total uptake could come from the subsoil. This is in good agreement with the results of Kuhlmann et al [19841. The rate of release of fixed K is increased by lowering the K concentration in the soil solution. Jungk et al. [19821 made detailed studies of K behaviour in the rhizosphere. Maize roots depleted K in solution from over 700 AM to 2 pM at the root surface in 2 days. This caused fixed K to be released within the root hair cylinder (about I mm from the root surface) to an extent of about twice the initial exchangeable amount (Figure 1). quantity of K released K conc. of soil solution 700- 100. 600- 200- 0 500- Herrenhausen 300- sandy soil 0 E 4NOc 4% clay z0 400 exch. K - A6 B0ten silt loam 300- 4cly100- 20 B~ilten silt loam 500 700nhuen /2121% clay 0 '1 0 1 2 3 4 5 6 0 1 2 3 4 5 6 distance from the root Figure I. Depletion of exchangeable K and soil solution K in the vicinity of 2-day old maize roots growing in sandy and silt loam soils. Arrows indicate initial ex- changeable K. (After Jungk et al. 1)982]). Many chemical extraction procedures for assessing the availability of fixed K have been devised including boiling acids, repeated extraction with Ca solu- tions, continuous cropping with ryegrass, electro-ultrafiltration and extrac- tion with exchange resins. Goulding [1984] continuously extracted K from soil suspensions with a Ca-saturated ion exchange resin. This method distin- guishes the different forms of K that are being released, and Goulding found that a multiple regression of the amount and rate of release of K.. and the rate of release of Kfj. gave the best correlation with crop uptake. These various 221 methods in their present form are most suited to characterising the long-term release of K from soils. They cannot provide a realistic assessment of the release of Kfi to crops in a single season, because they employ conditions (soil suspensions, shaking, electric fields) far removed from the soil-root situa- tion in practice. A method for assessing buffer power which will include the contribution made by Kfi, as well as K,, is urgently needed, not only for ac- curate modelling but also for practical recommendations. Diffusion coefficients are tedious to measure in soil but can be calculated from: D=DL0f/b (Nye [1979]). The product Db usually arises in transport calculations rather than D alone and this is readily calculated if f is known. Values of f for different soils can be found in Nye [1979] and Barraclough & Tinker [1981]. The above relationship assumes that all diffusion occurs in the liquid phase and that none occurs in the exchange phase. Evidence has been found (Staunton [1986]) that sodium diffuses in the exchange phase to a significant extent but, todate, no information is available for K. If K were to diffuse to any great extent in the exchange phase, this would be a major complication in the prediction of D. 3.3 Root growth parameters (k, r0 , rl) Root growth is the most important factor in uptake and one of the most difficult to measure and predict. The simplest measurements required are of root length and radius. It is still not clear whether different classes of root (e.g. seminal and nodal roots in cereals, and various orders of laterals) have different uptake characteristics. Kuhlmann andBarraclough [1987] found that most of the difference in K uptake rates by seminal and nodal systems of wheat could be accounted for by their different radii, but more information on this point, particularly for a wider range of crop plants, is required. If roots do have different uptake characteristics then it would be necessary to know the rates of growth and the radii of these classes individually. If they do not, then all roots can be treated as equal for uptake purposes and the determination of the total root length and the mean root radius of a system should be sufficient. These are relatively easy to measure but not to predict. Roots appear to grow and branch at constant rates under uniform conditions and elaborate descriptive models of root growth and architecture have been developed based on this (Lungley [1973/; Diggle[1988]).The growth and mor- phology of root systems, however, depends on shoot growth and numerous soil factors (temperature, moisture, oxygen, nutrients, mechanical impedance and porosity) many of which interact, so that mechanistic prediction is a formidable, if not impossible, task. Growth and branching have yet to be related to some of the above soil properties. The response of roots to soil K for example is unknown. Work in solution culture has shown that K may be unique in that K deficiency, unlike N and P deficiency, does not elicit com- pensatory root growth (Drew [19751). This has been confirmed in pots of soil for winter wheat (Table 2). The reason for this difference in behaviour is 222 unknown, but could be connected with minimum concentrations (either inter- nal or external) which must be reached before assimilate partitioning is shifted in favour of root growth. Table 2. Response of root/shoot weight ratio in winter wheat to N, P and K in soil (H-high, L =low) HP HP LP LP HK LK HK LK H N ...... 0.26 0.24 0.42 0.37 LN ...... 0.54 0.39 0.49 0.47 Root hairs can be important for K uptake (Claassen and Jungk [1984]) and most be included in any accurate model. Itoh and Barber [1983] modelled P uptake by hairs and concluded that hair length was more important than hair density since the depletion zones of densely packed hairs soon over- lapped. The measurement of root hairs in soil is extremely difficult and the contribution that hairs make to uptake will always be difficult to predict under all but the most artificial conditions. Hairs are affected by many soil factors (Cormack [1962]) but we cannot predict how far they are likely to extend in a given soil. The measurement of root growth in the field is especially difficult and prediction is confounded by the fluctuating environmental conditions. Nevertheless, progress has been made in relating root and shoot development (Klepper et al. [1984]) and root growth and depth to shoot growth (Burns [1980]; Barraclough et al. [1990]). A further complication in the field is that of root distribution, both with depth in the soil profile and within a given horizon. Distribution of root length with depth often conforms to an ex- ponential decay function (Gerwitz and Page [1974]), although the parameters in such a function have yet to be related to soil conditions. In a uniform soil roots are evenly distributed and this is easily characterized by a single, average value of the half-distance between roots. In natural aggregated soils however, soil structure may have a big effect on the spatial distribution of roots and heterogeneous distributions are neither easy to measure nor characterize. Random as opposed to regular distributions may not overly affect K uptake (Baldwin et al. [1972]), but extreme (clumping of roots could seriously reduce uptake. 4. Plant demand Plant demand can be specified empirically either as the uptake rate of a crop growing at its maximum rate (Table 1) or as the product of crop growth rate and some critical internal nutrient concentration. At a more physiological level, numerous studies in solution culture have established relationships be- 223 tween uptake per unit of root (flux) and concentration in solution. This pro- vides a natural link with diffusive supply. A Michaelis-Menten type relation- ship which is specified by Fmax and Km is most commonly used in models. These parameters are often linked to the number and kinetics of the putative ion carriers in cell membranes. The parameters vary with temperature, plant age and nutrient status and root/shoot ratio. The most sophisticated version of the relationship incorporates the internal concentration in the shoot as well as Fmax and Km as functions of concentration in the root (Glass and Sid- diqui [1984]). This approach emphasises the importance of concentration at the root surface and many studies have established Table 3. Supply (b,D) and demand ('max, K.) parameters used to model K uptake in Fig. 2. (L,=3 cm/cm3, ro-0.015 cm) imax Km b D Uptake (tmol/cm2/s) (PM) (cM2/s) (nmol/cm of root) a (5 hrs) ...... 1.5 10- ' 1 20 10- 15.6(5 hrs) A (24 hrs) ...... 1.5× 10- * 1 20- 10- 7 53.5 (24 hrs) B (24 hrs) ...... 5.OX 10- 6 50* 20 10-7 22.1 (24 hrs) C (24 hrs) ...... 1.5x 10- ' 1-* 2 10- 6 32.8 (24 hrs) D (24 hrs) ...... 5.Ox 10- 6 50 2 10-6 18.9 (24 hrs) * Seward et al. [1990] Wild et aL [19791 This behaviour was observed irrespective of buffer power, although a higher buffer power maintained a greater concentration between competing roots. There is very little evidence which shows that such rapid depletions occur in practice, although the results of Jungk et al. in Figure I are consistent with this. If such rapid depletions are commonplace, then a new way of specifying plant demand, which does not depend on concentration at the root surface, 224 would be required. Leigh [1989] has reported that barley maintains a constant K concentration in its sap throughout vegetative growth at a level depending on K supply. Given an adequate supply, a ceiling K concentration was reached which was not exceeded even at excessively high levels of soil K. The identifica- tion of physiological ceilings of K concentration could form the basis of a new and simpler approach to characterising plant demand. 100" a (5hrs) 0- 0CaB- C a A B C D CD Demand h h low h I Supply h h high I I 0- 0 0.15 0.30 Distance from root, r(cm) Figure 2. Potassium concentration profiles following diffusion to I day old roots accord- ing to the model of Claassen et al [1986]. (See Table 3 for parameter values). 5. Critical K concentrations in soil solution for high-yielding crops Virtually all K reaches roots by diffusion, and the concentration difference between the bulk soil solution, and the root surface needed to sustain any given uptake rate can be calculated from equation (2). Thus given a rooting density of 3 cm/cm3, which is modest for winter wheat (Table 1), on a wet, light soil an uptake rate of 5.2 kg/ha/day (Table 1) could be sustained by diffu- sive supply to a zero sink by a concentration of 51 pM K (5.2/0.101, see Table 225 4). In contrast, 91 l.M would be necessary in a moist, medium soil and, at the other extreme, in a dry, heavy soil, where diffusion is more difficult, 208 IM K would be necessary. Potassium concentrations in field soils are in the range 100-1000 pM (Mengel et al [1969]; Gregory et al [1979]; Barraclough [1986]), values at the upper end of this range being found on light, sandy soils. Superficially it would appear that K transport is only likely to become a limiting factor on dry heavy soils. Lower concentrations are adequate on light soils, but these soils have low buffer powers and they may not be able to sustain even low concentrations for very long. Critical solution concentra- tions calculated in this way would be of greater value if they could be related to K,, to give (practical>) isotherms. Table 4. Potential supply rates of potassium to roots by diffusion L, Potential supply rate (kg/ha/day/tM) (cm/cm3 ) Wet soil Moist soil Dry soil Light ...... 1 .028 .021 .014 soil ...... 3 .101 .075 .050 5 .187 .141 .094 M edium ...... 1 .021 .016 .011 soil ...... 3 .075 .057 .038 5 .141 .106 .070 Heavy ...... 1 .014 .010 .007 soil ...... 3 .050 .038 .025 5 .094 .070 .047 The critical soil solution concentrations as calculated above are conserva- tive values in some respects in that the contribution of root hairs has not been included; the greater absorptive area they provide would reduce the K concen- tration required. The above analysis illustrates the inter-dependency of roots and concentrations for nutrient uptake, but it cannot provide unequivocal answers as to when transport limits uptake. The reason for this is that the 226 reasonable to assume that most if not all roots would be contributing more or less equally to uptake. Clarkson [1981] has shown, for barley and marrow, that old roots retain a significant uptake capacity for K. The problem of ex- hausted soil zones remains and this is most likely to occur in the later stages of crop growth. The effect of ageing roots and exhausted soil on uptake could be assessed by comparing plant uptake rates and root growth. If the whole root system is contributing to uptake there will be a good relationship between the total amount of root and the plant uptake rate at any given time, whereas a better relationship between root growth rate and uptake rate would indicate that new roots growing into fresh soil zones were primarily responsible for uptake. 6. Conclusions Many problems remain in the modelling of K uptake. In principle all processes contributing to uptake can be modelled and even the detailed temporal and spatial variations in soil and plant properties which occur during growth can be taken into account using modern computers. What is lacking are the detailed measurements needed to validate and drive such models. In practice, one-off measurements and averaging procedures are unavoidable especially under field conditions and this will undoubtedly contribute to the failure of models to predict uptake as opposed to their failure as a result of incorrect principles being used. A method of assessing soil buffer power is needed which accounts for all the K that is available during growth. It is also important that the relative mobility of K in solution and exchange phases and the response of root sys- tems to soil K levels are established. The rate at which K is depleted at the root surface has important implications for potential uptake rates and for the boundary conditions to be used in modelling. If very low concentrations are rapidly established then Michaelis-Menten kinetics no longer apply and a new way of specifying plant demand independent of concentration at the root surface will be needed. Our inability to predict root growth is a major weakness in the modelling of K uptake and, in view of the difficulties, empirical values will continue to be needed for the foreseeable future. Ignorance of the 228 Glass, A. D. M. and Siddiqi, Q.: The control of nutrient uptake rates in relation to the inor- ganic composition of plants. In: Advances in Plant Nutrition, Vol. 1, P B. Tinker and A. Luchli (eds.), pp. 103-147. Praeger, New York (1984) Goulding, K.: The availability of potassium in soils to crops as measured by its release to a calcium-saturated cation exchange resin. J. Agric. Sci., Camb. 103, 265-275 (1984) Gregory, P J., Crawford, D. V. and McGowan, M.: Nutrient relations of winter wheat. 2. Movement of nutrients to the root and their uptake. J. Agric. Sci., Camb. 93, 495-504 (1979) Gregory, P J., Shepherd, K. D. and Cooper R: Effects of fertilizer on root growth and water use of barley in northern Syria. J. agric. Sci., Camb. 103, 429-438 (1984) Ingestad. T. Relative addition rate and external concentration: driving variables used in plant nutrition research. Plant, Cell and Env. 443-453 (1982) Itoh, S. and Barber S. A.: A numerical solution of whole plant nutrient uptake for soil-root systems with root hairs. Plant and Soil 70, 403-413 (1983) Johnston, A. E. and Poulton, P R.: Yields on the exhaustion land and changes in the NPK content of the soils due to cropping and manuring, 1952-1975. Rothamsted Report for 1976, Part 2, pp. 53-85 (1977) Jungk, A., Claassen, N. and Kuchenbuch, R.: Potassium depletion of the soil-root interface in relation to soil parameters and root properties. Proc. IX Int. Plant Nut. Colloq. pp.250 - 25 5 (1982) Klepper, B., Belford, R. K. and Rickman, R. W.: Root and shoot development in winter wheat. Agron. J. 76, 117-122 (1984) Kuhlmann, H., Claassen, N. and Wehrmann. J.: A method for determining the K-uptake from subsoil by plants. Plant and Soil 83, 449-452 (1985) Kuhlmann, H. and Barraclough, P B.: Comparison between the seminal and nodal root systems of winter wheat in their activity for N and K uptake. Z. Pflanzenern. Bodenk. 150, 24-30 (1987) Leigh, R. A.: Potassium concentrations in whole plants and cells in relation to growth. Proc. 21st Coll. Int. Potash Inst. 117-126 (1989) Lungley D.R.: The growth of root systems - a numerical computer simulation model. Plant and Soil 38, 145-159 (1973) Mengel, K., Grimme, H. and Nemeth, K.: Potential and actual availability of plant nutrients in soils. Landw. Forsch. 23/1. Sonderh., 79-91 (1969) Mengel D. B. and Barber, S.A.: Rate of nutrient uptake per unit of corn root under field conditions. Agron. J. 66, 399-402 (1974) Ministry of Agriculture, Fisheries and Food: Fertilizer recommendations 1985-86. MAFF reference book 209, HMSO, London (1986) Mullins, G. L. and Edwards, J. H.: A comparison of two methods for measuring potassium influx kinetics by intact corn seedlings. J. Plant Nut. 12, 485-496 (1989) Nye, P H. and Tinker, PRB.: Solute Movement in the Soil-Root System. Blackwell, Oxford (1977) Nye, PH.: Diffusion of ions and uncharged solutes in soils and soil clays. Adv. Agron 31, 225-272 (1979) Seward, P, Barraclough. PRB. and Gregory, R J.: Modelling K uptake by wheat (Triticum aestivum) crops. Proc. XI Int. Plant Nut. Colloq. Wageningen, Netherlands, 1989 (in press) Silberbush, M. and Barber, S.A.: Sensitivity analysis of parameters used in simulating potassium uptake with a mechanistic mathematical model. Agron. J. 75, 851-854 (1983) Silberbush, M. and Barber, S. A.: Phosphorus and potassium uptake of field-grown soya- bean cultivars predicted by asimulation model. Soil Sci. Soc. Amer. J.48, 592-596 (1984) Sinclair, A. H.: Availability of potassium to ryegrass for Scottish soils. 1. Effects of intensive cropping on potassium parameters. J. Soil Sci. 30, 757-773 (1979) Staunton, S.: The self-diffusion of sodium in soil: factors affecting the surface mobility. J. Soil Sci. 37, 373-377 (1986) 229 Welbank, PIJ, Gibb, M.J., Taylor, RJ. and Williams, E. D.: Root growth of cereal crops. Rothamsted Experimental Station Report for 1973, Part 2, 26-66 (1974) Wild, A., Woodhouse, R. and Hopper,M. I: A comparison between uptake of potassium by plants from solutions of constant potassium concentration and during depletion. J.Exp. Bot. 30, 697-704 (1979) Wild, A.: Potassium. In: Russell's Soil Conditions and Plant Growth, lth edition, A. Wild (ed.), Longman, UK, pp. 743-760 (1988) Willigen de, R and van Noordwijk, M.: Roots, plant production and nutrient use efficiency. PhD Thesis, Wageningen (1987) Woodhouse, R.J, Wild, A. and Clement, C. R.: Rate of uptake of potassium by three crop species in relation to growth. J. Exp. Bot. 29, 885-894 (1978) 230 Leaf Analysis for Standardising Soil Analysis J. Decroux* Summary The object of soil analysis is to measure as precisely as possible the extent of nutrient reserves in the soil and the power of a soil to provide for the needs of crops. The ease with which this nutrient stock passes from the soil to the plant depends on the interactions between crop species and variety, pedoclimatic conditions and nutrient reserves. This is the concept of bio-availability. While there is a critical nutrient level in the plant below which the crop will respond to application of that nutrient, experience shows that in the field it is usually difficult to establish a close relationship between a value determined by soil or plant analysis and response (cf. potash experiments in northern France and the Paris basin and survey of 110 vineyard plots in the Bordeaux area). Interpretation of analytical data for predictive purposes can be made by taking other factors which affect bio-availability into consideration - petiole K/Mg ratio for vine or K/CEC in soil in relation to clay content, for which interpretation diagrams have been established. Maintenance of leaf contents of major nutrients above the critical level indicated by reference values makes it possible to adjust fertilizer application in relation to soil nutrient supply. This is where leaf analysis at regular intervals, year after year, on the same sample trees adequately chosen and representative of various portions of the orchard is so useful. 1. Introduction The objective of agronomists who use soil analysis is to measure the quantity of nutrients which can be effectively taken up by a crop during the course of growth and developmento (Juste [1989). < * J. Decroux, La Grande Paroisse, 132 route d'Espagne, F-31057 Toulouse-C~dex/France 231 held. NH 4 acetate gives a good measure of soluble plus exchangeable Ca and Mg, but, in calcareous or dolomitic soils it also dissolves some carbonates so that the extract contains a part of the nutrients not immediately accessible to plants. The same argument applies to assimilable phosphorus determina- tions. It has been shown that 2% citric acid very well extracts tricalcium phos- phate applied as fertilizer. Thus, as well as measuring as exactly as possible the capacity of a soil to satisfy the nutrient requirements of a crop, soil analysis also gives a picture of the cultural history of a soil. It is important that this should be taken into account in the interpretation of analytical results, in the choice of method and in the quest for new methods. 2. Plant availability The preferred source of nutrients to the plant root is the soil solution, but the root also has access to less easily available sources. Bio-availability ex- presses the aptitude of a nutrient to be delivered from whatever source in the soil to the plant but does not describe the phenomena involved in the process. It depends on the interaction between [crop-variety] × [climatic conditions]. Anything which changes soil solution concentration can modify the passage of a nutrient towards the root and consequently uptake by the plant. The amount of a nutrient taken up very much depends upon (Figure 1): " soil and climatic conditions, * culture: species, variety, agronomy. For most plants, the effect of root temperature on nutrient uptake is greater than its effect on growth. Soil temperature around the root depends on soil and climatic conditions but also on cultural factors like method of tillage, and also on the type of root system: depth and branching. Cooper [1973] placed crops in 4 groups in this respect (Figure 2): For P uptake, apple falls in group B and maize in group D. For potassium, tomato and barley fall in group D. Tromp [1980] demonstrated an interaction between the effect of tempera- ture around the root and intensity of light falling on the foliage. Dry, matter content is greatly reduced in low light intensity with low root temperature. Uptake of calcium and potassium is much affected by these two factors (Table 1). Table I. Uptake of potassium and calcium by whole apple tree. Ca/K ratio in shoots, rootstock M9 Light Root Increase in dry Uptake (mg/tree) intensity temperature matter (g/tree) K Ca Ca/K High 180C 28.1 (100) 416 (100) 410 (100) 0.99 8C 18.5 (66) 195 (47) 392 (96) 2.01 18'C0 15.7 (56) 309 (74) 301 (73) 0.97 8 C 9.6 (34) 154 (37) 331 (81) 2.15 232 Plant Soil biomass Combined conditions Addsoriedti Adobd --- Bio-availability -- X 0 !/Crop, . z variety Exchangeable, Soil analysis for element n Soil Combined ~Pedoclimatic conditions - a 3 dore 0 Bio-availability -- variety 2 Exchangeable, -soluble Soil analysis for element n+l Figure I. Plant availability of soil nutrients: showing the way a nutrient passes from the soil into the plant and indicating the place of soil analysis (after Juste 119881). 233 x vZ C o >o - - >0 [1973J). Cooper (from the plant levels in nutrient 3. Critical For each nutrient, there is a critical level (whether in the whole plant or in a particular plant part) below which application of that nutrient will induce an increase in growth and production (Figure 3). Under constant climatic and cultural conditions, growth and production are not affected by fertilizer if the content exceeds the critical value. Yield levelCritical Optimum Lowering Lowering of yield of yield due to due to deficiency excess Nutrient levels Reference values Figure 3. Effect of an essential nutrient on plant yield or growth. 234 It is important that the levels of all nutrients should reach the critical level if maximum yield is to be obtained. Such a maximum is measured either by total biomass production (forage crops) or by yield of the commercially valuable part of the crop which is not necessarily attained at maximum production of total biomass: for example cereal grain yield under hot condi- tions or grapes for quality wines. 4. Interpretation of soil analysis The harvest obtained by the farmer is far removed from soil analysis con- ducted at a given time which indicated the available reserves of a nutrient. This is quite evident in field experiments which aim to correlate soil nutrient levels with response to graded dressings of fertilizer. Julien [1989] working on potassium fertilizer experiments in northern France and the Paris Basin, suggested three types of response in relation to exchangeable K level: * an «investment fertilization> range where there is normally response to applied K, " a «maintenance fertilization range in which only the more K demanding crops respond to K fertilizer, * a range of uncertainty. He maintains that it is difficult to base the interpretation of soil analysis on results of a single, even pluriannual, experiment. But, on the other hand, to consider a large number of experiments at the same time introduces a good deal of variation so that it is just as difficult to obtain correlation between soil analysis and crop response. Responses recorded in experiments show a wide scatter (Figures 4a and 4b). Figures 4a and b show similar variation in exchangeable K content and that many experiments show negligible response to applied K, even when soil K content is low. Figure 4b shows greater maximum effects on yield (5007o) than 4a (200%6). On the other hand, the density of points near the abscissa is proportionately higher in 4a. A: shows the increase in yield due to a heavy (2) K dressing over normal (1) maintenance K dressing; R, being maintenance K dressing (1) yield and R2 being heavy K dressing (2) yield: A = R2-Rl X 100 R I B: shows the lowering of yield in the absence of K fertilizer (0) as compared with the heavy dressing: B =- X 100 R2 235 In the latter case, if the values of B are low (- 0), there is little risk in apply- ing no K fertilizer. Yield increase 40 in % 30 20 10 ot4 5 . Exchangeable ...... " K,0 (ppm) 100 200 300 400 Figure 4a. Effectof exchangeable K on yield increase resulting from an investment dress- ing. Lowering of yield in % 50 .. 40 30 20 10 5 Exchangeable K20 (ppm) 100 200 300 400 Figure 4b. Effect of exchangeable K on the lowering of yield in the absence of fertilizer. 236 5. Plant analysis and standardising soil analysis The wide scatter of results connecting soil nutrient content and crop yield has led many to abandon the attempt to correlate yield with soil analysis and to concentrate instead on plant nutrient content as a criterion. Various methods are used: nutrient content of whole plant (often used for nitrogen), nutrient content of a plant part most representative of the plant's ability for uptake: leaf, petiole ... for P, K, Ca, Mg and trace elements. This was the method used by Etourneaud and Loud [1984/ in a study of 110 vineyards in the Bordeaux region in trying to establish critical nutrient levels for K and Mg in vine nutrition. Another objective was to improve the interpretation of soil analysis. They proposed diagrams relating the K/CEC ratio in the soil to clay content. Taking together all the results, there was no significant correlation between plant K content at either flowering or fruit initiation and exchangeable K con- tent of the soil. The true availability of potassium was not indicated by soil analysis. Just as important as soil K content were climatic and cultural factors and variety. Deeper examination showed that for K and Mg: " Potassium uptake was related both to soil K content and clay content. There was a good relation between petiole K and K/CEC of surface and subsoil; " Magnesium uptake was determined by both K level (K-Mg antagonism) and by soil Mg content. Petiole Mg was more negatively correlated with K/CEC than positively correlated with Mg/CEC of surface and subsoil. For confirmation, the 110 results were plotted on an interpretation diagram connecting K/CEC and clay content (Figure 5). On the diagram, the results are represented by signs showing K/Mg in petiole: less than 2, K deficiency probable; 2-8 K and Mg nutrition normal; > 8 Mg deficiency likely. Generally there was a good relation between K/Mg and K/CEC. Where K/CEC is designated «high 25 out of 29 cases show K/Mg above 8. Where it was olow or overy low 14 out of 19 cases show K/Mg below 2. There were few exceptions, and some of these could be ex- plained by the effect of rootstock. This interpretation of soil analysis is derived from a general interpretation of petiole analysis. It represents some progress in the case of vineyards and more especially so in the case of quality wines where the aim is high quality rather than maximum yield. A similar approach can be used with advantage for tree fruits - apples and pears (Figure 6). Here, the traces indicate the satisfactory levels for leaf K content for surface and subsoil. The data were obtained for the same trees over several years using samples of 100 leaves from 25 trees. It was found that when clay content was low, it was exceptional to find K deficiency in the trees if exchangeable K exceeded 70 ppm. However K defi- ciency is more frequent on clay soils even at an exchangeable K content of 350 ppm which would be considered sufficient for other crops like cereals. 237 This type of interpretation diagram is especially useful on establishment of a new orchard. It enables the calculation of the basal dressing required to bring K soil supply up to the desired level and this can be incorporated to depth during preparation of the land. Experience has shown that deep in- corporation of the basal dressing is especially effective in establishment of the orchard (more rapid growth, earlier bearing, increased yield). It is much more effective than several placement dressings applied later. K/CEC 8l - K/Mg ratio in petiole o < 2 19 5 -" + 2-8 62 S. • 1 ." x >8 29 31 7" - High 1 - 36 = Mddoc 4t/ ,- *8 37- 80=Libournais 4- . 81-1 10= Entre-deux-Mers +, . : No=a. r m a l 43 . ,4 Very low 430 5 10 15 20 25 30 Clayl%) Figure 5. Location of the 110 plots on the relation between K/CEC and clay contenr and the K/Mg ratio in the petioles. 238 Exchangeable K20 (ppm) 500" 450- Soil 350.400- 300- 250- Satisfactory 200- L Subsoil 150- 100- 50. 0 5 10 15 20 25 30 35 40 45 50 Clay % Figure 6. Interpretation of the K needs of surface and subsoil related to clay content, with a view to providing sufficient potassium for fruit trees. 6. Foliar analysis for monitoring fertilization of perennial crops When the orchard comes into bearing, leaf analysis is the only method which can be used to indicate the nutrient status of the trees; soil analysis is a poor indicator of true nutrient availability to the tree because: " it is not known how great a volume of soil is explored by the roots; o there is little information on the effect of climate and cultural methods (irrigation method, cover crop, fertilizer placement). Only continuous monitoring is sufficient to direct the fertilizer programme. A single analysis whenever it is made is of limited use, since it reflects past history more than the present state of affairs and cannot therefore be an ade- quate guide to the next or later years. But regular analysis, year after year, presents a picture of how the nutritional status of the crop is developing under the fertilizer treatment given and suggests changes required. One essential is to follow exactly the same sampling procedure from one year to another. In order to facilitate comprehension, analytical results are transformed into «compensated indices) related to a base of 100, following Kenworthy's [19731 method (Table 2). These are entered on a form which as well as showing current values for nutrient content and corresponding index, shows past history in graphical form with details of past fertilizer treatment. This gives at a glance the nutritional history of the orchard: 239 " Variations due to climate are separated from those due to fertilizer treat- ment, " Inter-nutrient interactions are clearly shown, * Such a presentation demonstrates the operation of the law of the minimum and compensatory phenomena in tree nutrition. Table 2. Form of report on leaf analysis SADEF Result of leaf analysis Laboratoire d'analyses agricoles Sample No: LS23-3 Lestenaque Orchard Sampling date: 00/07/89 Lamouzie Saint Martin Arrival date in laboratory: 28/07/89 24130 L Force Date of printing the form: 10/11/89 Laboratory No. 40703 Planting year: 77 Plot No.: Bloc 23 Apple tree Smoothee Rootstock: INRA MM 106 Trees/ha: 666 Part analysed: whole leaf Sampling date: 105 days after full flowering (F2) Target yield (t/ha): 60 Recommended fertilization Result of anaylsis evel Kenworthy ind. Very low Low atis- High factory 1000 leaves weight (g) 322 84 in 07 of DM: Nitrogen N 2.43 99 N Phosphorus P 0.15 87 P Potassium K 1.27 81 K Calcium Ca 1.64 114 Ca Magnesium Mg 0.3 109 Mg Sulphur S 0.17 K/Mg 4,2 in mg/kg DM: Boron B 40.1 94 B Manganese Mn 104 113 Mn Zn Zn 30.2 92 Zn Iron Fe 113 Copper Cu 24.9 101 Cu Satisfactory level 83-117 50 83 117 150 240 Evolution of levels N(.) P K Ca (x) Mg(*) Year Low Isatisiac. High Low Satisfac. High Low Sarislac. High Low Satisfac. High 1982 1983 1984 1985 1986 > 1987 1988 1989 x=leaf weight Dressing (kg/ha) Yield N P2 05 K2 0 MgO CaO (t/ha) 145 46 70 64 135 46 70 54 130 80 51 124 100 64 129 70 80 68 130 80 100 65 171 80 147 66 102 52 138 120 300 A summary of 17 years foliar analysis on apples and pears and 8 years on peaches has justified the idea of establishing permanent reference values unique to each species which correspond to the best obtainable yields (Table 3). These long-term measurements have been made in a large number of or- chards representative of diverse cultural conditions but giving satisfactory yields. They show stability in development and principal leaf nutrient contents for high level production. The variability around these values is quite accept- able, standard errors being of the order of 8% for nitrogen, a remarkably low figure, to 30% for potassium which is usually the most variable nutrient. Variability in trace element content is always large, reflecting the use of spray chemicals or foliar fertilizers by the grower. Mean leaf weight reflects the capacity of the tree to intercept solar energy and, hence, fruit production. There is a close inverse relationship between fruit crop and leaf development for a given set of cultural conditions. The value of foliar diagnosis has been verified by the statistical evaluation of reports received from 286 commercial peach growers in S.E. France (1975-1982). There was a close connexion between yield and manuring with an opposing relationship between leaf N and P contents and weight per 1000 leaves. There was no direct relation between fertilizer applied and leaf nutrient contents. 241 Table 3. Standard leaf nutrient contents for apple, pear and peach Regions S.W. France S.E. France Crops Apples Pears Peaches >7 years old Measures Reference value CV Reference value CV Reference value CV (%) (0) (%) 1000 leaves weight (g) 400 18 340 16 210 14 N (%) 2.45 8 2.10 11 3.70 14 P (%) 0.18 20 0.16 10 0.27 18 K (%) 1.70 24 1.65 23 3.20 12 Ca (%) 1.40 17 2.00 19 2.15 20 Mg (%) 0.27 21 0.36 21 0.41 12 Mn (ppm) 69 75 93 140 72 64 Cu (ppm) 24 - 24 - - - Zn (pprn) 39 65 43 64 68 26 B (ppm) 46 57 36 24 40 16 Three diagrams (Figures 7, 8 and 9) have been established to group yields and corresponding leaf nutrient contents in relation to the amounts of fer- tilizer applied. (For N and P 20 5 the steps from class to class are of 15 units and for potassium 30 units). Three observations can be made: I. There is a strong correlation between yield and nitrogen fertilization. 2. There is a relation between PK fertilizer application and yield, but it is looser, especially for phosphorus: there is a good deal of fluctuation. 3. Similarly, leaf nutrient contents vary little with fertilizer level: they are around 3.75% for N, 0.27% for P and 3.100 for K. This last observation indicates that the trees are well able to obtain the nutrients necessary to achieve their yield potential according to the seasonal conditions. Mineralisation of organic matter in these orchard soils is always insufficient to assure the nitrogen needs of the trees. The growers are able to use leaf analysis to adjust N fertilization in accordance with yield. Whatever the yield level, leaf N content is maintained above the critical level of 3.70%. In practice, it is usually possible to slightly reduce the level of N fertilizer without bad consequences in yield. If the relation of PK fertilization and yield is less close, this expresses the contribution to supplies of these nutrients from the soil and this varies a good deal from orchard to orchard. Phosphate and potash fertilizer needs depend much on soil P and K status and annual monitoring by leaf analysis shows 242 how conditions are developing and suggests what modifications of the fer- tilizer programme may be needed. Figures 8 and 9 show that leaf P content is maintained at about the critical reference value of 0.27% while K content is always lower than the critical reference value of 3.2%. These observations show that fruit growers tend to treat their orchards more generously with phosphate than with potash. On average: P fertilizer used = 4 and K fertilizer used =1.3 Crop P removal Crop K removal There is still room for progress in the matter of adjusting P and K fertiliza- tion to fruit tree requirements, though present practice is not operating to the detriment of fruit production. More rigorous use of foliar diagnosis car- ried out at regular intervals would enable fertilizer dressings to be better ad- justed to yielding capacity of orchards and designed to sustain yield and fruit quality. Yield It/ha) Leaf N (%) Leaf N "- 40J ..... ~, 14 • -' , ",. - " % ~- 3.7 20- -2.0 100 200 300 Kg N/ha Figure 7. Yield and leaf nitrogen content in relation to nitrogen fertilization. 243 Yield t/ha) Leaf P (%) 30 ].0.30 30- Leaf P I % 0.27 25 %.V.1 25 -0.2 15 Yield 10- 5 20 60 100 140 180 kg P20./ha Figure 8. Yield and leaf phosphorus content in relation to phosphorus fertilization. Yield t/ha) Leaf K (%) I N3.2 3 0 - - ," v '. --.. ,,9.''-- . 3 3.0 Leaf K 25- 20- -2.0 15- 10. 5 100 200 300 400 kg KO/ha Figure 9. Yield and leaf potassium content in relation to potassium fertilization. 244 7. References Boulay, H., Decroux, J. and Diris, J: Pratique de la fertilisation raisonne. Pommiers et Poiriers - SCPA/AZF, 1984 Cooper, A.J.: Root temperature and plant growth. Res. Rev. 4, CAB 31-43 (1973) Etourneaud, F and Loud, A.: Le diagnostic p~tiolaire de la vigne en relation avec l'interpr6- tation de l'analyse de sol pour le potassium et le magnesium. VI' colloque international pour l'optimisation de la nutrition des plantes, Vol. 1, 189-198 (1984) Julien, . L.: Dtermination des normes d'interpr~tation d'analyse de terre en vue de la fertilisation potassique. Science du sol 27, 2, 131-144 (1989) Juste, C.: Mobilit et biodisponibilite des oligo - 616ments - Les deuxikmes journes de I'analyse de terre. Ed. Frontires, B.P. 33, 91192 Gif-sur-Yvette, 1989 Kenworthy, A.: Leaf analysis as an aid in fertilizing orchards. Soil testing and plant analysis, S.S.S.A., Wisconsin, USA 381-392, 1973 Martin-Prdvel, P., Gagnard, J. and Gautier, P: L'analyse v6g6tale dans le contrOle de I'alimentation des plantes tempirdes et tropicales. Lavoisier, Techn. Et Doc., Paris, 1984 Ramon, J: Le diagnostic foliaire du bld. VI' colloque international pour l'optimisation de la nutrition des plantes. Vol. 2, 473-482 (1984) Tromp, J: Mineral absorption and distribution in young apple trees under various environ- mental conditions. Mineral nutrition of fruit trees, 173-182, 1980 245 Tissue Protein as Indicator for the K + Nutritional Status of Plants Y M. Heimer, A. Golan-Goldhirsh and H. Lips* Summary Potassium requirement for normal growth of plants and maximum yield stems from its involvement in plant metabolism. Consequently, early detection of potassium deficiencies has been a great concern of farmers. The working hypothesis of the present work assumed that potassium deficiency constitutes a stress like other environmental stresses to which the plant needs to respond and adjust by changing gene expression. This is manifested as alteration in protein profiles. Potassium deficiency was associated with increased abundance of two polypeptides of molecular mass of 60 and 62KDa in hydroponically grown sunflower plants. These poly- peptides were detected in young subapical region of the stem and in main veins + petioles of young leaves. The increased abundance could be detected at least as early as three days after the transfer of plants from potassium containing medium (5 mol • m - 3) to medium containing suboptimal concentrations (I mol •m- 3 or less). Deficiencies of other nutrients such as nitrogen, magnesium and phosphorus, or increased salinity of the growth medium were not associated with these protein markers. It is proposed that these polypeptides could be the basis for early diagnosis of potassium deficiency. 1. Introduction The plant can be described as a machine generating energy to fulfil two over- lapping purposes, maintenance and growth. Plants rely on the continuous supply of nutritional factors on one hand and adequate environmental condi- tions on the other. However, environmental conditions and availability of nutrients are constantly changing. Only under very strictly controlled labora- tory setups, can ambient conditions be maintained constant for extended periods of time. Changes which cause deviation from the current homeostatic steady state affect maintenance and growth. Consequently plants adapt to changing conditions by alterations in gene expression. There is ample literature describing the requirement of potassium for nor- mal plant growth and maximum yield, which stems from its involvement in plant metabolism (Evans and Sorger [19661; Clarkson and Hanson [1980]; Munson [19851; Mengel and Kirkby [19871). Clearly, the nutritional require- Prof. Dr. Y M. Heimer, A. Golan-Goldhirsh and H. Lips, Center for Desert Agrobiol- ogy, The Jacob Blaustein Institute for Desert Research, Ben-Gurion University of the Negev, Sede Boker Campus, 84990, Israel. 247 ment for K4 is unlike that for most other nutrients. K4 was not shown to be a structural part of organic compounds or an essential and intrinsic co- factor of enzymes. Still, its concentration in the cytoplasm is probably main- tained at 100 mol -m - ' (Evans and Sorger [1966]; Clarkson and Hanson [19801). It is an abundant mineral which occurs in considerable amounts in most soils. However, the forms of potassium available for plants are often deficient in soil (Larson, Barnes and Runge [1985]) which leads to increased use of potassium fertilizers (Pretty and Stangel [1985]). Furthermore, plant species may differ considerably in exploiting soil K 4 . Thus, of two plant species growing on the same soil, one may suffer from K deficiency and the other not. Rye-grass (Loliumperenne) was shown to be a strong competitor to red clover (Trifolium pratense) under conditions of low potassium in the soil (Mengel and Steffens [1985]). This was due to higher root length of the former. Recommendations for K * fertilization are generally based on soil analysis. For many soils however, this procedure is not adequate since the potassium forms in soil contributing to plant nutrition cannot be distinguished by this general test. This is particularly true for soils rich in K + -bearing primary minerals (Sparks [19861) and for soils rich in interlayer K+ (Mengel and Wiechens [1979]; Feigenbaum et at. [1981]). There are currently two tests for the assessment of potassium status which are based on plant analysis as alter- natives to soil test. One is the analysis of total potassium content found in plant tissues (Bergmann [19881). This test does not distinguish between the contents of various cell and tissue compartments such as the apoplast (cell wall), and symplast. There is indeed some evidence that it is the potassium concentration in the cytoplasm which controls growth (Leigh and Wyn Jones [1986]). The other test is based on the activity of the enzyme pyruvate kinase which is correlated with the K+ nutritional status of the plant (Besford [19781). This procedure has not gained wide application and its implementa- tion meets with difficulties especially vis A vis its specificity and sensitivity. Thus, there is still a need for a reliable, accurate, sensitive, early and easy test for potassium status (Munson [1985/). We contend that mineral deficiencies like other environmental extremes constitute stresses to which the plant needs to respond and adjust. Most stresses have been linked to the activation or inactivation of specific genes resulting in changes in protein profiles (Bewley, Larsen and Papp [1983]; Jack- son et at [1985]; Sachs and Ho [19861; Fritsmeier et at [19871; Heuss-La Rosa, Mayer and Cherry [1987]; Mansfield and Key [19871; Mohapatra, Poole and Dhinsda [1987]; Ramagopal [1987]; Vartanian, Damerval and de Vienne [1987]; Bohnert et at [1988/). The polypeptides which appear in anaerobiosis (Sachs and Ho [1986]) are enzymes of the glycolytic pathway; exposure of plants to heavy metals triggers the synthesis of metal binding proteins (Jack- son et at [1985]); infections with pathogens cause the elevation of phytoalexin biosynthetic enzymes (Fritsmeieret at. [19871; and salt stress causes the activa- tion of several genes associated with carbohydrate and amino-acid metabolism in halophytes (Bohnert et al. [1988]). Under conditions of phos- 248 phate Ihnitatnons increased secretion of acid phosphatase by roots was reported for seseral plant species /Goltdsten, Baert'n cd Datt [1989]). Fhis enzyme is believed to be responsible or utilization of organic phosphate asailable in the arbient Nit rogen star aion cansed the secret iot i a unique protein by tomato roots (Go/dswi, IBaeri/ein and Danon 119891). With respect to polassium nutrition. except for a few recent reports. starva- tion metabolism has not been examined at the molecular level. (hanges in The phosphorviation of membrane proteins of barley roots grown under potas- sinm limitation as xsell as changes in the pro filC of soluble proteins ,wec reported (Kuipa et al. /1988,, ernando et /1988/)./t. kt hae observed recently (Golan-Goldhirsh, Fieimer and Iips /1990/) chanles ittprotcin pro fiCs of sunflower plants exposed to potassi un depri ,ation. We assumed that the extent of the response and the adjustmnent to the en- vironmental conditions is proportional to the magnitude of the deviation from the current condition. Thus, in order to detect the changes it was neces sarv to amplify it by employing extreme deficiencies. kno" nu thth ventually they inms- be detrimental. We iepoit here onr observations on the cihanitgces in protein profiles in sunflower associated ev ith potassiur deficient.. 2. Materials and methods Sunflower (tielianhiU anmius 1. c. SaItola) plants wee grmwn hvdroponi- cally on long Ashton nutrient solution (Htewitt [1952]) %kithvarious concen- trations of K ' as indicated. Plants were also gron inmedia deficient or one or other of the fkllowtng minerals: nitrogen, phosphorus or magnesium. The effect of 75 tool in Na(I in the medium on protein protiles 'Aas AsotCsted Plams "ere grovn under 12 h light (300 l m , 5I at 23 ( and 12 Ih darkness at 184Cs ih contiluous acrtior. At the selected tnes plant parts "sere quick y removed, ,seiglied, placed itliquid niturogen and stored at -70, (. After rapid freeiing in liquid nitrogen sanpes were grond to a fe p-oxer withl mortar and pestle. [ he poder eawashed once wit Ii cold ( 2) ( ) 8M1 accone and 4 times with cold (20 () 00% acetone and air dried at roor temperature. Protci s w e extracted rom the dried powder ,itih IX Laetninli sample buffeftrletn/i [1970/i containing 2 iol rn I' t21 [A, 5 nioll - in ascorbate, 5 mol in 'NaliSo ,mmo i n m leupeptin and 2 mmol -m phetrvrchhsulton lfuoride. Sodium dodeolvulau pohaoryllld& gd electrophoresis (SDS- [M(tIJ) ,as carried out according to aernm li(laenni /197Zf) on 10% posacrv antidc gel. 4lie same atno u sci proreh ,;ere apilied to each lane (50 ilicrograms). 'clxpeptide hands ysere %isttaliicd on the gel by staining x it (ootnia"sie Brilliam Blue es. 5oein "as denernitned hi the Iolin phenol reagen t toi; et at 11951/h. 249 3. Results and discussion Sunflower plant, itl, forbI an e'ttlcd period o1 little wivithout potlttlstLt, S'1.lety oeIu educed grot! h, reduced leaic size and necrotic Ic 'ions tthich a ppe!at till on the c t tiutleeavsl (U.latUGl,dhirsh, tlt?7r andc! L..n /1990/j. I hc qesition wit ch then atose kih ispet to detectioi of a posiblI J)IIlCmII Ilari l 'erwas ICther il addilio to tilh qunltllitalive Cfect potassiutit dclty cotx htid o sintloxer, IheTC waa q ualitati eCffeclt on the protCin COil- postli as woell. I ' ptik in protiles ot ext cts obliiltd tom plant groxN I tn 5 motl nl K ant io pi s, transfere"d to 0 K ' and harvested at the time of appear- Acll' ofth visual symploIns of delicicatC\ 14 21 da\s) are hot it) Figlre . a I) 62 6--_ = K DEF P ~l* qw 54 aw / 650 52 1 3 o 6-2" -1 MWV L V S L V S R MW I K iK K .K I i' I. Souhitosl cv~tt~tthilelpl vtpLti cnIcer 2 d\,d layh tt ertIla\ ,c Ivd "hen the c lUAtu srptt (nlio- - t l l I ol K deliti'l c \Nere olbsI ed the t tlle'l ltlit' tl ot "lcleaip tied Io' I lane, ic I p kiwt\p d'tl It-hici-l inccaed ii abuti daic !in orgall" 01 dli- acil plat;I I pokyxpltide yth i dec'lcaCid in thundance K 1)11 desie ISO Several bands increased in intensity while others decreased in extracts riom potassium deficient plants as explained in the legend. The most pronounced changes in i he aerial parts were the increase in the inensi> of two polpep- tides with molecular mass of 60 and 62KDa, as well as decrease in the in ensitv of two polypeptides with molecular mass of 52 and 54KDa in plants trans- ferred to 0 K - . These changes could be clearlIN observed in young subapicai eLions of stems,. The increased abundance ol the 60 and 62K Ia polvpeptides w,hich w'e putatively named K-D I-P's for K deficiency proteins, could also be observed in blades and petidoles , main Weins of voung leaves (ffigure la). Changes in the protein profiles were also noticed in root extracts where the most pronounced was the decrease in abundance of the 50KDa polkpeplide in K deficient plan s (Figure Ib). The 60 and 62Kl)a pol peptides increased also in hypocot Iys of K deficient seedlings grown in the dark (Figure 2). - -975 6o/ - -45 --- 31 A sk -o-215 K K MW fJ ~ic"2 [!c p e l e Ih plfl l,l 'vK [)I iPm 1h"j)' , o~r- m 1 dol dIC g1flilldr~ {! Kricl~ >5dltq~lcIvkt 251 ]ie potential sc of he K- DI I ', a malkers Lit K dcfllctIckrlsite, [Ici carl> dctccIablIt t aitte WI I c orectie mcea ure , can 1k emploced cI ectixci' inhI ti Icd Ihie icase in abundanec of these Ivo polv\pcplide, t, t tuctOmilat the Loniz) itlik (itK ' i! thC 1lrient &OUlItOnj, and [he youith o1 me of feto\th it K dei icet i cdia, unti i/ed in Figuc 3. I le jliac , ot tht K F)0II', wis apparmnt lt least as carl> a, 3 las after lhit tia itt]i o pott itl[ 'Lufliciegi plats ito K df[li'CII con0di[ion. 1'his 5' ,s o ne b 't 'te h. y~pearanc of" the k nown m1I~o rp ho.logical] d e.fii ecyklCsvnrlp 0 nIIswhrich tl'ialkl bcomeo cxidenlt ol the I4t h day or iate! after the trais etr. f lie i en ittide of [heui tr eae hin abunda nc of the K I)- P's. as judged fit \ int e\'alllaltoltiof the pr o iles ater Mtlining the SI)S eel, with ( oomas- sic Brilliant BLe was simlilart M ill iihoptial concentrations oi K used Mld it! ll ltllts Qsatllined I*- Da,', kiler ri el ---- M\VV A 0 24 3 b 10 13 24 . . ... 62 P 'K 5 \602 V aE TUi"o r L}1,I 'iIl ,CI Mit, td6KI ltoh'p ilokp lHd {} +lick, 210"' L t'd ci ' ' li{do m klll %ii ok\iv }Il[fi l Ia ll, '<( i fO{ I hV ICSJI MZHItS]h'k aic jifj I,, ill )C tjj iI, ilhk ipimm l ,tl*{vd~q I4Jl\ *I ~ hon \ ht on];l 'ki [I II M (J 11 1 litd K1 1, 0p l i l o te,1 1 ' - k The specificity of the markers with respect to nitrogen, magnesium and phosphorus deficiencies as well as increased salinity was also examined. The increased abundance of the 60 and 62KDa polypeptides seemed to be ap- parent only in potassium deficiency (data not shown). The biological significance of the K-DEF-P's is not at all clear at this stage. They may reflect impairment of normal development of various organs, mainly stem, brought about by various stresses including K 'deficiency. On the other hand, they may be related more directly to K '-mediated or K -de- pendent metabolism and acquisition. The increased abundance of K-DEF-P's may indicate a response of the cells to scavenge any K + ion available in the ambient, or maintain the highest endogenous concentration of potassium possible under conditions of low external availability of the cation. In this context the K-DEF-P may be related to K + transport or binding. It has been shown recently, that withdrawal of K caused rapid modification in the phos- phorylation of a plasma membrane associated polypeptide and changes in the profile of soluble proteins (Kulpa et al. [1988]; Fernando et al. [19881). K + is known to be translocated between various organs and tissues of the plant (Evans and Sorger [1966]; Ben-Zioni, Vaadia and Lips [1971]; Lips et al. [1987]; Clarkson and Hanson [1980]; Touraine, Grignon and Grignon [19881). Thus, the K-DEF-P's could increase in response to K * status within the responding organ directly, or be triggered by a signal transmitted from a different organ. It is tempting to hypothesize that the signal for increased abundance of these polypeptides is the reduction in potassium influx. The early and easy detection of the increased abundance of the K-DEF-P's at all suboptimal concentrations of K + tested, makes them favorable candi- dates as markers for diagnosis of K + deficiency in plants. Purification of the 60 and 62KDa polypeptides is under way with the objective of raising antibodies. The antibodies will be used for accurate quantification of the K-DEF-P's. A detailed study into their specificity with respect to other nutri- tional imbalances and other environmental stresses and universality with respect to other plant species, is in progress. Acknowledgement We acknowledge financial support from the Dr. Herman Kessel, Applied Biology Research Fund in the memory of Mr. J.J. van Ransbury. 4. References Ben-Zioni, A., Vaadia, Y and Lips, H.S.: Nitrate uptake by roots as regulated by nitrate reduction products of the shoot. Physiol. Plant. 24, 288-290 (1971) Bergmann, W.: Ernahrungsstdrungen bei Kuhurpflanzen, Entstehung, Visuelle und Analy- tische Diagnose. VEB G. Fischer-Verlag, Jena (1988) Besford, R. T: Use of pyruvate kinase activity of leaf extract for the quantitative assessment of potassium and magnesium status of plants. Ann. Bot. 42, 317-324 (1978) 253 Bewley. £D., Larsen, J D. and Papp, J. 7 Water stress induced changes in the pattern of protein synthesis in maize seedlings mesocotyls: a comparison with the effect of heat shock. J. Exp. Bot. 34, 1126-1133 (1983) Bohnert, H. J, Ostrem, .A.. Cushman, J. C, Michalowski, C. B., Rickers, ., Meyer, G., DeRocher, E. J, Vernon, D. M., Krueger, M., Vazquez-Moreno, L., Velten, J, Hoefner, R. and Schmitt, J M.: Mesembryanthemum chystalinum, a higher plant model for the study of environmentally induced changes in gene expression. Plant Molec. Biolog. Reports 6, 10-28 (1988) Clarkson, D. T and Hanson, J B.: The mineral nutrition of higher plants. Ann. Rev. Plant Physiol. 31, 239-298 (1980) Evans, H. J and Sorger G. : Role of mineral elements with emphasis on the univalent cations. Ann. Rev. Plant Physiol. 17, 47-76 (1980) Feigenbaum, S., Edelstein, R. and Shainberg, 1: Release rate of potassium and structural cations from micas to ion exchangers in dilute solutions. Soil. Sci. Soc. Amer. J. 45, 501-506 (1981) Eernando, M.. Siddiqi, M. Y., Kulpa, J and Glass. A. D. M.: Changes in soluble proteins in barley roots associated with K + deprivation. Plant Physiol. 86 (supplement), 491 (1988) Fritsmeier, K. H., Kretin, C, Kombrink, E., Rohawer, F, Taylor, ., Schell, D. and Hahl- brock, K.: Transient induction of phenylalanine-ammonia lyase and 4-coumarate: COA Ligase mRNAs in potato leaves infected with virulent or avirulent races of Phytophthora infestans. Plant Physiol. 85. 34-41 (1987) Golan-Goldhirsh, A.. Heimer, . M. and Lips, S. H.: Effect of potassium on growth and the electrophoretic pattern of leaf proteins of sunflower supplied with ammonium or nitrate (submitted) (1990) Goldstein, A. H., Baertlein, D. A. and Danon, A.: Phosphate starvation as an experimental system for molecular analysis. Plant Mol. Biol. Report. 7, 7-16 (1989) Heuss-La Rosa, K., Mayer, R. R. and Cherry, J H.: Synthesis of only two heat shock pro- teins is required for thermoadaptation in cultured cowpea cells. Plant Physiol. 85, 4-7 (1987) Hewitt, E. J: Sand and water culture methods used in the study of plant nutrition. Com- monw. Bur. Hortic. Plant Crops (GB), Tech. Commun., 22 East Mailing, Kent (1952) Jackson, R J., Naranjo, C. M.. McClure, P R. and Roth, E. J: The molecular response of cadmium resistant Datura innoxia cells to heavy metal stress. In: J. L. Key and T Kosuge (eds.), Cellular and Molecular Biology of Plant. AR Liss, New York, pp. 145-160 (1985) Kulpa, J. Fernando, M., Siddiqi, M. Y. and Glass, A. D.M.: Regulation of K + influx may involve phosphorylation of membrane proteins. Plant Physiol. 86 supplement), 477 (1988) Laemmli, U. K.: Cleavage of structural protein during the assembly of the head of bacteri- ophage T4. Nature 227, 680-685 (1970) Larson, W.E., Barnes, R. F and Runge, E. C. A.: In: R. D. Munson (ed.), Potassium in Agriculture. Amer. Soc. Agronomy, Crop Sci. Soc. Amer. and Soil Sci. Amer. p xvii (1985) Leigh, R. H. and Wyn Jones, R.G.: Cellular compartmentation in plant nutrition: The selective cytopolasm and promiscuous vacuole. Adv. Plant Nutrition 2. 249-279 (1986) Lips, S. H., Soares, M. L M., Kaiser, . and Lewis, O. A. M.: K + modulation of nitrogen uptake and assimilation in plants. In: Inorganic Nitrogen Metabolism (eds. W R. Ull- rich, P. Aparicio, P . Syrett and E Castillo) pp. 233-239. Springer Verlag (1987) Lowry, 0. H., Rosebrough, N., Farr, A. and Randall, R.: Protein measurement with Folin phenol reagent. Jour. Biol. Chem. 193, 265-275 (1951) Mansfield M.A. and Key, J. L.: Synthesis of low molecular weight heat shock proteins in plants. Plant Physiol. 84, 1001-1006 (1987) 254 Mengel, K. andKirkby, E. A.: Principles of Plant Nutrition. International Potash Institute, Bern (1987) Mengel, K. andSteffens, D.: Potassium uptake of rye-grass (Loliumperenne) and red clover (Trifolium pratense) as related to root parameters. Biol. Fert. Soils 1, 53-58 (1985) Mengel, K. and Wiechens, B.: Die Bedeutung der nicht austauschbaren Kaliumfraktion des Bodens fur die Ertragsbildung von Weidegras. Z. Pflanzenern. Bodenkunde 142, 836-847 (1979) Mohapatra, S. S., Poole, R. J. andDhinsda, R. S.: Changes in protein patterns and translata- ble messenger RNA population during cold acclimation of alfalfa. Plant Physiol. 84, 1172-1176 (1987) Munson, R.D. (ed.): Potassium in Agriculture. Amer. Soc. Agronomy, Crop Sci. Soc. Amer. and Soil Sci. Amer. p. 1223 (1985) Pretty, K. M. and Stangel, PJ.: Current and future use of world potassium. In: R. D. Mun- son (ed.), Potassium in Agriculture. Amer. Soc. Agronomy, Crop. Sci. Soc. Amer. and Soil Sci. Soc. Amer. pp. 413-423 (1985) Ramagopal,S.: Differential mRNA transcription during salinity stress in barley. Proc. Natl. Acad. Sci., USA, 84, 94-98 (1987) Sachs, M. M. and Ho, D. T H.: Alteration of gene expression during environmental stress in plants. Ann. Rev. Plant Physiol. 37, 363-376 (1986) Sparks, D. L.: Potassium release from sandy soils. In: oNutrient balances and the need for potassium . Proc. 13th Int. Potash Inst. Congress, Bern, 93-105 (1986) Touraine,B. N., Grignon, N. and Grignon, C: Charge balance in N0 3-fed soybean. Estima- tion of K + and carboxylate recirculation. Plant Physiol. 88, 605-612 (1988) Varlanian, N., Danerval C. and de Vienne, D.: Drought induced changes in protein pat- terns of Brassica napus var. oleifera roots. Plant Physiol. 84, 989-992 (1987) 255 Coordinator's Report on the 3rd Working Session Prof. Dr. A. van Diest, Department of Soil Science and Plant Nutrition, Agricultural University, Dreijenplein 10, 6703 HB Wageningen, Netherlands; member of the Scientific Board of the International Potash Institute With the term «soil fertility> appearing in the title of this session, Johnston and Goulding apparently deemed it appropriate to present a definition of < 257 tween nutrients in soil solution and nutrients on soil colloids. Therefore a soil extractant that can be expected to remove nutrients from a solution shortly after they appeared in the solution, is likely to be useful as universal soil extracting agent for soil testing purposes. The electro-ultrafiltration (EUF) technique can be mentioned in this respect. The advantage of this method is that, indeed, nutrients are removed from the aqueous medium shortly after they were released from the soil. In this way, the soil is then induced to release more nutrients to restore a disturbed equilibrium. Thus, an estimate can be obtained of the capacity of the soil to release nutrients from its solid constituents. Another advantage of the EUF method is the fact that no chemicals are needed for the extracting procedure and that therefore the environment is not polluted with chemical waste products. Disadvantages of the method are the high prices of the apparatus and of the filters to be used, and the extended period of time one soil sample occupies the machine. The number of samples that can be handled per day is therefore limited. The resin method can also be seen as a universal extracting method, especially when a combination of anion - and cation resins is employed. The resins are expensive, but since they can be used over and over again, the cost per soil sample is limited. Here again we observe a procedure in which nutrients are removed by the resins from the liquid phase soon after they were released by the solid phase. In this way the action of the resins simulates the action of roots, and this close simulation seems to justify the assumption that the resin method can serve well as a universal extraction method for soil testing purposes. The salts used for removing the nutrients from the resin, namely NaCI or NaHCO 3, are relatively cheap and have no serious polluting effect, environmentally speaking. It must however be realized, as was pointed out in Wessolek's presentation, that crop species may differ widely in root density, and therefore also in their ability to withdraw potentially available nutrients from a soil. It was, however, pointed out clearly in Barraclough's paper that for the time being it is still difficult to identify this quantity of potentially available K or, in other words, the soil buffering power with respect to K. Likewise, difficulties are met in measuring and predicting the growth and morphology of roots. At present, only simplified diffusion models are available for predict- ing the potassium concentration needed for high crop yields. Uptake of nutrients is affected not only by soil factors, but also by climatic factors, such as soil temperature and light intensity. It was emphasized by Decroux in his paper that in fruit production soil analysis alone is not enough to guarantee optimal nutrition of fruit trees. The earlier mentioned climatic conditions can exert strong influences on the nutritional status of the trees, and such influences can best be monitored through leaf analysis. With regular foliar analysis there is a possibility to examine the efficiency of established fertilizer application policies, and to reevaluate such policies when needed. 258 One of the difficulties associated with leaf analysis is that often the results of the analyses become available to a grower too late to enable him to take corrective actions. It was therefore interesting to note that Heimer and coworkers could show that an early appearance of certain polypeptides in plants can be viewed as an early warning system for an impending K deficiency in these plants. If indeed growers would be able to detect such a deficiency themselves at an early stage of plant growth, it would leave them enough time to rectify the situation. In such a case, the crop would have ample time to respond to the rectification, so that the eventual yield would not be much affected. If such techniques could become available to farmers and growers, not only for K but for other nutrients as well, I think important progress would have been made in safeguarding the optimal nutrition of our crops. 259 Chairman of the 4th Session Dr. Ch. Pieri, Sous-Directeur des Ressources Naturelles, CIRAD/IRAT, BP 5035, 34032 Montpellier Cedex 1, France; member of the Scientific Board of the International Potash In- stitute 4th Session Evaluation of Field Experiments and Fertilizer Recommendations 261 Field Experiments for Fertilizer Advice - Design - Execution - Interpretation 0. Jourdan and P Villemin* Summary The aim of advice is to ensure that the farmer obtains a just return for his investment in fertilizer through improvement of yield and its stability and quality. Consideration of factors which modify fertilizer effects and their interactions, leads to discussion of ex- perimental procedures to measure these effects in the field in both short and long term work. Great importance is attached to careful planning of the experiment. In long term work which may involve a preparatory phase to establish different soil fertility levels, provision must be made for eventual splitting of plots to accommodate test treatments. It is also advisable, where possible, to conduct a uniformity trial on the experimental area. There is a need for great care in recording and cultural operations throughout the period of the trial. Information on crop performance should be supplemented by investiga- tion of effects on the soil, necessitating careful sampling for analysis at intervals. Matters of statistical significance and interpretation of results by mathematical modelling are con- sidered. 1. The aims of fertilizer advice The farmer invests in fertilizer. The agronomist investigates the profitability of this investment in the short, medium and long term. Fertilization raises crop production (yield) and improves quality of the produce. 1.1 Improvement of yield Generally speaking, potassic fertilizer increases total biomass production. In cereals, for example, it increases grain yield and also the amount of straw. This action is an expression of Liebig's law of the Minimum and there is frequently positive interaction between potassium and other factors limiting crop yield, notably nitrogen supply. * Dr. 0. Jourdan and Dr. P Villemin, Centre de Recherches SCPA, F-68700 Aspach-le-Bas, France 263 1.2 Quality improvement This may be expressed in: - Sugar content of beet (Villemin et al. [1990/) - Taste (sugar content and acidity) and keeping quality of fruits (Martin- Prdvel [1989]) - Milling quality in cereals by increasing flour content (1000 grain weight) - Oil content in rapeseed, sunflower and other oilseed crops. - Storage properties and commercial value of potatoes: decrease in internal blackening, higher proportion of marketable tubers (Gravoueille [1987]; Beringer [1989]). 1.3 Drought resistance Potassium is the principle agent in maintaining turgor and stomatal regulation responsible for water economy (Lindhauer [1985/). 1.4 Earliness Ripening of cereals is hastened, permitting earlier harvesting (Chevalier [1975/). 2. Plant response to potassium fertilizer 2.1 Plant requirements Figure I illustrates two aspects of K requirement of maize: - the total amount of K taken up (M) indicates total needs for K (some 200kg/ha K20 for maize), - the maximum daily uptake rate (F) indicates the rate at which the soil must make K available to the plant (8-12 kg/ha/day for maize). Other aspects, such as the plant's ability to extract potassium from the soil are important. A crop which can obtain sufficient K from a soil low in potas- sium and does not respond to K fertilizer is less demanding: this is the case with wheat, despite the fact that it has considerable total needs (over 300 kg/ha K2O at the time of maximum vegetative growth) and maximum rate of uptake (6-8 kg/ha/day). Intensity of the farming system also affects this aspect. For example, forage grasses are extremely demanding under intensive cutting management but less so under an extensive system. 264 M 200- o 150- Silking stage 50- May June July August september I Figure I. K uptake pattern in maize. 2.2 Analysis of crop response 2.2.1 Response pattern When yield is plotted against rate of fertilizer we can identify the rate of fer- tilizer corresponding with economic criteria such as maximum gross margin (Figure 2). Yield Maximurn-__ yield slight Economic gradient optimum Ij Cost/price ratio S eIo Fertilizer rate X opt X max Figure 2. Response curve. 265 2.2.2 Response factors 2.2.2.1 Soil The part played by soil K supply in plant nutrition is made clear from Figure 3 which shows that response is the greater the poorer the soil. Comparing curves I to 4, it is obvious that soil 4 can obtain all its requirements from the soil and the yields recorded in the absence of potassium fertilizer indicate the extent of K reserves in the soils (4 > 3 > 2 > 1) (Jourdan[19881). Another important point about these curves is that no matter how much potash fer- tilizer is applied on soils 1, 2 and 3, which have lower soil reserves of K, the yield will never attain the level recorded on soil 4. This shows the particular value of the residues accumulated in the soil as a result of previous potash fertilizer treatment. Yield (4) Fertilizer rate 0 E X Figure 3. K response on soils of differing K status (1) low (4) high. 2.2.22 Plant Crops having the highest demand for K show the greatest response to applied potassium. In the example of Figure 4 (Villemin et al. [1990]) the highest responses are given by sugar beet and the lowest (virtually nil), by wheat and barley. 2.2.2.3 Time Response varies between years in two ways: - differences in weather can change the response pattern as shown in Figure 5: " increase up to the highest rate as in 1984 and 1986; " increase up to the medium rate of potash, but no increase above that level in 1979, 1981 and 1983; " no response in 1980 and 1985 on account of other limiting factors. 266 - because soil K content is increased by the potash dressings applied while it falls on plots receiving no potash. The contrast between 1981 and 1983 soil analysis data is shown in Figure 6. Evidently, these changes have af- fected the response patterns of Figure 5. o- IKx - Ko (%) O Beet Kx Wheat, barley 70 - Oats x Lucerne 60. + Peas 50. 40- 30 X 20' 1968 1972 1976 1980 1984 Figure 4. St Jean sur Moivre experiment. It is an illusion to expect that a response curve will tell us everything about how a crop responds to increasing rates of K fertilizer. Though nutrition is of dominant importance in determining yield, there are a great many other factors which also affect it. These may not all be known, but the way in which response is affected by climatic conditions shows how important they can be. In experimentation to identify the optimum rate of fertilizer one must take note of such variations and it is useful to think in terms of the frequency of favourable results in relation to the number of cases where response is small or negligible. The agronomist will always aim to arrive at the situation exemplified by soil 4, where soil K reserves are high and yields good, but he has also to reckon with the cost of soil improvement (passing from soil I to soil 4). 267 Yield index 120- (K0 = 100) 1986 110- 1984 100- 90 80 I I I K0 K, K, 130 - 1983 120- 1979 110- 10 1981 10- I I I I o0 K, K, 1980 100- 90- 1985 (Triticale N limiting) 801 K0 K, K, Figure 5. Different types of response to K fertilizers by cereals (wheat, barley, triticale) (Jourdan 11988J). 268 30, Kx - K annual/mean Index (%) Kx 20- 1983 100 10. 0 .,, 1981 0 - . ..-----.-----.--...... ------.. Annual -30 K0 K, K, K3 K Figure 6. Change in response curve at St Benoit sur Seine (Villemin [1990]). 3. Choice of experimental procedure 3.1 General principles Spectacular response to fertilizer treatment on a poor soil is one thing, but there is more to it than that, especially in the case of potassium because of interactions with environmental factors: - Climate - Soil - Other nutrients - Plant. The results of one experiment can apply only to the conditions under which the experiment is conducted (given soil type, given region etc.). It is preferable to investigate one thing at a time rather than to search for the solution to several different problems at the same time. It is better for instance to inves- tigate the response to K fertilizer separately from studying the fate of applied potassium. 3.2 Choice of design The aim of the experiment and its duration are inseparable and choice of design must be based on this. 269 3.2.1 Short term experiments (1 or 2 years) The advantage of this type of experiment is that the results are not affected by the residual effects of fertilization. On the other hand, its applicability is limited because of the lack of inter-annual variation. To get a better appreci- ation of such seasonal effects, it is better to grow all the crops, which will be differently affected by climatic variation, in the same year. There are several categories of design possible: - Fundamental experiments to study a mechanism (e.g. movement of potas- sium within the plant) and, for this, precision is more important than representativity. The design is not simple and requires adequate replication. - Experiments to identify the economic value of a treatment where, again, accuracy is the main consideration. Here several sites should be used to obtain information on the basis of a geographical or climatic region or a soil type. Experiments should be simple; if they do not individually produce significant results, treating them as a group will enable more precise evaluation of the results. - Multifactorial experiments (e.g. the effects of N, P and K and their interac- tions). These can be accommodated on a minimum number of plots. 3.2.2 Long-term experiments Nutrients like P and K are less mobile than N and if treatments are applied year after year positive (or negative) balances accumulate in the soil; it is rare in the case of P or K for applications and removals to be exactly in balance. The nutrient balances to aim for should pivot around nil (applications balanc- ing removals) where fertilizer applied compensates for removals in harvested crops and drainage. If the balance is strongly positive the soil may be enriched after a few years to the extent that losses in reversion or leaching become high. If the balance is much in deficit, yields decline and the crop will exploit the soil to obtain its needs (Figure 6). Whatever the case may be, with fertilizer treatments repeated year after year, after quite a short time we are comparing cultural systems rather than only differing rates of K fertilizer. For this reason it is preferable to keep the treatments simple because interactions between potassium and other nutrients also increase with time. An experiment at Laval (Figure 7) illustrates these points. This was a 3Nx3Px3K, 27 plot factorial. Soil analysis after 20 years showed that there was a greater difference in exchangeable K content between treatments N3K3 and NIK 3 (69 vs 145 ppm) than between K, and K3 (35 vs 103 ppm), the reason being that the high yields under the high N treatment resulted in in- creased K removal by the crops. Soil K content under N3 K3 was similar to that under N, K2 though the K3 rate was designed to enrich the soil. Nitrogen had as much effect in determining the eventual K content of the soil as did K itself. 270 160.0' K20 content (ppm) 140.0- 120.0 100.0- 80.0. K 40.0" K, 20.0-• 0.0 N, N, N3 Figure 7. Soil K contents at the end of the Laval experiment. Long-term experiments, as well as giving information on the economics (comparison of gross margins on different treatments), yield information on the way in which treatment effects change with time. Example 1. Change in response to K by wheat at Ozoir le Breuil (Figure 8). The response index decreased with time probably indicating a lack of balance between K, Ca and Mg. 20.0%" 100 (K2-K) /K , 15.0%. -' 10.0%" - 5.0%" -5.0 %- N -10.0 % %- %'N,NN , -15.0 %. 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 Figure 8. Change in K response by wheat in the Ozoir leBreuil experiment. 271 Example 2. Change in K response at St Pourcain/Besbre (Figure 9). Response increased with time due to K impoverishment under treatment K. and enrichment under K2. Long-term experiments require less replication than short-term since repeti- tion over the years increases statistical precision. The NK interaction can even- tually be measured by regular rotation of the N treatments. 25.0 % 100 (K, -K ) / K, N, 20.0 % 15.0 %_ %Nj 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Figure 9. Change in KCresponse with time in the St Pour~ain/Besbre experiment. 3.2.3 Experiments with two phases - A preliminary phase where two or three treatments designed to impoverish or build up the soil and possibly to maintain soil fertility are applied over several years. - A test phase of short duration in which increasing rates of K are tested to show the effect of soil K level on response to fresh K (Figure 3). In this type of experiment, treatments in the first phase should be contrasting as regards physical (method of cultivation) chemical (nutrient balance) and economic (financial return resulting from enrichment/impoverishment) fac- tors. There is a great difference between potassium effects in the long term on tilled and uncultivated soils (Figure 10). Plots in the preliminary stage should be large enough to permit subsequent subdivision. There must also be sufficient replication so that inter-plot effects are not confounded with fertilizer treatment effects. Consequently costs are high. 272 Exch. K20 (ppm) 0 100 200 300 400 500 0- 16--A- 10- - 20- I . -- U----Tilled Ko 0 - I .- .--..--- Tilled K, n30 I I * Unti led K 40 -- e--Untilled K, 50 Figure 10. The effect of potassium application on tilled and untilled soils. 3.3 Some experimental designs Design must not be an end in itself; it has to meet the needs of the experiment (Tranchefort and Philippeau [1971]). The results must be susceptible of statistical evaluation and a choice has to be made between precision and cost. There has to be some compromise between the number of treatments under test and the degree of replication. 3.3.1 Short term experiments - Single factor experiments " Completef'y randomised " Randomised blocks - Multi-factor experiments " Randomised block * Split plot * Latin square * Confounded design 273 The last design can accommodate a large number of treatments on few plots but the reliability of results is limited by the lack of true replication and it gives limited information on interactions over time. 3.3.2 Long term experiments It is possible to include only a limited number of treatments and suitable lay- outs are complete randomisation or the randomised block. If it is wished to test nitrogen levels, it is well to rotate the N treatments in order to even out K removals. The N treatments will leave different residues but such differences are small relative to the direct effects of N on nutrient removals (cf. Laval expt. Figure 7). 3.3.3 Two phase experiments (see 3.2.3) The same kind of procedure as for long-term experiments must be followed in the preparatory phase but, in addition, the plot size must be greater so that there is room to apply the treatments needed in the test phase. 4. The siting of experiments 4.1 Precautions Environmental conditions are important in considering the plan of the experi- ment and its management. 4.11 Statistical aspects The degree of replication is governed by the complexity of treatments. It is necessary to have an idea, from previous experience, of the likely variability (plot error) in relation to the magnitude of the treatment differences expected. Often, such information is lacking and this obliges the experimenter to allow room for manoeuvre by, for example increasing the number of replications so that there is room for later modification in the light of early results. 4.1.2 Management Ploughing causes lateral displacement of the soil. To some extent this effect can be minimised by alternating the direction of ploughing annually. If ploughing is at right angles to the length of the plot, soil mixing will be minimised. Spreading of fertilizer requires consideration. * If fertilizer is applied before ploughing it is well incorporated but subject to lateral displacement; this is avoided by applying after ploughing. " The rate of application must be carefully checked. 274 * Spreading must often be by hand. There are a number of tricks to ensure even distribution. If it is at all possible it is preferable to spread by machine, thus avoiding human error. " The same precautions must be observed in spreading other (non-treatment) fertilizer dressings. Sowing should be done with care so that all plots are treated equally (same number of passes of the seed-drill, for example). The same applies to phytosanitary measures which must be uniform over the whole experiment, including discards. Irrigation, even when done with the greatest care can introduce hetero- geneity. It is essential that every treatment should represent all parts of the heterogeneity gradient. It should be possible to achieve this by arranging blocks perpendicular to the line of advance of irrigation equipment. The area of plot sampled for harvest has to allow for all crops of the rotation (width of seed-drill and harvesting equipment are determining factors here). Care must be observed to ensure that crop residues are restored to the ap- propriate plot, mixing across plot boundaries must be avoided. In the final analysis, a compromise must be reached as regards plot size. Allowance must be made for all the crops of the rotation in the knowledge that yield variability varies with crop - for instance sunflower is more variable than wheat. 4.1.3 Precautions in soil analysis If there is time several analyses should be conducted before laying down the experiment. For this five samples should be taken across any likely gradient in soil nutrient content. Both surface and subsoil should be sampled. 4.2 Establishment in the field 4.2.1 Soil sampling It is essential to have a full description before the experiment is started so that variation in chemical and physical properties of the soil is known. Sampling of soil and subsoil should be done for each plot and the samples should be stored moist since drying can affect the results of analysis. Non-experimental variants (e.g. mechanical analysis, CEC etc.) may be exa- mined block by block, while K content should be examined on the individual plot basis. Obviously, it is important to obtain the samples before any treat- ments are applied. Size of soil sample depends on the determinations required. 4.2.2 Uniformity trials To conduct a «dummy run ) is often thought to be a luxury but it does increase the reliability of results: an appreciation of natural variability in the field 275 which later appears as experimental error comprises overall variability indi- cated by yield but also some further detail shown up in yield components, variation in nutrient content of the crop etc. Generally speaking, examination of the results of a uniformity trial enables one to avoid erroneous conclusions in the experiment proper. Allowance can be made for variability indicated by the uniformity trial in comparison of the eventual treatment effects. Note 1: For safety's sake it is wise to include one or two extra blocks in the uniformity trial; then unusually variable areas can be excluded from the experiment proper. Note 2: Time spent in doing the uniformity trial may uncover some defect in the origi- nal conception of the experiment. The experimenter is given a year during which he has the opportunity in the light of experience on the site to modify his approach; once treatments have been applied change is impossible. 5. Results 5.1 Collecting the results Precautions must be taken not to increase the error inherent in recorded values. - by preserving objectivity (critical examination of any action by field staff) - by maximum precautions in recording information. Plot data must be en- tered immediately in the field in a standard form. It is often thought that the recording of information is a simple matter but perhaps a little tedious where there is great heterogeneity (in pasture experi- ments for example). The routine for recording at harvest should be tightly laid down by a competent person. 5.2 Handling of results Theoretically, the statistical treatment of results is decided before the experi- ment is started and should present no great problem. In any case, calculations should be checked through all stages from the field records. 5.3 Interpretation This covers a wide field but there are several essential rules: 276 - All results are subject to inherent variability and their reliability can be as- sessed by considering the size of an effect in relation to this variability (ex- perimental error). - The size of a treatment effect is compared with the error and its significance is expressed in terms of probability, in other words as risk that the effect is not real. The degree of risk acceptable in applying a chosen treatment depends on the nature of the treatment: * in the case of fertilizer, significance at p=0.05 infers a I in 20 risk that the treatment may prove ineffective. In practice this risk is small and a rather higher risk (p= 0.10 or 0.20) might be acceptable because the risk that the treatment may not be effective in one year is lessened because the residue left in the soil benefits later crops. * in the context of plant breeding, however a significance level of 0.05 is too low when the choice of variety may affect a planted area of 100000 ha. - Appreciation of the results of fertilizer experiments is improved by fitting them to a mathematical model, using the method of least squares. It may be wrong to consider that a correlation coefficient near to unity is always good while a lower value is poor. A frequent error is to attribute cause to effect when two variables are adjusted to a mathematical model. Models have been set up to simulate observed facts such as less than propor- tional yields as in the Mitscherlich equation: - c y=A (I-e (x+ b)) where y=yield, A=maximum yield, c and b are constants and x=fertilizer rate, or in an ADAS model of the form: y=a+cx+brx where a, b, c and r are constants. Again, on the basis of the Liebig's Law of the Minimum one may use the intersecting straight line model of Boyd et al. [1976]. The purpose of all these models is to smooth the experimental values but the adjustment does not necessarily imply reality. 6. Conclusion The aims of fertilizer experiments are quite different from those of other agro- nomic experiments concerning, for instance, plant protection measures or plant breeding. The effect of fertilizer rate is only partly attributable to rate since the soil contributes to the nutrient supply of the plant. Again, within the domain of fertilizer research, P and K experiments differ from those with nitrogen since the residual effects are generally confounded 277 with direct effects. The research worker has to decide whether he is interested in the immediate effect of a fertilizer treatment or in the cumulative effect of that treatment over a period. Fertilizer experiments require the greatest care and only experience can en- sure that nothing that may affect the outcome is neglected from the very begin- ning of the work. In long-term work, care taken in the establishment of the experiment determines its success and the question of balance of a nutrient is more important than the effect of fertilizer rate itself. The old adage that a chain is only as strong as its weakest link applies just as surely in fertilizer work as in any other activity. 7. References Beringer, H.: Influence of potassium nutrition on the quality of root and tuber crops. 5th Academia Sinica IPI Workshop, Kunming and Wuhan, China (1989) Boyd el al.: Factors governing the effective use of nitrogen. Span 19, 2, 68-69 (1976) Chevalier, H.: Fertilisation azotte, phosphat&e et potassique de ]a prairie temporaire exploi- t6e au rythme de la p~ture. Fourrages No. 62, 133-159 (1975) Gravoueille,J. M.: Fertilisation et qualite de la pomme de terre. La Pomme de Terre fran- qaise, No. 442, 213-228 (1987) Jourdan, 0.: Conception d'un r6seau d'essai de fertilisation au champ. Dossiers Agrono- miques dAspach le Bas, No. 3 (1988) Lindhauer, M.G.: Influence of K nutrition and drought on water relations and growth of sunflower (Helianthus annuus L.. Z. PflanzenernAhr. Bodenk. 148, 654-669 (1985) Martin-Prvel, P.J.: Physiological processes related to handling and storage quality of crops. Proc. 21st Coll. Int. Potash Inst. 255-283 (1989) Villemin, P., Ballif J.L. and Lefevre, F: Fertilisation potassique en terres de Craic. Centre de Recherches SCPA, 68700 Aspach le Bas, 1990 Tranchefort, . andPhilippeau, G.: Dispositifs exp~rimentaux usuels dans les essais agrono- miques de plein champ. Recueil de plans-types, ITCF, mars (1971) 278 Soil and Plant Test Data in Computerised Fertilizer Recommendations 1. Collin* Summary The costs of a fully equipped laboratory for soil analyses demand that it should serve a large area; it should operate on acountry rather than on a regional basis. The interpretation of crude analytical data to provide sound advice on such a scale necessitates the use of computerisation. The computer program has to be based on a bank of regional data estab- lished from the results of long-term experiments and must be capable of dealing with data from routine soil analyses and more detailed investigation including biological (Stanford de Ment) tests. In some cases it is necessary to simplify the model used to a level which still gives satisfactory practical results. It is essential that the results of analysis and con- sequent advisory recommendations should be available quickly. There is always some dilemma between objective advice and the tailoring of that advice to make use of commer- cially available fertilizer formulations and the program must allow for this. 1. Introduction Formerly, soil laboratories were established to serve particular regions, but the cost of equipping a laboratory to provide a full range of services is now so high that the establishment has to serve as large an area as possible, in our case the whole country. Over 30 million hectares are under cultivation in France and this area is divided into some 400 natural regions within which conditions are uniform. This means that programs for interpretation of analytical data have to allow for regional adjustment of the resulting advice and have to allow for variation in soil type within the region. Samples sent for analysis by farmers may originate from one or several fields. The advice requested may be either to monitor the evolution of soil fertility or to answer a specific question relating to a crop or field. Every sample is accompanied by an information sheet showing at least the geo- graphical location, intended cropping and target yield. The objective of the processing of analytical data is to provide advice and for this to be effective that advice must be available to the farmer within three weeks. It is provided on a standard form (Appendices I and 2) giving analyti- * D. Collin, Centre de Recherches SCPA, F-68700 Aspach-le-Bas, France 279 cal data plus specific recommendations in such a manner as to obviate the need for long discussion between advisor and farmer. The slight risk of error in such a system is considered acceptable but the local advisor is in a position to 2. Present day bases for fertilizer advice 2.1 Establishment of norms for interpretation The SCPA disposes of a mass of data on potash response, and to a lesser extent on phosphate response, originating from more than 2000 replicated field experiment results over the period from 1960 to the present. Much data is also available from pot culture experiments. An accepted basic principle is that a soil with long-standing nutrient residues of potassium and phosphorus has a higher yield potential than one which has only recently received corrective treatment or a low nutrient soil even when heavy dressings of P or K are given. The objective therefore is to identify for each nutrient and for each crop rotation a soil level above which fertilizer applied at above the maintenance rate (removal) will give no yield response with the most responsive crop of the rotation (van der Paauw[1952]; Villemin [1990]; Villemin (1987]). For potassium, the criterion K exch./CEC is usually preferred to exchangeable K alone. Norms may be established on alternative bases according to locally available data: a. Local basis (Figure 1). Yield indices are stated in terms of the frequency of response at different soil nutrient levels. The norm to be used depends on the frequency of K-demanding crops in the rotation with an acceptable risk of divergence from that norm. To establish this type of norm, many field experiments are needed, which is costly, and it is risky to extrapolate from the experiment to other conditions. Frequencies of responses 100%. X X , I Norm I x K/CEC Yield with K-Yield of control Figure I. Frequencies of responses (yield index = Yield of control 280 b. Universal basis (Figure 2). A figure relating yield response to the ratio K exch./CEC has been developed from the results of SCPA field experiments. With very few exceptions the trace on this figure separates the conditions where response to K is to be expected (11) from those where it is not (Jo). This somewhat empirical approach has the merit of suggesting a basic norm which can be used in the absence of precise knowledge of a local situation (Quemener [1976]) while it can be modified as more regional data become available. The general relationship can be stated in the form: K20= 4(0.077-CEC-0.035)' KO CEC Figure 2. General norm calibrated according to SCPA field trial results. The same kind of reasoning is used for other aspects (P, Mg, pH, organic matter, trace elements) borrowing some data from other research organisa- tions (Remy et at. [19741). One or more analytical properties which affect nutrient availability (Table 1) are used to assist diagnosis and other conditions like structural stability, microorganism activity may be taken into account. Table I. Bases for soil diagnosis Bases of diagnosis Element Clay* Organic pH Total matter carbonates P ...... X M g ...... X Organic matter ...... X X pH (CaO ) ...... X X Z inc ...... X C opper ...... X Boron ...... X * Clay and CEC are generally highly correlated I CEC determined by Riehm method 281 2.2 Practical advice For P, K and Mg, advice is directed to: - eventual correction of soil nutrient level up to the desired norm - maintenance of that level by applying fertilizer to replace losses. Corrective dressings take into account the difference between soil nutrient level and the required norm, soil mass and fixation coefficient. In the case of potassium, corrective dressing= required increase in exchangeable K20x soil massx fixation coefficient. Mean values may be used for soil mass and fixation coefficient. In the case of magnesium, the theoretical value for MgOxsoil mass is empirically reduced in inverse proportion to clay content. For practical and financial reasons, the total corrective dressing should be applied over a period of several years. Maintenance dressings have to replace nutrient losses occurring in three ways: - Removals by crops depend on the treatment of residues (straw, tops, cobs etc. ploughed in or removed). Standard values may be used for grain/straw or top/root ratios and for nutrient content of forages (yield expressed in dry matter) and these may vary from region to region. Allowance must be made for crop variety, use of straw shorteners etc. Plenty of data is available from field experiments. - Fixation. Simple replacement of crop removals is insufficient to maintain soil nutrient level. Allowance must be made for aging of phosphate and potassium fixation. This will be of the order of 50% for phosphate and 10% for potassium. - Leaching of potassium and magnesium depends on clay content and miner- alogy, rainfall and exchangeable nutrient content. Such data for a region is difficult to obtain. For given climatic conditions, leaching is inversely proportional to clay content. In the absence of local lysimeter data, average values may be used. Soil Ca content is normally sufficient for crop nutrient requirement, but recommendations for liming are based on pH, organic matter and clay con- tent. Advice regarding organic matter and trace elements is empirical in line with current practice. Advice for nitrogen fertilizer is approximate and ad- justed locally since Nmin. is not determined. To draw the attention of farmers to an eventual sulphur problem in high yielding or sulphur demanding crops (brassicas), and in regions where there is no adequate sulphur mineralisation, removals of S are stated in the reports. 282 3. Regional soil studies 3.1 Potassium Exchangeable K is a poor indicator of the likelihood of crop response. K ex- tracted by a stronger reagent such as sodium tetraphenylboron is a better indi- cator (Figure 3) (Quemener [19741). Less readily available potassium plays an important part in the nutrition of cereals and grasses and the relationship between K extracted by NaTPB and exchangeable K for each soil type is used in the adjustment of norms (Figure 4). The linear regression between the two is: K extr. by NaTPB= ai exch. K±0 (r2 >0.9) K uptake (ppm) 3a K uptake (ppm) 3b + + + + + + +4 + ++ +Z+4+++ + Exch. K : K extr. by NaTPB Figure 3. Relation between potassium extracted by pot cultures for a set of various soils and exchangeable potassium (3a) or potassium extracted by NaTPB (3b). K extr. by NaTPB Exch. K regionalized NN general norm (IN) Exch. K CEC Figure 4. Soil type modulation of norms: GNi=GN ,G according to the average K extracted by NaTPB. ai 283 The relation between the value ai for a particular soil type and that for the general norm ac, is used for adjustment of the norm. For each soil type, the specific fixation coefficient is taken into account in estimating corrective dressings. General models are of the type: Fixation capacity= f(CEC, organic matter, total carbonate, exch. KzO) CEC is preferred to clay content. Figures 5a and 5b illustrate the variation between and within soil types (Colin et al. [19901). Where K extracted by NaTPB is low in relation to exchangeable K, exchangeable K content is a good indicator of requirement. Soils with a very high fixation capacity require higher K fertilizer dressings. K extr. by 5a Measured fixation 5b NaTPB A 100%- capacity (van der Marel) A OAAA AA &A++A A A 0 Estimae *A* A fixation capacity A A 100% Figure 5. Relations between K extracted by NaTFE and exch. K (5a) and between van der Mare fixation capacity and estimated fixation capacity for 3 soil types (5b). The reliability of these data is checked with local references, principally field experiments, before inclusion in the computer model. This is done be- cause: - intermediate K is particularly important for cereals and grass, - buffer capacity is important when requirements are high on shallow soils - no account is taken of the volume of soil explored by roots. Where there is contradiction, we have recourse to microculture and field ex- perimentation. In practice, a regional study calls for collaboration between Aspach and a regional expert (soil scientist, agronomist). Some 50 samples of surface and subsoil are subjected to routine and more detailed analysis (fixation capacity (van derMarel), NaTPB, CEC (A ddiscott). Other determinations like wet fixa- tion and exchangeable K on moist samples may not be used for advisory pur- poses but contribute to fundamental studies. The soil units used in computerised advice are identified by geographical coordinates and certain soil characteristics (chiefly clay content or CEC and total carbonates (Villemin [1987). 284 3.2 Other nutrients The analytical service has been expanded to include other nutrients and soil factors. P205: Relations between the three main methods used in France (Dyer,Joret-Hebert, Olsen), fixation capacity (Fardeau)(Villemin et al. [1990]. CEC: Clay content, relation between different methods (Metson, Addiscott). CEC is obligatory for routine interpretation. If clay is not individually determined, it is estimated from CEC and organic matter content (Villemin [1987). pH and CaO: For pH (water), simple linear regressions between pH and exchangeable Ca or saturation of the CEC are easier to use than more sophisticated methods used in fundamental research. Trace elements: Extraction with DTPA or EDTA give correct indication of availability and adjustment for soil type and climatic conditions will give a good indication of the likelihood of deficiency in crops. Trace element status of the subsoil the contribution from which is often under-estimated is important. 4. Cropping system The sequence of crops in the rotation, their frequency and specific nutrient needs (high K-demanding row crops for example) is used to modulate recom- mendations using data from regional field experiments. A predominantly cereal rotation would have a lower K need than one including root crops. Two other general points must be observed: - information on the sheet accompanying the sample may be poorly com- pleted or the farmer may not yet have decided his cropping plan. This might lead to contradictory advice for similar analytical situations. - change in crop rotation; the area under oil crops has increased by a factor of five over the past ten years. Three main cultural systems are distinguished for the purpose of advisory recommendations: - Orchards and vineyards. Weighted values for surface and subsoil are used and emphasis is placed on bringing the soil up to the desired nutrient status before planting these permanent crops. This is important because it is not easy to apply corrective dressings later. 285 - Permanent grass. Under grazing, allowance has to be made for potassium recycled through the animal. Depending on grazing system recycling is as- sumed to be 25 to 50% of potassium uptake. Recycling also depends on the number of years under grazing management and the stocking rate. The level of N fertilizer recommended depends on the presence of legumes in the sward. The dense rooting system of the sward with ability to extract potassium from non-exchangeable reserves makes it difficult to recommend corrective dressing except for definitely deficient conditions. - Market gardening. Such systems have long received attention from research workers and advisors. For market garden crops, the farmer uses a different coloured form but the information on intended cropping is not always com- plete and the advice sheet is headed «vegetable norms used. The norms are based on the cropping system rather than soil type. 5. Computerisation 5.1 General principles Software is designed for ease and speed of operation and is menu-based with limited and simple codification which allows novice operators quickly to at- tain a speed of 300 interpretations per day. Analytical results are usually downloaded direct from the laboratory; the two other main inputs are crop rotation and geographical location (Lambert coordinates or, more usually, the commune which is the smallest administra- tive unit [about 50 km 2 ]). The latter gives information on the physical unit (liability to leaching, cropping potential etc.) and soil type which combines natural region and soil analytical features (CEC and total carbonates). Files are accessed using specific keys in succession or generalised data if specific information is lacking. 5.2 Supplementary advice IPAS (Software for soil analysis interpretation [Interpr&ation personnalis~e des analyses de sols]) gives advice in terms of nutrient units, avoiding reference to commercial products, so choice of the best available commercial formula may present difficulty. To aid the fertilizer supplier and/or farmer a special program has been devised to suggest how the necessary amount and nutrient ratio for any particular case can be made up from locally available compound and straight fertilizers. Non-quantified aspects such as pH range for the crop, chlorine content for Cl sensitive crops, sulphur content, are taken into ac- count. In the absence of soil analysis, maintenance dressings are recom- mended. 286 5.3 Hardware support IPAS was designed for a mainframe computer before the advent of the univer- sal micro-computer and this had disadvantages - because computers were scarce and expensive, the software could not be widely distributed and de- velopment of software was slow. The (Videotex system, introduced in 1985 (Claustriaux [1986]; Collin et at. [1985]) met with only limited success because there was a tendency suddenly to shower the farmer with advice, some of which may have been contradictory, and the results were not in a form suitable for later modification in the light of more recent soil analysis results. The WAS program was rewritten in 1986 to run on micro-computers oper- ated by DOS or Prologue. Since then it has processed over 50000 analyses. 6. Checks on the success of the method There are three indicators of the success of the system: - Acceptance of the advice by farmers and field advisors: sometimes the rates of fertilizer used are higher than advised (particularly profitable crops) or lower (traditional crops). - Statistical examination of the results on a regional basis and changes in fertility status of soils (Villemin et al. [1989J) usually agree with the opinion of local experts. - Success is indicated by the fact that the software has been widely taken up by cooperatives and private laboratories throughout the country. 7. Conclusion Originally conceived for the purpose of disseminating advice and the results of fundamental and applied research at Aspach, computerised handling of analytical data has proved to be a success. Now more than 50000 farmers have access to a personal computer and a release of IPAS using the expert system approach should be planned for the future. 8. References Balland, D. and Quemener, J.: Comment mettre au point des normes d'interpr tations? Exemple du Sud-Ouest. Forum Fertilisation, Atelier 2, COMIFER, 1982 Clausiriaux, JJ.: Le Vidotex agricole: utopie ou r6alit6 - Les informarma-g-iciens. Namur, November, 1986 Collin, D., Villemin, P, Quemener, J. and Bornard, J. C: Interpretation personnalise d'analyses de sols (IPAS) par Vid~otex. Convention Informatique SICOB, 20 Sep- tembre, 1985 287 Collin, D, Villemin, P and Koller, R.: Variabilit6 r~gionale et interr~gionale des caract~ris- tiques analytiques concernant le potassium. Dossiers Agronomiques d'Aspach-le-Bas, N- 4 (1990) Collin, D. and Merelle, F: lntrt d'une banque de donnees oligo-6l6ments - cartographic - exploitation pratique au profit des agriculteurs. 2' journe d'analyses de terre GEMAS Blois, 1989 Jourdan, 0.: Potash recycling on pastures. Society of chemical industry, London, 1987 Laurent, F., Taureau,J.C., Thevenet, G. and Bregeon, E: Methodologie pour un diagnostic des situations carences en cuivre. Perspectives Agricoles No. 11 (1986) Quemener, J and Rolland, D: Application de ]a technique de Stanford et De Ment AI'ex- traction du potassium des sols. Ann. Agron. 21 (6), 819-844 (1986) Quemener, J.: The measurement of soil potassium. IPI Research Topics No. 4,48 p. (1978) Quemener, £ : Le conseil de fumure potassique tire de l'analyse de sol A la SCPA. Dossier K20 No. 6 (1976) Quemener, J, Bornard, I C. and Villemin, P.: Informatique et r6gionalisation du commen- taire d'analyse de sol. C.R. Acad. Agri. de France, 71, No. 4, 403-412 (1985) Quemener, J: Complementarit6 de l'expirimentation et des 6tudes de la dynamique du potassium pour la r~gionalisation du commentaire d'analyse de so]. C.R. Ac. Agr. France, 71, No. 4, 389-401 (1985) Quemener, J: The use of reagents containing sodium tetraphenylboron for the extraction of soil potassium: applications to the study of fertility problems. Proc. 10th Congress Int. Potash Inst. Budapest (1974) Remy, Marin-Lajlrhe. L'analyse de terre: ralisation d'un programme d'interpritation au- tomatique. Annales Agronorniques, 25, 4, 607-632 (1974) Van der Paauw, E: Evaluation of methods of soil testing by field experiments. Ass. Int. Sci. Sol. Dublin, 1952 Villemin, P., Quemener, J., Bornard, I C. and Collin, D.: IPAS - Logiciel d'interpr6tation personnaliste des analyses de sols. Annales de Gembloux, 92, 1-12 (1986) Villemin, R: Classification automatique des formations superficielles en vue de l'interpr6- tation de lanalyse de terre (in press) Villemin, P.: Fertilisation potassique en terres de craies. SCPA, 1990 Villemin, P and Fardeau, £ C.: Limons sur argiles. SCPA, DAA No. 5 (1990) Villemin, P: Translation of laboratory K-data into K fertilizer recommendations. Proc. 20th Coll. Int. Potash Inst., 199-210 (1987) Villemin, P and Boiffin, J.: Evolution de Ia fertilitd des sols de IAisne. Synth~se de 15 000 analyses de sol, INRA-SCPA, 1989 Villemin, P.: Diagnostic de la fertilisation phosphat6e des sols. Premitre journ6e de lana- lyse de terre, GEMAS Blois, 1987 288 Appendix 1. Soil analysis References: 90 10 015 340 000001 201 Laboratory No. 036 Data sheet No 23001 Date: 09/02/90 Sample No. smallhedge - -I Field area Sampling date: January 90 CLAUDE MICHEL Received: 16/01 THE BIG FARM " Laboratory approved by * 10120 STGERMAIN 378 " the Ministry of Agriculture* L FERTILIZER ANALYSIS-] Soil content Norm Very tow Medium Medium High (me/kg Ig/kg) g/kgl l low high * pH lwater) 6.5 N "Organic matter 26 19 I (C.Org. x 1.73) 0 Total nitrogen 1.5 C/N 9.8 A * * Assimilable phos, (P,O) DYER 0.45 0.5 eC.E.C. 120 "Exchang. cations Potassium (KO) 4.3 0.20 0.23 Magnes. (MgO) 4.5 0.09 0.11 Calcium (CaO) 103.9 2.19 3.08 Sodium (Na2O) 1.3 0.04 " Saturation level 95% 2o Micronutrients Iron (Fe) DTPA 26 20 Mangane(Mn)TPA 21 8 Copper (Cu) 0TPA 0.20 0.35 Zinc (Zn) OTPA 0.60 0.88 Boron (B) 0.50 0.49 * Miscellaneous Chlorisis index : 0 This soil has been surveyed in 1989. The norms have been in- Active carbonate : 0 creased by 1.1.The estimated K fixation capacity is 65%. Granulometric analysis (g/kg) 5 fractions (free CaCO,)+O.M.+resid. moisture=1000 * Total carbonate (CaCO,) 0 * Clay : <0.002 mm 213 Silt :0.002-0.02 mm 270 Coarse silt : 0.02-0.05 mm 302 Sand : 0.05-0.2 mm 132 Coarse sand :0.2-2mm 53 Texture Sandy clay loam: Very stable soil 1 ;S All results expressed in relation to dry fine earth SADEF Laboratory approved by the Ministry of Agriculture, Rue de Ia Station - 68700 Aspach-le-Bas T61. 89 48 91 67 -Tlex 881 101 - T6lcopie 88 48 79 03 289 Appendix 2. Fertilizer advice for the rotation (kg/ha) Crops Rapeseed Wheat Grain maize Wheat Residues Ploughed in Ploughed in Ploughed in Exported Yield 3.2 t/ha 7.5 t/ha 9.5 t/ha 7.5 t/ha Nitrogen 180 220 190 1 220 in2 in2 in1 in2 spreading(s) spreading(s) spreading(s) spreading(s) Sulphur removal (SO,) 130 55 50 80 PO, K20 MgO PO, K20 MgO P,0, KO MgO PO, K20 MgO Removal 50 30 15 75 45 15 70 45 15 85 135 25 Maintenance fixation 15 20 0 25 15 0 20 15 0 25 15 0 Leaching 0 20 30 0 20 30 0 20 30 0 20 30 Total Maintenancefertilizer (E) 65 70 45 100 80 45 90 80 45 110 170 55 Reductionorcorrection(R) -25 40 40 -35 40 40 -30 40 0 -40 40 0 TOTAL(E)+(R) 40 110 85 65 120 85 60 120 45 70 210 55 Distribution crop rotation 65 170 120 55 120 40 65 145 70 55 120 40 Phosphorus: Soil content being very high, maintenance dressing has been diminished by 35%, possibly for 5 years. A new analysis is advised in 5 years. Potash: maintenance fertilizing has been increased by 40 kg/ha annually for 4 years. Total supplementary fertilizer: 160kg/ha. Liming: Apply 2.4 t/ha CaO to reach a pH of 7.0. Do not plough in more than 1.5 t CaO at a time. Magnesia: The annual amounts being low, apply 210 kg/ha every three years. Nitrogen: Nitrogen fertilization must be modified mainly in relation to climatic condi- tions, and organic manure ploughed in. Copper: Plough in for correction 5 kg/ha of copper (Cu). Zinc: Plough in for correction 3 to 4 kg/ha of zinc (Zn), I A7 290 Soil Fertility Data Banks as a Tool for Site-Specific K-Recommendations E. Andres* Summary Accounting for a whole range of site factors is indispensable, if fertilizer prognosis is to be efficient. This applies in particular to the nutrient potassium, as its availability depends on a number of site factors which are not considered in conventional advisory practice based on soil test data alone. Neglecting the influence of site factors gives rise to lack of precision in fertilizer recommendations. This is the background on which a concept for formulation and propagation of site- specific K fertilizer recommendations was developed. Based on a dense network of long- term K fertilizer trials on different sites, automated management of information is achieved by means of a geographic information system. This system is capable of identifying all areas which are cartographically compatible with the site factor combination of the area under study. Within the calculated validity areas experimentally confirmed information for ecologically and economically optimizing the soil K status and K fertilizer rates will be made available to farmers and advisors for a given site and crop. Information systems like (KALIPROG> offering POTASH PROG(nosis) accounting for site characteristics constitute progress both in site-specific advisory practice and future test programs. 1. Introduction The aim of K fertilizer application is to make up for the difference between the K status of soil and the K demand of plants and to ensure that sufficient nutrients are available in all developmental stages. For economic and ecological optimization of fertilizer input it is necessary to determine this difference as precisely as possible. Conventional approaches based e.g. on soil test data, nutrient offtake tables and balance sheets give only approximate estimates of soil K reserves and K demand of plants. Precise prediction of optimal K fertilizer rates in the medium- and long- term range requires profound knowledge on the soil-plant system's response to fertilizer input accounting for the specific climatic and management condi- tions at farm level. How this knowledge is obtained and how it is applied in the assessment of optimal K fertilizer rates for a given site with the aid of a geographic information system will be described in the following. * Dr. E. Andres, Agricultural Advisory Service, Kali and Salz AG, Friedrich-Ebert- Strafle 160, D-3500 Kassel, Fed. Rep. of Germany. 291 2. Site factors and potassium availability Availability of soil potassium for plant uptake depends on a number of factors: " reserves of exchangeable K, " reserves of non-exchangeable K, " soil texture, " clay content and type of clay, " soil structure, " organic matter content, " effective rooting depth, " water, heat and gas economy, * specific K uptake efficiency of the plant. Yield (grain equivalent) t/ha % Clay % Clay 0 Fluvi-Eutric-Cambisol 30 L Stagno-Dystric Gleysol 34 A Orthic Luvisol 24 0 Eutric Cambisol 37 fl Fluvi-Eutric-Cambisol 37 U Vertic Cambisol 39 9- y =75,95-0.27 •x 8- r2=0.34 7- 6- 5- wrong interpretation of data 10 20 30 40 50 60 70 80 K-content of soil* (mg K20/100 g) Calcium-Acetate-Lactate extraction (Schfiller) Figure 1. Site specific relations between K-content of soil and plant yield (Data: Orlovius [1984). 292 The combination of these factors is variable depending on the individual site and gives strongly site-specific relationships between K reserves determined by soil tests and K uptake or crop yields observed in field trials (Schlichting and Sunkel [1971]). Orlovius [19841 was able to demonstrate in long-term field trials on heavy soils in southwest Germany that a characteristic relationship between soil K reserves and crop yield existed for each of the investigated sites (Figure 1). This result was corroborated by field trials on an enlarged scale showing that this differentiation even occured at similar clay contents (Andres [19891). If all points of measurement in Figure I are calculated together, the result reveals that increasing the soil K status does not give a yield advantage, although yield increases had been measured on each of the 6 sites. For this reason works relating yields or yield increase by fertilizer K on different sites to soil K reserves (cf. Schachtschabel [1985]) do not provide a sound basis for assessing the soil K status of a given site. Test results by Wehrmann and Kuhlmann [1983] also show that yield response to fertilizer K is greatly influenced by factors which are not accounted for in conventional systems of soil-test-based fertilizer advice: In 24 fertilizer trials with sugar beet the fertilizer optimum ranged between 0 and 500 kg K20/ha on the different sites independent of the soil nutrient contents. These results are not surprising, as the nutrient index according to which the soil nutrient contents are classified is only based on the soil test ratings and clay content. Advisory practice at individual farm level should therefore aim at greater precision in the prognosis of K fertilizer requirements by accounting for a whole range of site factors in the interpretation of soil test results. 3. Possibilities of accounting for a number of site factors in fertilizer prognosis As long as the influence of all site factor combinations on the K status in soil cannot be quantitatively assessed, long-term field trials will constitute the most important data basis from which optimal soil K status and optimal K fertilizer rates can be derived. Field trials reflect the influence of a whole range of site factor combinations, including the changing local weather condi- tions in the course of a year. In conventional fertilizer prognosis based on soil analytical results as com- monly practiced by agricultural testing and research stations the test sites are at best grouped according to clay (and partly also to humus) content; mean values of the desired soil K status and mean K fertilizer rates to raise the soil K status to the desired level are calculated from test results of groups having the same clay contents. This deprives the farmer of decisive information on the response of the individual site and thus of the actual precondition for adapting the K fertilizer rates to the site-specific requirements. 293 Applying site-specific information obtained on one test field to other locali- ties with identical site characteristics would mean a considerable improvement in fertilizer prognosis. Figure 2 shows how this theoretical model is translated into practical advice: A network of long-term field trials on different sites is established in which the optimal soil K reserves and the respective K fertilizer rates are determined for each site, followed by detailed site descriptions of the locality. Suitable maps are evaluated according to areas with identical site factor combinations and K dynamics. These are presented in the form of special-purpose maps offering both to farmers and advisors experimentally confirmed specific information with regard to optimal soil K status and respective K fertilizer rates for given soil and climate factors. on differenLomtem S isAssessment fiW soil-K status andof optimum optimum on dlfferent sites K fertilizer rates Application to areas Aplio tCorrelatewt ste-specfc crop response charactedstics (soil characte s factors, climatic factors) Utilization of maps Supporting research into (Geol., soil, climate) causal relationships Figure 2. Procedure of incorporating site factors in potash recommendations. 4. Geographic information systems in advisory practice The procedure of preparing these special-purpose maps by hand is very time- consuming. Automated management of information with the aid of geo- graphic information systems is more convenient (Figure 3). Geographic infor- mation systems are computer systems capable of recording, examining and printing out data for defined areas. Real landscapes are reproduced in the form of geographic models by means of computer-linked methods. This may be achieved with the aid of thematic maps representing the spatial distribution of landscape characteristics like relief, utilization, geology, soils. Furthermore, these maps can be super- imposed in such a way as to obtain areas with a combination of selected criteria. These areas correspond to functional spatial units; with regard to the nutrient potassium they correspond to regions of identical K status. A detailed description of geographic information systems and how they work is given by Schaller [19871. 294 Real landscape Input Output Input a Materials Information Output Energy Money Presentation of models and results Static Quasi-dynamic Dynamic Geographic information Dynamic feedback system models Relief Geology Soil Groundwater zonalGeogr. presentation, models, Surface functional spatial waters units, networks Utilization without variation Biotope0 in time, transit- ion matrix Techn. infra- structure Type 1 Type 2G Time-related Function-relatedlee[ Time ne ASe uences Themat. maps Statistics 1980 82 83 Figure 3. Reality and models of geographic information systems (after Schaller and Spandau [19871). 295 Kiel 6Se • Harrover so 0I 0 0 O0' O 00 0 Berlin Oaiei 0g : 0 0 0 0 0 Munc~hen Figure 4. Longterm field experiments on arable land of the Advisory Service of Kali und Salz AG. 296 Extraordinary progress has been made in the field of data processing. The Agricultural Advisory Service of Kali and Salz AG is making use of these advantages and has elaborated a new concept for the formulation and propa- gation of site-specific K fertilizer recommendations (Andres and Orlovius [1987]). The main features of this concept are: (a) a dense network of about 150 long-term K fertilizer trials in almost all agricultural areas of the Federal Republic of Germany allowing to obtain empirical basic data for economic and ecological optimization of soil K status and fertilizer rates (Figure 4). (b) a geographical information system with digital storage of information on soil and climate factors (thematic maps) for complete coverage of the respective areas. This information system called KALIPROG is used for POTASH PROG(no- sis) in advisory practice. A model is presented in Figure 5. SofiMap Geoloial Ptcpiaioi Tempe-rature_ Map of Geographical Natural Areas" Map Digital Cneso Input of Results and Production Site Parameters of of Fiek Trials Thematic Maps Identification of Aras Quotng wit Deffinite Site o Factor CombinationsStenfmaos Optimal Soil-K- Content and Fertil- izer Rate Figure 5. Structure of the site information system KALIPROG®. 297 The prerequisite for computer assisted processing of maps is the digitaliza- tion of criteria represented in the map (qualitative properties: e.g. soil type; quantitative characteristics: e.g. amount of rainfall), i.e. the transformation of the analogous graph of the map into a numerical form. Detailed instruc- tions and a literature survey for digital mapping are given by Bach [1987]. To provide complete coverage of the Federal Republic of Germany with a recommendation network the maps selected for this purpose have to fulfil the following three conditions (after Franzle [1983]): - the maps should cover the whole area, - they should be directly or indirectly related with the factors pertaining to the site, - the data material should be available to all users. In the Federal Republic of Germany only geoscientific and climatic maps at a scale of 1: 1 mio. or 1:2 mio. meet these requirements. For regionalized in- terpretation of long-term field trial results, i.e. tests with a duration of at least 6 years and/or two crop rotations, the 1: 1 mio. scale was chosen as a com- promise. The survey area is.covered by a reference grid in the form of a series of Ix I mm squares. On 1:1 mio. maps I mm corresponds to a length of I km in nature. Accordingly, each grid cell represents an area of 1 km2 on the ground. The cartographic basis of the system is given in Table 1. Administrative boundaries and natural landscape complexes are delineated on geological, pedological and climatic maps. Table 1. Cartographical basis of the site information system KALIPROG>' Map Author Orig. scale Digital conversion Soil map of the Fed. Hollstein 1:1 Mio. Federal Research Institution Republic of Germany [1963] for Nature Protection and Landscape Ecology Soil map of the Fed. Roeschmann 1:1 Min. Federal Institution for Geo- Republic of Germany [1986] sciences and Raw Materials Geological map of the Anonym 1:1 Min. Federal Institution for Geo- Fed. Republic of Germany sciences and Raw Materials Mean annual precipitation Schirmer 1:2 Mio. Kali und Salz AG (mm/year) [1979] Mean annual air tempera- Schirmer 1:2 Mio. Kali und Salz AG ture (0C; year) [19851 Natural regions of Ger- Meynen et 1:1 Mio. Kali und Salz AG many at. 11962] Administrative areas of the Vgrenz 1:1 Mio. Institute for Applied Geodesy Fed. Republic of Germany 298 The digital site information system offers the possibility - to obtain individual information for given coordinates, - to generate factor combinations from different maps, - to present areas with identical factor combinations on screen or printout, - to relate areas with identical desired soil K status and respective economi- cally optimal K fertilizer rates to existing field trial network (Figure 5). A practical example is given in the following. 5. Working principle of the site information system < Table 2. Site description of the field experiment Luedinghausen Geographic location Municipality Luedinghausen District Coesfeld Major natural region Westphalian Lowland Basin Main natural region Central Muenster Land Mean annual precipitation 747 mm Mean air temperature 9.30C Soil type Gleyic Luvisol Parent material Boulder loam of the Saalean glaciation Land appraisal* SL 3 D 59/60 Taxation of natural yield potential based on soil and climatic conditions The experiment was carried out as a static fertilizer trial with 50 m2 plots and 4 replications. Initially the soil contained 21 mg K2OCAL (Table 3). Sugar beet or maize were grown in rotation with cereals. To study the influence of different soil and crop management on optimal K fertilizer rates the experi- ment was divided into two sets: one with crop residues (sugar beet leaves, straw of maize and cereals) removed from the field, the other with crop residues left behind and incorporated. During 11 years of the experiment the following results were obtained ac- counting for the actual price/cost ratio: - With crop residues removed from the field the highest net profits are ob- tained at annual K dressings of 300 kg K20/ha in the rotation with abun- dant leaf matter. In that case the soil K contents remain almost constant irrespective of the positive nutrient balance (+90 kg K20/ha and year). 299 + + + + 0 0 UU 0 2 U U U 00CW er- Considering'th differen respons ofla n eelcptoftizeK thenetproit oul vnb m rvdb plig30k 2 at ua 3000-r-- m '- a fertilizer had been applied, its maximum is reached at rates of 150 kg/ha K20 to sugar beet and cereals and 300 kg/ha to maize. These are also the ecologically and economically optimal rates for the maintenance of soil K reserves. This experiment shows that in high-yielding cropping systems K fertilizer rates adapted to the requirements of the specific site are well able to mobilize yield and profit reserves even under conditions where the soil K status is relatively high. Some authors suggest (Schachtschabel [1985/) to redress the nutrient balance in soil by simply returning the amount of nutrients removed by the crops. This is, however, not sufficient and would neither allow the realization of optimal yields nor ensure the maintenance of soil fertility. The next step is to utilize the test results for agricultural production at farm level. For this purpose the site information system has to identify all areas which are cartographically compatible with the factor combination of the test site. Figure 6 shows that large areas in Northern and Southern Germany are covered with loamy sediments of the Pleistocene. In the Federal Republic of Germany soils from loamy Pleistocene deposits constitute a total area of about 2 mio. ha. However, only a small part of these soils correspond in their ecological characteristics to the specific conditions of the test site. To determine the potential validity area of the experiment the following soil and climate criteria were combined: Soil type - gleyic luvisol Geol. parent material - loamy sediments of the Saalean glaciation Annual precipitation - 700-800 mm Annual temperature - 9-10 0 C In the Federal Republic of Germany this particular combination applies to about 24 700 ha (Figure 7). Like the area under study, all these areas are situ- ated in the major natural region of the Westphalian Lowland Basin. Within the calculated validity area experimentally confirmed information for op- timizing the soil K status and K fertilizer rates of a given site and crop will be made available to farmers and advisors. 301 Area: 19.683 km 5 4 Kiel Hamburg Wiesbaen' arbr(* S cBerlin- Figure 6. Soils from loamy pleistocene deposits. 302 West phalian Osnabr(Jck Lowland Basin Melia Horstmar , 40iee Herford •osed k MOnster Sassenberg C154 9%- GOterslo, Paderborn 0 6 54 Soest Gelsenkirh 04 Essen. -- Dortmund 41 045 wattensche" Duisburg Figure 7. Potential validity area of the field experiment Luedinghausen. 6. Discussion and future outlook Accounting for the specific nutrient dynamics of the individual sites is a deci- sive prerequisite in fertilizer prognosis to achieve optimal fertilizer use effi- ciency in plant production. In this respect, the concept of (site-related fertilizer recommendations by KALIPROG is based on the following model: - Localities with identical site factor combinations have the same optima with regard to K status and K fertilizer rates, - deviations in the site factor combination effect changes in these optima. Exception: the deviations are very small or neutralize each other, - functional spatial units (areas with identical K status) can be determined by reproduction of real landscapes in a geographic information system. The efficiency of the whole system depends to a large extent on the density of the field trial network and the scale of the maps recorded in the data bank. 303 Spatial information systems at a scale of 1:1 mio. show a high degree of aggre- gation. In view of the great variability which may occur within a small area (cf. Webster and Beckett [1971]; Mutert et al. [1979]) it is therefore only pos- sible to give guidelines for the area concerned. It depends on the advisor to modify these guidelines according to his specific experience, if site and management conditions of the field deviate from those of the test field. A diagram facilitating decision-making can be used for this purpose (Andres and Orlovius [19871). If a geographic information system is utilized for large areas, it is indispens- able to evaluate as many field trials as possible and to determine the respective areas of validity. This gives information on the already existing recommenda- tion network and is superior to the conventional approach of determing the fertilizer requirements from soil analytical data alone. Moreover, the areas for which fertilizer recommentations have to be worked out in the future can also be identified. Accordingly, information systems like KALIPROG consti- tute progress both for site-specific advisory practice and future test programs. 7. References Andres, E.: Die Bedeutung des Kali-Bodenvorrates fur das Ertragsgeschehen im Hinblick auf eine standortbezogene Optimierung der Dtingung. VDLUFA-Schriftenreihe, Kon- grefiband 1990, in press Andres, E. and Orlovius, K.: EDV-gesttatzte, standortbezogene Dangeberatung for Kalium auf der Basis langjihriger Feldversuche. VDLUFA-Schriftenreihe, 23, Kongrellband 1988, 239-261 Anonym: Geologische Karte der Bundesrepublik Deutschland 1:l 000 000. Herausgeber: Bundesanstalt for Geowissenschaften und Rohstoffe, in preparation Bach, M.: Die potentielle Nitrat-Belastung des Sickerwassers dutch die Landwirtschaft in der Bundesrepublik Deutschland. Dissertation. Landw. Fakultit der Universitat Gattin- gen, Fachgebiet Bodenkunde (1987) Frlnz/e, 0.: Regional reprasentative Auswahl der Baden for eine Umweltdatenbank - Exem- plarische Untersuchung am Beispiel der Bundesrepublik Deutschland. Umweltfor- schungsplan des Bundesministers des Innern, Forschungsbericht 106 05 028 (1983) Hollstein, W.: Bodenkarte der Bundesrepublik Deutschland 1: 1000000. Published by Bun- desanstalt fur Bodenforschung (1963) Meynen, E., Schmithusen, J., Gellert, J., Neef, E., Miller-Miny, H. und Schultze, J.H.: Handbuch der Naturrtiumlichen Gliederung Deutschlands. Bd. I und 2; published by Bundesanstalt fur Landeskunde und Raumordnung, Bad Godesberg (1962) Mutert, E., Lamp, J. and Kneib, W.: Zur regionalen Variabilitt von Baden in Schleswig- Holstein. Mitteilgn. Dtsch. Bodenkundl. Gesellsch., 43/11, 655-660 and 679-684 (1979) Orlovius, K.: Mehrjahrige Feldversuchsergebnisse zum EinfluB unterschiedlicher K-Gehalt im Boden auf den Ertrag schwerer Baden in Baden-Wiirttemberg. Landw. Forsch. 37, Kongrelband 1985, 674-681 Roeschmann, G.: Bodenkarte der Bundesrepublik Deutschland 1:1 000 000. Published by Bundesanstalt fur Geowissenschaften und Rohstoffe (1986) Schachtschabel, P.: Beziehung zwischen dem durch K-DOngung erzielbaren Mehrertrag und dem K-Gehalt der Baden nach Feldversuchen in der Bundesrepublik Deutschland. Z. Pflanzenerndhr. Bodenkd., 148, 439-458 (1985) 304 Schaller, J: Anwendung flAchenbezogener Informationssysteme for aktuelle Fragen des Bodenschutzes. Mitteilgn. Dtsch. Bodenkundl. Gesellsch. 53, 61-67 (1987) Schirmer, H.: Mittlere Niederschlagsh6hen (mm), Jahr. In: H. Schirmer and V. Vent- Schmidt. Das Klima der Bundesrepublik Deutschland, published by Deutscher Wetter- dienst, Offenbach/M. (1979) Schirmer. H.: Mittlere Lufttemperatur (*C), Jahr. In: A. Meyer and H. Schirmer: Das Klima der Bundesrepublik Deutschland, published by Deutscher Wetterdienst, Offenbach/M. (1985) Schlichting, E. and Sunkel, R.: Nthrstoffgehalte und Dflngewirkung in einigen Boden Wflrttembergs. Landw. Forsch. 24, 170-192 (1971) Vgrenz: Digitale Datenbank der administrativen Grenzen der Bundesrepublik Deutschland 1:1 000000. Institut fur Angewandte Geodasie, Frankfurt Webster, R. and Beckett, P.H.T: Soil spatial variability - a review. Soils and Fertil., 34, 1-15 (1971) Wehrmann, . und Kuhlmann, H.: Kali diingen ohne ROcksicht auf die Gehaltsklassen. DLG-Mitt. 23 (1983) 305 Experience with Fertilizer Recommendations in Eastern Europe R. Czuba* Summary Agricultural laboratories for soil, plant and fertilizer analysis have been organised in East European countries: Bulgaria, Czechoslovakia, GDR, Hungary, Poland, Romania. The results are used for formulating fertilizer recommendations. Laboratories are sited through- out the countries and carry out soil analysis at regular intervals. It is standard practice to determine pH, N, P, K, Ca and Mg contents with microelements (B, Cu, Mn, Mo, Zn) contents as required. Plant analysis is also used for fertilizer recommendations. Various methods of chemical analysis are used. The results are usually used to show up any trends in soil fertility. Soil and plant analyses are also used in pollution control. 1. Introduction Soil analysis has been used for many years in Eastern Europe for the formula- tion of fertilizer recommendations. pH, phosphorus, potassium and mag- nesium contents are always determined and usually boron, copper, man- ganese, molybdenum and zinc; occasionally iron and sodium contents are de- termined. For several years now, soil nitrogen (NO 3, NH 4) contents are also used. Plant analysis is widely used in Czechoslovakia and the GDR and, more recently in Poland. The laboratories have recently started measurement of toxic elements for the purpose of environmental protection. 2. Organisation of analytical work The laboratories are run on a regional basis. Bulgaria. The ((Nikolai Pushkarov > Institute of Soil Science and Yield Programming in Sofia as well as carrying out large series of analysis, super- vises the work of 48 laboratories all over the country. Eight of these laboratories determine pH, N, P and K while the rest measure only mineral N content. * Prof. R. Czuba, Research Department for Methodology of Agrochemical Service, Insti- tute of Soil Science and Plant Production, pl. Engelsa 5, 50-244 Wroclaw, Poland 307 Samples taken each year represent 800000-900000 ha. The results show a trend to soil acidification, the proportion of acid soils having increased from 34 to 4707o between 1970 and 1988. Czechoslovakia. The Central Institute for Inspection and Research in Agriculture in Prague is in charge of the work and conducts all the soil analysis for pH, available P, K, Mg, B, Cu, Mn, Mo, and Zn. 51 area laboratories do plant analysis. Soils are analysed every three years, covering about I million ha a year. German Democratic Republic. Soil investigation is done by the Institute for Plant Nutrition in Jena and a Department for Agrochemical Research and Consultation (ACUB) has been formed recently. Up to 1987, all soils from the whole country were analysed in 8 cycles for pH, available P, K, Mg and B, Cu, Mn, Mo and Zn. Some of the work is done by area laboratories in Dresden, Halle and Rostock. As well as the standard analysis, Jena also does some additional determinations in soil and plant material. Hungary. The Plant Protection and Agrochemistry Centre of the Ministry of Agriculture in Budapest organizes research and supervises 20 centres in the country. Routine soil analysis includes determination of pH (KCI), avail- able P, K, Mg, Na, B, Cu, Mn, Mo, Zn, total N, N-min, lime requirement, S04-S, organic matter content, compactness and salinity. Poland. Administration of soil work is in the hands of the Institute of Soil Science and Plant Production Department in Wroclaw. Routine soil analysis (pH [KCIJ, available P, K, Mg and B, Cu, Mn, Mo, Zn) is done by 17 regional stations, each being responsible for about 1.1 million ha. State farms (2307o of farming land) contract with the stations for soil analysis every 4 years. There is no charge for determination of pH and major elements; minor ele- ments are determined on request for a small charge. Analysis is also done free of charge for private farms at intervals determined by the farmers. Soil sampling is done by the farmers themselves under the direction of the station staff. All state farms are regular users of the service; about 60% of private farms use the facilities. Romania. The Research Institute for Pedology and Agrochemistry in Bucharest cooperates with 38 area laboratories who do analysis and make fertilizer recommendations. Advice for fruit is based on plant analysis and N-min in the top I m of soil by Wehrmann's (modified) method. The centres in all countries organize field experiments which are used to check recommendations. These experiments measure the main effects of in- dividual nutrients and their interactions. They establish criteria for interpret- ing results of analysis and also test new fertilizer materials. During the last 15 years workers of the different countries have cooperated in series of field experiments particularly with reference to the effects of fertilizer treatment on crop quality. 308 3. Method of soil analysis and interpretation of results There are differences between the methods used in the different countries. So far, attempts to standardize have failed. Table I summarises the methods used in three countries. Table I. Soil analysis methods used in Czechoslovakia, GDR and Poland Determination Czechoslovakia GDR Poland pH ...... 0.2 M KCI 0.1 M KCI I M KCI P ...... Egner Egner-Riehm Egner-Riehm K ...... Schachtschabel Egner-Rehm Egner-Riehm Mg ...... Schachtschabel Schachtschabel Schachtschabel B ...... Berger-Truog Berger-Truog I M HCI (Rinkis) Cu ...... EDTA Westerhoff I M HCI (Rinkis) Mn ...... EDTA Schachtschabel I M HCI (Rinkis) Mo ...... Grigg Grigg I M HCI (Rinkis) Zn ...... EDTA Trierweiler- I M HCI (Rinkis) Lindsay The laboratory methods used depend to some extent on the availability of apparatus and the range of determinations made has been widened every- where in recent years. More attention is being paid to environmental protec- tion and, in addition to routine soil analysis, soil and plant contents of As. Cd, Cr, Co, Pb, Hg, Ni etc. are frequently measured. There are also inter- country variations in method of interpretation of results. For macro-nutri- ents, 5 classes of availability are used; for minor elements 3. For pH, there are 5 or 6 classes (in Czechoslovakia 7). These classes take other soil properties such as mechanical composition and organic matter content into account. The content classes used for potassium in mineral soils in Poland are shown in Table 2. Table 2. Potassium content classes used in Poland for mineral soils (mg K/100 g soil; Egner-Riehm method) Soil Content class Very light Light Medium heavy Heavy Very low <2.1 <4.1 C<6.2 <8.3 Low 2.2- 6.2 4.2- 8.3 6.3-10.4 8.4-12.5 Average 6.3-10.4 8.4-12.4 10.5-16.6 12.6-20.7 High 10.5-14.5 12.5-16.6 16.7-20.7 20.8-24.9 Very high > 14.6 > 16.7 _>20.8 25 4. Computation of fertilizer recommendations Soil analysis is the main basis of recommendations. These recommendations are designed mainly for large or cooperative farming enterprises, except in 309 Poland where the majority of farms is held privately. These are worked out by computer and print-outs giving recommendations for one or several years are sent out to the farms. There is good cooperation between countries in designing the computer programmes. In addition traditional methods, includ- ing the distribution of coloured maps, are also used. In most countries this service is provided free of charge, sometimes for a small fee. The work is organized on a long-term timetable. In Poland, recom- mendations for private farms (the majority) are based on soil analysis at 4 year intervals. Soil sampling is arranged by the stations to cover the whole of a village on one day. The farmers take their own samples using apparatus distributed by the centre. They receive the recommendations in the form of a coloured map indicating variation in soil fertility within the farm accompa- nied by the recommendations. There are 5 classes for lime requirement and major elements; three for minor elements. Examples of fertilizer recommendations are shown in Appendices 1 and 2. 5. Plant analysis and fertilizer recommendations The first countries to adopt these methods were Czechoslovakia (Baier [1977) and the GDR (Bergmann and Neubert [1976]; Podlesak et al [19881). This method is used to a varying extent in all Western Europe. In Czechoslovakia, the method is used on cereals on several hundred thousand ha. Leaf samples of winter crops are taken at the 6th leaf stage (Feekes 2) and for spring sown crops at the 5th leaf stage for determination of N, P, K, Ca and Mg contents. Recommendations are based mainly on inter- nutrient ratios. In the GDR norms for leaf contents of major (N, P, K, Ca and Mg) and minor elements (Cu, Mn, Zn) at particular growth stages (Feekes 4-10) have been established for all cereals, while for other crops (maize, sugar beet, rape, potato, red clover), the boron content is also measured (Podlesak et al [1988]). The results are used for: - diagnosing the cause of poor growth and yield, - control over plant nutrition in later growth stages, - determining measures to be taken to improve plant nutritional status. Plant analysis is also widely used in Poland. Here it is used as a guide to spray application of nutrients (Czuba [1989]). Spray applications are made 3-6 times during the growing season using aqueous solutions of urea at con- centrations adjusted to the susceptibility of different crops. In Bulgaria, Hungary and Romania, foliar analysis is used mainly for fruit and other permanent crops. 6. Soil analysis in the control of pollution Soil analysis is used in all the countries to give warning of pollution. This applies mainly to farming land, but it is also used in other cases. In the GDR, 310 the main emphasis is on heavy metals and analysis is done on soils, plants, water supplies, waste materials, fertilizers, pesticides and sewage. In Poland there are 1100 permanent bench-mark sites in regions exposed to the greatest industrial emissions and on other areas free of pollution. Soil and plant samples are taken at regular intervals. Hungarian laboratories offer a compre- hensive programme to control pollution by agrochemicals. 7. Long-term aspects Many centres carry out long-term statistical evaluation of the results of soil analysis with a view to establishing changes in soil properties over time. Some results obtained in the GDR and Poland are quoted in Tables 3-5. In the GDR (Table 3) there was a decrease over a 10 year period in the area in need of liming and soil P, K and Mg status improved. In Poland, there was some improvement in soil fertility over the first ten years, but this was followed by a decline despite increased lime usage. This was ascribed to an increase in sulphur emissions from industry. On the other hand, there was a distinct increase in soil P and K content but deterioration with respect to magnesium. Such results give a good picture of trends on a national scale. The most comprehensive long-term soil analytical data of this kind is to be found in Bulgaria, the GDR and Hungary. Table 3. Fertility of arable soils in the GDR (Podlesak et al. [1988/) Per cent of the area Element Year I 2 3 4 5 Ca ...... 1978 44 17 22 I1 6 1984 56 17 14 8 5 1987 57 17 15 6 5 P ...... 1978 14 22 48 12 4 1984 32 31 32 4 1 1987 41 30 26 2 1 K ...... 1978 53 25 16 4 2 1984 64 20 13 2 ! 1987 68 18 11 2 1 M g ...... 1978 42 9 30 12 7 1984 46 12 36 4 2 1987 51 13 33 2 1 -Ca: I - liming not required P, K, Mg: I - very high content 2 - very small liming requirement 2 - rather high content 3 - medium liming requirement 3 - medium content 4 - liming needed 4 - medium low content 5 - liming essential 5 - very low content 311 Table 4. Soil reaction in 3 cycles of investigations in Poland Period of investigations Very acid and Slightly acid Neutral and acid soils soils alkaline soils pH <5.5 pH 5.5-6.5 pH> 6 .5 By 1965 ...... 58* 25 17 1966-1975 ...... 56 26 18 1976-1987 ...... 61 24 15 * % of the area Table 5. The P, K and Mg content in Polish soils (Czuba 11989J) Mg Period of K investigations Low Medium High Low Medium High Low Medium High By 1965 56* 29 15 65 28 7 33 33 34 1966-1975 47 33 20 53 30 17 43 34 23 1976-1987 21 35 44 43 26 31 46 34 20 * % of the area 8. References Ap/tauer, J.: Krit&ria hodnoceni v ,sledkil rozbord. Agrochemicke zkougeni pild. Praha (1985) Baier, J.: Principy zAkladnich vztahd ve vfliv6 rostlin. Okresni vybor Svazu drulstevnich rolnikd v Osti nad Orlici Praha, 1977 Bergmann, W and Neubert, R.: Pflanzendiagnose und Pflanzenanalyse. VEB Gustav Fischer Verlag Jena, 1976 Czuba, R.: Naziemna i agrolotnicza technologia dolistnego dokarmiania zb6z roztworem mocznika I4cznie z mikroelementami i pestycydami, lUNG Pulawy, 1989 Debreczeni, B.: Auswertung und Darstellung der Bodenuntersuchungsergebnisse in Un- garn. Empfehlungen for Praxis, Budapest, 1985 Donev, P.: Improvement of the organization of the agrochemical supply for the agricultural production in Bulgaria. Documents of conferences, Agard, Hungary, October (1988) Hera, C., Borlan, Z. etal.: Tabele si nomograme agrochimice. Editura Ceres. Bucuresti, 1982 Podlesak, W. et at: Agrochemische Untersuchung und Beurteilung von Boden und Pflan- zen. Ak. der Landw. der DDR, 1988 Schilling, G., Ansorge, H., Borchmann, W., Markgraf G. and Peschke, H.: Pflanzen- ernahrung und Dtingung, Teil I1 - Dflngung. 2nd edition, VEB Deutscher Landwirt- schaftsverlag Berlin (1989) Sroczydiski, W el al: Zalecenia okreqowej stacji chemiczno-rolniczej. Poznari, 1989 VintilO, J., Borlan, Z., Rdutd, C., Daniliuc, D. and Tigdnas, L.: Situatia agrochimicA a solurilor din Romania. Editura Ceres, Bucuresti, 1984 312 Appendix 1. K fertilizer recommendations (kg K/ha) in the GDR (after Schilling ei al. [19861) Crop Soil Yield K supply level* group (t/ha) 2 3 4 5 Winter wheat ...... 2 4.8 54 104 134 174 3 6.0 42 102 202 202 4, 5 6.0 37 102 202 202 W inter rye ...... I 3.5 77 102 122 147 2 4.5 64 114 144 184 W inter barley ...... 1 3.5 88 113 133 158 2 4.5 77 127 157 197 3 6.0 87 147 247 247 4, 5 6.0 82 147 247 247 Oats ...... 2 4.2 92 142 212 212 3 5.5 107 167 267 267 Rape ...... 2 3.0 82 132 202 202 3 3.0 57 117 217 217 Early potatoes ...... I 19.0 123 148 168 193 2 20.0 87 137 207 207 Maincrop potatoes ...... 1 28.0 182 207 227 252 2 30.0 148 198 268 268 4, 5 30.0 118 183 283 283 Sugar beet ...... 2 40.0 205 255 325 325 3 45.0 215 275 375 375 4, 5 45.0 205 270 370 370 Grassland, low intensity ...... 1,2 20.0 36 86 136 136 3 20.0 26 86 126 126 4, 5 20.0 16 86 121 121 Grassland, high intensity ...... 1,2 45.0 130 180 230 230 3 50.0 140 200 240 240 4, 5 50.0 130 200 235 235 * 1-5; I=very high (no K fertilizer applied), 5=very low 313 Appendix 2. Fertilizer recommendations in Poland (kg N, P2O, and KzO/ha) after Sroczyriski el al. [1989] N requirement Crop Soil Target group yield Very High Medium Low Very (t/ha) high low W inter wheat ...... 1, 2, 10 5.0 135 115 100 85 65 3, 4, 8, 11 4.4 140 120 105 90 70 5 3.2 135 115 100 85 65 Rye ...... I, 2, 10 4.2 125 110 100 90 75 3, 4, 8, II 3.9 125 115 105 90 80 5, 6, 9 3.0 120 110 95 85 70 7, 12, 13 2.3 115 105 90 80 70 Winter barley ...... 1, 2 4.6 95 80 65 55 40 3, 4, 8 4.2 100 85 75 60 45 5 3.7 95 85 70 55 45 6, 9 2.4 90 80 65 50 40 Summer wheat ...... 1, 2, 10 4.4 135 120 100 85 70 3, 4, 8, 11 4.0 140 125 100 90 75 5 3.7 135 115 100 85 65 Oats ...... 1, 2, 10 4.2 90 80 70 60 50 3, 4, 8, 11 3.9 95 80 70 65 55 5 3.5 90 80 70 60 50 6, 9, 12 2.3 85 75 65 55 45 13 2.0 80 70 60 50 40 Grain maize ...... 1, 2 6.0 120 110 95 85 75 3, 4, 8 5.7 125 115 105 90 80 5 5.4 125 110 100 90 75 Sugar beet ...... 1, 2, 10 45.0 140 125 110 95 80 3, 4, 8, 11 42.0 140 125 115 100 80 5 36.0 140 125 110 95 80 Industrial and forage potatoes 1, 2, 10 32.0 130 120 100 85 65 3, 4, 8, 11 29.5 135 120 100 85 70 5, 6, 9, 12 28.0 130 115 100 85 70 Ware and seed potatoes ..... 1, 2, 10 32.0 100 90 80 70 60 3, 4, 8, 11 29.5 95 85 75 65 55 5, 6, 9, 12 28.0 95 85 75 65 55 Grassland ...... 1, 2, 10 50.0 340 320 300 280 260 3, 4, 8, 11 47.0 345 325 305 285 260 5 44.0 340 320 300 280 260 6,9, 12 35.0 330 310 290 275 255 314 P supply level K supply level Very Low Medium High Very Very Low Medium High Very low high low high 110 70 45 30 25 110 90 75 65 20 95 65 40 25 20 100 80 65 55 20 75 55 35 20 20 80 70 60 50 20 100 65 40 30 25 125 100 85 70 30 95 65 40 25 20 115 90 75 65 30 95 70 45 30 25 115 90 75 60 30 70 50 30 15 - 85 70 55 45 - 100 60 30 25 20 110 85 65 55 20 90 60 25 20 20 105 80 60 50 20 90 65 35 25 20 105 85 70 60 25 50 40 20 20 20 55 45 35 30 20 110 75 40 25 20 140 115 95 80 20 75 60 30 20 15 120 100 75 65 - 75 55 30 20 15 95 95 75 65 - 100 65 40 25 20 130 100 85 75 20 95 60 40 20 20 120 90 75 65 20 100 70 40 25 20 115 100 95 80 25 60 40 25 20 20 65 55 50 40 20 45 25 20 20 20 50 30 20 20 20 170 110 65 50 40 210 175 150 130 40 160 105 60 45 40 200 160 140 120 40 180 130 85 50 45 215 215 185 155 65 240 160 95 70 60 225 185 155 140 50 220 145 85 60 55 210 170 145 125 45 220 160 100 60 55 190 190 165 140 60 95 55 30 15 - 185 145 125 I10 30 90 45 20 15 - 170 130 110 90 25 100 70 40 20 15 175 155 125 100 40 95 60 30 15 - 195 140 120 110 35 95 50 20 - - 185 135 110 100 30 100 70 40 20 15 175 155 125 100 40 195 135 85 65 55 215 180 160 145 70 185 125 80 60 50 205 170 155 135 65 200 145 95 60 55 195 195 165 150 75 150 110 70 45 40 155 145 125 110 55 315 Experience with Fertilizer Recommendations in Italy and in Southern Europe M. Perelli* Summary Countries of Southern Europe have many common features, such as: climate, crops, low and non-homogeneous level of fertilizer use, few soil analyses and little attention to research on fertilizer use. In Italy, no more than 50000-60000 soil samples are analysed per year (average one analysis per 280-340 ha of agricultural area). There are not less than 30 laboratories, but about 50% of tests are made by eight laboratories, seven of which are located in Northern Italy. Public Administration is active in soil analysis with 10 laboratories, which make an average of less than 1000 tests per year. In Greece, about 40000 soil samples are analysed per year for advice to farmers by three laboratories, equivalent to one analysis per 225 ha. In Spain, public laboratories make ana- lyses only for research and analyses for fertilizer recommendation are made by many private laboratories, which often are very small. In Southern Europe internationally well-known methods of analysis are used, with, in Italy, a differentiation of the nutrients extraction related to the soil pH. Some Italian labora- tories determine potassium fixation. In Italy different approaches (direct Mitscherlich's equation, SLAN, BCSR) are used to interpret soil analyses of available nutrients and to make fertilizer recommendations. Today, the most used approach is the procedure Demetra 1), which uses modified norms and integrates soil analyses and environmental and agronomic data. oDemetra lo is discussed in detail. 1. Introduction There are some characteristics which connect the countries of the South Mediterranean Europe, apart from climatic conditions. Generally, there is a predominance of calcareous soils, as reported by the Commission of the Euro- pean Community [1985, and crops of the <(Mediterranean Triad)) such as wheat, grape and olive are still a peculiarity of the landscape. Moreover, we can observe a low level of fertilizer use, as compared with other European countries (Table 1). As a matter of fact, data shown in Table I are only an average: in all Mediterranean countries intensive cropping and a high level of fertilization (e.g. Po Valley, Valencia, Macedonia) adjoin areas with extensive crops or poor pastures. * M. Perelli, Agronomist, Via Puccini, 11-30034 Mira-Venezia, Italy 317 Table 1. Use of fertilizers in Southern countries of European Community in 1985 (FAO 119871, mod.) Country N P20 K2 0 Total N P20 K20 Total kg ha - ' Index EEC= 100 Greece 48.9 19.5 6.0 74.4 69.5 58.7 17.5 54.0 Italy 61.3 40.2 20.6 122.1 87.1 121.0 60.1 88.6 Portugal 41.6 21.3 10.3 73.2 59.1 64.1 30.1 53.1 Spain 29.6 15.2 10.2 55.0 42.1 45.8 29.8 39.9 EEC 70.4 33.2 34.3 137.9 100.0 100.0 100.0 100.0 In Southern European Countries, there is also a low number of soil analyses aimed at fertilizer recommendation. In the last 20 years, research on soil tests and fertilizer use was lacking and discontinuous in Greece, Italy, Portugal and Spain. In the CAB data base there are 998 records with the following key words: soil test, soil analyses or fertilizer recommendation. Only one of them refers to Greece, two to each Spain and Italy and none to Portugal. 2. Soil Analyses 2.1 Italy As shown by the characteristics exposed in Table 2, there are many differences between Italian soils and those of other countries with developed agriculture. Obviously, the content of available nutrients changes a lot, as all over the world, but the Italian situation is even worse, due to high land fragmentation, which implies a high differentiation in management; the average area per farm is 7.9 ha in Italy, against, for example, 14.6 ha in West Germany, 25.9 ha in France and 66 ha in the United Kingdom. Table 2. Frequence of selected soil properties in a sample of Italian soils (Maggiolo and Schippa [19881, mod.) Org. matter Frequence pH / Frequence C.E.C. Frequence (g kg ) (%) (070) (cmol kg-) (97o) < 5 9.2 <5.5 9.4 < 5 0.5 5- 10 43.9 5.5 -6.0 8.8 5 - 10 10.7 10- 15 31.2 6.0- 6.8 6.2 10- 15 19.7 15-20 9.7 6.8-7.3 4.3 15-20 20.6 20-25 3.1 7.3-8.0 38.2 20-25 16.3 25-30 1.1 8.0-8.5 31.4 25-30 10.7 >30 1.8 >8.5 1.7 >30 21.5 318 Before 1950, soil analyses in Italy were discontinuous, rare and research- oriented, with no consideration to fertilizer advice, with the remarkable but restricted exception of Friuli-Venezia Giulia, in North-Eastern Italy (Visintini Romanin [19821). In the 1950s, the National Association of Sugar-beet Growers (ANB) made more than 85 000 soil analyses, 84% of which were in Northern Italy. Later, soil analyses suddenly decreased and subsequently resumed following fertilizer price increase in 1972-1975 and environmental protection requirements. Soil samples analysed per year were more than 6500 in 1974 and 45000 in 1987 (Genevini [1988]). Today, we estimate that no more than 50000-60000 soil samples are ana- lysed per year in Italy which, compared with the Italian agricultural area (about 17 - 106 ha), is equivalent to an average of only one analysis per 280-340 ha. It is very difficult to have an exhaustive outline of soil analysis in Italy. There are not less than 30 public and private laboratories, but about 507o of tests are made by eight laboratories, seven of which are located in Northern Italy. The number of samples analysed per year ranges from 100 to 5000 and more per laboratory, with an average of 2500 samples (Genevini [1988]). Public Administration is active in soil analysis with 10 laboratories, but none of them is dependent upon the Central Government, which delegated responsibility to Local Administrations (Regions and Provinces). As reported by Consalter et al. [1988] (Table 3) only three public laboratories make more than 1000 tests per year, the average being less than 1000 tests per year. Table 3. Characteristics of ten public soil laboratories in Italy (Consalter et aL [19881, mod.) Years of activity ...... < 5:4 labs > 5:6 labs Analyses per year ...... < 1000:7 labs > 1000:3 labs Soil analyses in data base ...... <4000:5 labs >4000:4 labs 2.2 Greece In Greece, soil anaylses are performed mainly by the Soil Science Institutes of Athens and Salonika and by the Soil Analysis Laboratory of Hellenic Chemical Products and Fertilizers Co. There are also about ten small private laboratories in the Country. About 55 000 soil samples are analysed in Greece per year, but only 40000 samples are for advice to farmers, because about 15 000 are for research pur- poses. Considering the total Greek agricultural area (about 9 - 106 ha), it is equivalent to an average of one analysis per 225 ha. The Greek government provides for establishing new official soil analysis laboratories with a total capacity of 800 000 soil samples per year (about one analysis per year per 10 ha). 319 2.3 Spain It is very difficult to get a clear picture of soil analyses in Spain also. Public laboratories make analyses only for research and analyses for fertilizer recom- mendation are made by many private laboratories, which often are very small. Some adviser organizations, as the "Center of Vegetal Nutrition> (CNV, Barcelona) supplement soil analyses with plant analyses, in order to have bet- ter fertilizer recommendations. 3. Methods used in soil analyses 3.1 Italy The first attempt of coordinating soil analyses methods was made in 1976 by the Italian Society of Soil Science (SISS [1976). At the beginning of 1980s, the considerable development of soil analysis and the introduction of the new methodology required a unification of methods used in soil analysis. In the same period, two books regarding analyt- ical methods were published by the Italian Society of Soil Science (SISS [19851) and by the Italian Association for Unification in Chemical Industry (Unichim [1985]). Both were essentially a normalization of internationally well-known methods, already applied in most important Italian laboratories. Three years later, Unichim [1988] published a new text on semi-automatic soil analysis, with the same methods applied on automatic colorimeters. It is important to say that in Unichim methods there is a differentiation of the nutrients extraction related to the soil pH. This was founded on many studies on inapplicability of many widespread methods on calcareus soils, which are dominant in Italy (Maggioni et al. [1984]). Obviously, the meeting of the two organizations was inevitable and a new unified text has been written (SISS-Unichim [1990), preserving the differenti- ation on the basis of soil pH (Table 4). Table 4 requires a comment, bearing in mind that analytical methods are sug- gested both for pedological research and for routine tests. For instance, for texture two methods are suggested: pipette for pedological characterization of soil and hydrometer for routine tests. The organic matter content is estimated by multiplying the organic carbon concentration by 1.724. As everybody knows, this factor is too low for many soils, and, consequently, the organic matter content is underestimated (Nelson and Sommers[1982]), but this factor is well-known and applied by the Italian law on fertilizers. In some laboratories, elementary analysers are used for total nitrogen ana- lyses of soils. Some analysers give satisfactory results on total sulphur and carbon analyses (Maggiolo et al. [1987). The last mentioned data may be 320 used for total carbonate (lime) estimation. Some have started to use elemen- tary analyses for organic carbon and soluble sulphur analyses (Maggiolo,per- sonal communication). Many laboratories do not make direct analyses of cation exchange capacity, which require great care and often give inadequate results (Doll and Lucas [1973]). Generally, for routine applications they prefer to use the total ex- changeable cations, as suggested by FAO (Cotennie [19801). Table 4. Analytical methods suggested by SISS-Unichim [1990] for Italian soils Analysis Methods A I soils. pH Water suspension soil: water ratio 1:2.5 buffer pH KCI IN suspension soil: water ratio 1:2.5 Conductivity Water suspension soil: water ratio 1:2.5 Texture Pipette or hydrometer Cation exchange capacity BaC12 Organic carbon Walkley and Black Total nitrogen Kjeldhal (with or without HF digestion) Soluble boron Hot water extract Soluble chloride Water extract Sulphate HCI extract Mineral nitrogen 2M KCI extraction Potassium fixation McLean quicktest (modified) Non acid soil (pH>6.5): Total lime CO2 evolved by HCI dissolution Active lime Drouineau (modified) Available phosphorus Olsen Exchangeable K, Mg, Ca, Na BaCI2+Triethanolamine Available Fe, Mn, Zn, Cu Lindsay and Norvell Gypsum requirements Saturation Acid soils (pHsd6.5): Available phosphorus Bray and Kurtz Exchangeable K, Mg, Ca, Na I N ammonium acetate Available Fe, Mn, Zn, Cu Lakanen and Ervio Lime requirements Shoemaker et al. 3.1.1 Potassium Fixation Owing to the tendency of individual soils to fix different amounts of potas- sium, we need a good knowledge of potassium fixation in order to make a good fertilizer recommendation. In order to perform an accurate measurement of potassium fixation in soils, a 60-day equilibration has been used (McLean et al. [1982]). A quicktest proce- dure has been studied as an alternative basis for making potassium fertilizer recommendations. Since then, the quicktest methodology has been improved 321 by appling a mathematical equation adjusting the quicktest recommendations values (McLean et al. [1982]; Pichtel et at [1986]). Specific research has been done to test the applicability of the quicktest to Italian non-acid soils (Perelliet al [1988]), comparing the 60 day equilibra- tion with 2 h and 16 h quicktest on 100 soils with a wide range of chemical features. For the 2 h quicktest, the relationship between the quicktest values and the 60 day equilibrium values was: Y = 1.209 • X + 10.717 (r2 = 0.809) (1) and for the 16 h quicktest was: Y = 1.060 • X + 9.070 (r2 = 0.856**) (2) The same work showed a high variability of potassium fixation in Italian soils (Figure 1). Therefore, neglecting potassium fixation induces a remark- able bias in potassium fertilization (Table 5). Frequency (%) 15- 0105- 0 .. <10 10-20 20-30 30-40 40-50 50-60 >60 Potassium fixation (%) Figure 1. Frequency distribution of potassium fixation (60 day equilibration) in a sample of Italian soils (Perelli et al. [1988]). Relationships between potassium fixation and other soil features are very feeble (Perelli[1988a]) and, therefore, analytical determination of potassium fixation is necessary and useful for fertilizer recommendations. Potassium fixation quicktest is useful for research too, and it was used in potassium balances in crops (Table 6). 322 Table 5. Frequency distribution of bias in potassium fertilization if fixation is neglected (Perelli el aL. [19881, mod.) Bias of fertilization Frequency (%) (kg K20/ha) Grain corn Winter wheat -60 to -40 4 9 40 to -20 37 43 -20 to 0 21 21 0 to +20 12 12 +20 to +40 18 4 +40 to +60 6 4 > +60 2 7 Table 6. Potassium balances for tobacco in sandy soils, including potassium fixation (Cristanini [1990] mod.) Soil A Soil B 1988 1989 1988 1989 Fixation 21.2% 20.70 kg K20 ha-' Inputs: ...... 1369 1644 1124 1218 Soil (exchangeable) ...... 964 1154 764 730 Organic manure ...... 113 140 128 191 M ineral fertilizer ...... 292 350 232 297 Outputs: ...... 1423 1410 1095 1074 Soil (exchangeable) ...... 1236 1214 952 924 Crops ...... 125 122 95 89 Fixation ...... 62 74 48 61 Balance ...... -54 + 234 + 29 + 144 * Leaching, release, etc. Rainfall (May-October) 1988: 239 mm; 1989: 574 mm 3.2 Greece and Spain Analytical methods used in Greece are similar to those used in Italy (Table 4), but without differentiation on soil pH basis. Therefore, the Olsen method is always used for available P and exchangeable potassium and magnesium are extracted only by ammonium acetate. Spanish soils are generally calcarous and therefore analytical methods used in Spain are about the same as exposed in Table 4 for the Italian non-acid soils, but always using ammonium acetate for the extraction of exchangeable cations. 323 4. Soil analyses interpretation and fertilizer recommendations 4.1 General characteristics of the soil In Italy for particle-size classification (texture), both limits suggested by the U.S. Department of Agriculture (USDA) and by the International Soil Science Society (ISSS) are used (Table 7). Generally, the first is applied especially for pedological purposes and the ISSS classification is used for advice to farmers. Table 7. Particle-size limits according to classification schemes applied in Italy Particle size (mm) Classification SISS (USDA) Unichim (ISSS) Gravel ...... > 2 > 2 Sand ...... 2 -0.05 2 - 0.02 Silt ...... 0.05 - 0.002 0.02 - 0.002 Clay ...... < 0.002 < 0.002 Soil pH is measured in a water suspension, and classification is made, accord- ing to the same limits used worldwide, in the traditional classes of highly acid, acid, slightly acid, neutral, slightly alkaline, alkaline and highly alkaline. In most Italian soils, the lime content (calcium carbonate and often dolo- mite) is very high. Therefore, measurements of lime and Table 8. Classification of soils related to lime and active lime (Giardini [1986], mod.) Soil Total lime Active lime g CaCO3 kg-I Poor ...... < 25 < 20 Medium ...... 25- 100 21 - 50 Rich ...... 101 -250 51 - 100 Excessive ...... ------>250 > 100 Soil organic matter in Italy is quickly mineralized, because of the typical climatic conditions of the Mediterranean area. Therefore, classification of soil organic matter content must be adapted to the particular situation, and a value of 150 g kg- ' is generally considered as good. There is also a general agreement on evaluating cation exchange capacity, which is considered medium at 10-20 cmol(+) kg- . 324 4.2 Available nutrients Today in Italy, four different approaches are used to interepret soil analyses of available nutrients and to make fertilizer recommendations. The first uses directly the well-known Mitscherlich's equation: Y= Ym I I-exp[-cI - (d+x)] 1 (3) where Yni is the maximum yield, c a parameter related to the efficiency of fertilizer, Y is the expected yield obtained by application ofx units of fertilizer nutrient to a soil containing d, the amount of < 4.3 «Demetm I More functionally, many authors suggested intermediate approaches to soil analysis interpretation, applying variable sufficiency levels of available nutrients, computed as a function of other soil parameters, especially for potassium (Dolland Lucas [1973]; Henkens [1977]; Buchner and Sturm [1980]; McLean and Watson [1985]). Some procedures for soil analysis interpretation, as the French 326 both in use of modified norms and in integration between soil analyses and environmental and agronomic data (Figure 2). Crop gonmcCrop Si management history (rotation) Environment analyses Yield Modified L goal norms Nutrient Soil requirements fertility Optimum fertilization Report for crop rotation on soil Figure 2. Flow-chart of procedure 'Demetra® Lu for soil analysis interpretation and fer- tilizer recommendation. Considering the increasing importance of <"Demetra I)> it is worth giving more unpublished pieces of information about this procedure. The first step is the determination of fertilizer requirements ofa crop rotation in the specific environmental situation, apart from soilfertility. This is made considering crop requirements, including crop removal, crop nutritional ca- pacity and efficiency of fertilizers. The yield goal is calculated on the basis of past yields, considering field characteristics and management too. The main parameters considered are shown in Table 9. Table 9. Main environmental and agronomic data required by The second step is the calculation of required values of available nutrients («modified norms ), which depend on other soil characteristics, as reported in Table 10. In other words, reference parameters differ soil by soil and they result from specific functions. 327 Table 10. Main soil characteristics considered by «Demetra b) to calculate modified parameters (Perelli [1987]) Considered soil characteristics Nutrient pH Clay Lime O.M. CEC P K Mg Ca Nitrogen ...... + + + Phosporus ...... + + + + + + + Potassium ...... + + + Magnesium ...... + + + Calcium ...... + + + Sodium ...... + + + + + Iron ...... + + + + + Manganese ...... + + + Zinc ...... + + + Copper ...... + + + Boron ...... + + +4 Such functions were derived, on an empirical basis, from information coming from different sources (Perelli [19851) and are updated frequently, as soon as new data are available. For instance, the reference value for available phosphorus (PR) in soils with pH ! 6.5 (Olsen) is calculated in relation to the organic matter (OM), active lime (L) and clay (C) content in soil, by the equation: PR=a+b. OM 2 +c. OM+d exp(e- L)+f- C (6) where a, b, c, d, e, f and g are parameters, with values depending on the soil types and the environmental area. For exchangeable potassium, in the eastern part of the Po Valley, the refer- ence value K R is a function of the exchangeable magnesium (Mg,,) and cation exchange capacity (CEC): KR=38+0.221 - Mgc +4.1055 • CEC+0.025 . CEC (7) The quantity of nutrient in soil is considered adequate if it equals the reference value, with a ±20% degree of tolerance. In Table 11 are reported different classifications derived by equation (7). Table II. Example of different classifications for the exchangeable potassium calculated from equation (7), in function of the cation exchange capacity and the exchange- able magnesium (mg K kg-') C.E.C. Mg Exchangeable potassium (cmol kg-') (mg kg-') Very low Low Medium High Very high 10 100 <40 40- 80 80-120 120-160 > 160 10 200 <50 50-100 100-140 140-190 >190 10 300 <60 60-110 110-170 170-230 >230 30 100 <60 60-130 130-190 190-260 > 260 30 200 <70 70-150 150-220 220-290 >290 30 300 <80 80-160 160-250 250-330 >330 328 In the third step, the ratio between soil analyses (S) and the reference value (R) allows to calculation of a multiplication coefficient K with the following equation (cf. Table 12): K= 1-0,9 • log, (S/R) (8) The coefficient K is used to modulate fertilization, multiplying the fertilizer requirements of crops, as determined above. Table 12. Multiplication factor for fertilization, as function of the ratio between soil ana- lysis and reference value, calculated from equation (8)and used by ( Further improvements are made to exclude too high or too low values and to fit fertilization to the rotation, paying attention to the specific require- ments of different crops. Finally, fertilizer recommendation is improved by considering potassium fixation and farm management, in particular organic manure and the number of top dressings. Obviously, « 4.3.1 Effects of fertilizer recommendations with tDemetra 1o Before speaking about the effects of soil analyses and fertilizer recommenda- tions it is important to remember that generally in Italy only the better farmers, which have better yields, request soil analyses. Therefore the returns from fertilizer recommendations are very feeble in terms of yield. More frequently benefits are appreciable on crop quality: sugar content in beets and grapes, commercial quality of tobacco, reduction of bitter pit in apples etc. Fertilizer recommendations often affect fertilizer consumption, especially for phosphorus, which in Italy is applied in large amounts (cf. Table 1). Table 13 reports some data on fertilization of grain corn in 1000 soils in an area of about 3000 ha in the plain between river Adige and river Tione (south of Verona). 329 As shown by Table 13 there is a remarkable variation in fertilization after soil analyses: we had no variation in yields, but the consumption of fertilizers was changed, with a total increase in studied area of 87.3 t of K20 and a total decrease of 302.4 of P20 5, which according to Boswell et al. [1985], is equivalent to a saving of 11.3 • 1012 J, or 3.8 GJ/ha. Table 13. Fertilization of grain corn in 1000 soils in plain between river Adige and river Tione in 1988: soil analyses results, fertilization used by farmers and fertilizer recommendation calculated with Fertilization kg P20 ha' kg K 2 Ohat Farmer Demetra I Farmer Demetra I M inimum ...... 80 0 0 0 Maximum ...... 230 250 210 250 Range ...... 150 250 210 250 M ean ...... 156.3 55.6 123.7 152.8 Standard deviation ...... 44.7 59.1 55.2 55.2 Number of samples - 0- 50 kg ha ' P20 5 or K20 .. 0 584 119 69 50-100 kg ha' P205 or K2O .. 167 212 251 106 - 100-150 kg ha ' P20 or K20 .. 325 134 297 297 - 150-200 kg ha ' P205 or K2O .. 335 48 270 363 200-250 kg ha-' P2Os or K20 .. 173 22 63 165 Differences kg P205 ha- kg K20 ha-' M inimum ...... -230 -200 M aximum ...... + 150 +230 Range ...... 380 430 M ean ...... - 100.8 + 29.1 Standard deviation ...... 77.4 80.9 330 Table 13. (continued) Number of samples -250to -200 kg ha-' P20 5 or K20 100 2 - 200 to -150 kg ha- P205 or K20 218 20 - 150 to -100 kg ha-' P 0 or K20 245 43 - 2 5 -100 to - 50kgha ' P2O 5 or K20 223 108 - 50 to 0kg ha P20, or K20 106 218 0to + 50kgha- P 20 5 or K20 69 243 ± 50 to +100 kg ha-' P20 5 or KzO 30 158 - *+100to +150 kg ha P205 or K2O 9 145 +150 to +200 kg ha P20 or K20 0 54 +200to +250 kg ha - ' P20 or K20 0 9 4.4 Soil analysis interpretation in Greece and in Spain In Greece, there is not enough experience in soil analysis interpretation and therefore fertilizer advice is generally made on the basis of foreign experience. The same applies in Spain and other Mediterranean countries. 5. References Boswell, E C., Meisinger, J. J. and Case, N. L.: Production, marketing, and use of nitrogen fertilizers. p. 229-292. In: 0. P Engelstad (Ed.). Fertilizer technology and use. Soil Science Society of America, Madison WI. Buchner, A. and Sturm, H.: Gezielter Dlingen. Verlangsunion Agrar. Frankfurt (Main) (1980) Calvet, G. and Villemin, P: interpretation des analyses de terre IPAS. Centre de Recherches SCPA, Aspach le Bas, France (1986) Commission of the European Community: Soil map of the European Communities 1: 1000 000. Luxembourg (1985) Consalter, A., Clamor L. and Giandon, P.: Situazione dei laboratori per analisi di terreni e vegetali in strutture regionali e provinciali. p. 395-417. In: SISS-Unichim. Atti delle Giornate di studio sull'analisi del suolo, Verona, September (1988) Cotennie, A.: Soil and plant testing as a basis of fertilizer recommendations. FAO Soil Bulletin 38/2 (1980) Cristanini, G.: Ricerche sull'ottimizzazione della nutrizione del tabacco. p. 30-47. In: Cooperativa Tabacchi Verona, Relazione annuale 1989 (1990) Doll, E. C. and Lucas, R. E.: Testing soils for Potassium, Calcium and Magnesium. p. 133-151. In: L.M. Walsh and J D. Beaton (Ed.). Soil testing and plant analysis. Soil Sc. Soc. of America, Madison WI. (1973) Eckert, D. J.: Soil test interpretations: basic cation saturation ratios and sufficiency levels. p. 53-64. In: J R. Brown (Ed.). Soil testing: sampling correlating, calibration and in- terpretation. Soil Sc. Soc. of America, Madison, WI. (1987) FAO: Fao Fertilizer Yearbook, vol. 36, Rome (1987) Federico Goldberg L. andArduino, E.: La valutazione della fertilitA, p. 511-530. In: P Sequi (Ed.). Chimica del suolo, Patron, Bologna (1989) Genevini, P L.: Le analisi di routine in Italia e all'estero. p. 384-394. In: SISS-Unichim. Atti delle Giornate di studio sull'analisi del suolo, Verona, September (1988) 331 Giardini, L.: Agronomia generale. 3rd edition, Patron, Bologna (1986) Henkens, Ch.H. (Ed.): Adviesbasis voor bemesting van landbouwgronden. Bedrifs- laboratorium voor Grond- en Gewasonderzoek, Oosterbeek (NL) (1977) Maggiolo, R. and Schippa, M.: Campionamento dei suoli agrari: rappresentativita in fun- zione dei parametri da determinate in laboratorio. p. 418-459. In: SISS-Unichim. Atti delle Giornate di studio sull'analisi del suolo, Verona, September (1988) Maggiolo, R., Perelli, M. and Tarocco, M.: Use of vNA 1500)> in soil analysis. In: Atti delle Giornate di studio sulle metodiche di analisi elementare, Padova, June (1987) Maggioni, A., Pinion, R., Varanin, Z., Barbera, A. and Perelli, M.: Sulla dotazione nutri- zionale dei terreni dei Colli Euganei, nota la: estrazione e dosaggio di microelementi assimilabili. Agricoltura italiana, 113, 95-114 (1984) McLean, E. 0: Contrasting concepts in soil test interpretation: sufficiency levels of avail- able nutrients versus basic cation saturation ratios. p. 39-45. In: T R. Peck et at. (Ed.). Soil testing: correlating and interpretating the analytical results. Am. Soc. of Agronomy, Crop Sc. Soc. of America, Soil Sc. Soc. of America. Madison, WI. (1977) McLean, E.O. and Watson, M.E.: Soil measurement of plant-available potassium. p. 277-308. In: R.D. Munson et al (Ed.). Potassium in Agriculture. Am. Soc. of Agronomy, Crop Sc. Soc. of America, Soil Sc. Soc. of America. Madison, WI. (1985) McLean, E. 0Q,Adams, J. L. and Hartwig, R. C.: Improved corrective fertilizer recommen- dations based on a two-step alternative usage of soil test. 11. Recovery of soil- equilibrated K. Soil Sci. Soc. Am. J. 46, 1198-1201 (1982) Melsted, S. W. and Peck, T R.: The Mitscherlich-Bray growth function. p. 1-18. In: T R. Peck et al (Ed.). Soil testing: correlating and interpretating the analytical results. Am. Soc. of Agronomy, Crop Sc. Soc. of America, Soil Sc. Soc. of Amercia. Madison, WI. (1977) Mombiela, F, Nicholaides, J. J.and Nelson, L. A.: A method to determine the appropriate mathematical form for incorporating soil test levels in fertilizer response models for recommendation purposes. Agron. J. 73, 937-941 (1981) Nelson, D. W. and Sommers, L. F.: Total carbon, organic carbon, and organic matter. p. 539-579. In: A. L. Page, R. H. Miller and D. R. Keeney (Eds.). Methods of soil analy- sis, Part 2. Chemical and microbiological properties, 2nd ed. Am. Soc. of Agronomy and Soil Science Soc. of America. Madison, WI. (1982) Olson, R. A., Frank, K. D., Grabouski, P H. and Rehm, G. W.: Economic and agronomic impacts of varied philosophies of soil testing. Agronomy J. 74, 492-499 (1982) Olson, R. A., Anderson, F N., Frank, K. D., Grabouski, PH., Rehm, G. W. and Shapiro, CA.: Soil testing interpretations: sufficiency vs build-up and maintenance. p. 41-52. In: . R. Brown (Ed.). Soil testing: sampling correlating, calibration and interpretation. Soil Sc. Soc. of America, Madison, WI. (1987) Perelli, M.: Una nuova metodologia per l'interpretazione delle analisi del terreno elottimiz- zazione della concimazione. L'Informatore Agrario 41 (34), 81-95 (1985) Perelli, M.: Le analisi del terreno. L'Informatore Agrario 42 (6), 35-56 (1987) Perelli, M.: Fissazione del potassio e caratteristiche del terreno. p. 177-178. In: Atti del VI Convegno nazionale della SocietA Italiana di Chimica Agraria, Udine, September (1988a) Perelli, M.: Metodologie di interpretazione delle analisi del terreno. p. 460-480. In: SISS- Unichim. Atti delle Giornate di studio sull'analisi del suolo, Verona, September (1988b) Perelli, M., Maggiolo, R. and Corazzina, E.: La fissazione del potassio. L'Informatore Agrario 44 (22), 65-70 (1988) Pichtel, J.R., McLean, E. 0., Dick, W. A. and Esmaeilzadeh, H.: Refinement of quicktest methodology for improved potassium fertilizer recommendations. Agron. J. 78, 772-774 (1986) Setti, G.: Concimare col computer. Terra e Vita 29 (37), 57-59 (1988) 332 SISS: Metodi normalizzati per I'analisi del suolo. Bollettino SocietA Italiana di Scienza del Suolo, 10 (1976) SISS: Metodi normalizzati di analisi del suolo. Edagricole, Bologna (1985) SISS-Unichim: Metodi normalizzati di analisi del suolo. In litteris (1990) Tombesi, L., Moretti, R., Francaviglia, R. and Favola, G.: Climatologia e valutazione della produttivith. Supplemento agli annali dell'lstituto sperimentale per ]a nutrizione delle piante, Roma, p. 1-165 (1985) Unichim: Analisi dei terreni agrari - parte 1: metodi manuali. Manuale 145/1, Milano (1985) Unichim: Analisi dei terreni agrari - parte 11: metodi semiautomatici. Manuale 145/11, Milano (1988) Visintini Romanin, M.: La ricerca chimica agraria egli studi pedologici nel Friuli-Venezia Giulia. 10 aggiornamento dell'Enciclopedia monografica del Friuli-Venezia Giulia, p. 219-242 (1982) 333 Experience with Fertilizer Recommendations in USSR A.A. Sobachkin and . M. Bogdevitch* Summary Scientific principles and fertilizer use recommendations are worked out by some 300 research institutes, departments of universities and higher educational establishments, ex- perimental stations, consolidated into the geographical network of fertilizer trials. The fertilizer use recommendations are put into practice by collective and state farm specialists with the aid of more than 200 regional prospecting stations of the State Agrochemical Service, who conduct short-term field trials on fertilizer effect evaluation and systematic (once in 4-6 years) inspection of soils for potassium and other element contents. Potassium fertilizer rates are determined according to the content of mobile or exchange- able forms of potassium, established by various methods on different soil types. When working out recommendations, texture, acidity, organic matter content, content of non- exchangeable forms of potassium, biological peculiarities of mineral nutrition of agricul- tural crops, weather and agrotechnical conditions arc taken into consideration. Materials of agrochemical inspection of soils and the results of trials on potash fertilizer effectiveness are used for working out soil fertility models and constructing, on their basis, computer programs for calculation of rates, dates and methods of potash fertilizer usage together with all other means of agricultural chemistry. 1. Introduction USSR is the largest potash producer in the world. Deliveries to State agricul- ture in 1985-1988 averaged 6.8 mio. t K20 annually. About 300 research institutes, departments of agricultural higher educa- tional establishments and universities, experimental stations, situated in different regions and consolidated into the geographical network of fertilizer trials work out potash fertilizer recommendations and advise on other aspects of farming (organic manures and fertilizers, ameliorants, pesticides etc.). Fertilizer recommendations are put into practice by collective and state farm specialists with the help of workers of more than 200 regional prospecting stations of the State Agrochemical Service, where short-term field fertilizer trials are carried out and routine soil analysis is done every 4-6 years for potas- sium and other elements. * A.A. Sobachkin, All-Union Research Institute for Fertilizers and Agrology and L M. Bogdevitch, Byelorussian Institute for Soil Science and Agrochemistry, Minsk, USSR 335 2. Methods of soil analyses Available potassium content appears to be the best indicator of potash fer- tilizer requirement. Various methods are used according to soil type, of which there is great variety, and region (Table 1). Soils are found in 6 categories of K-availability: I-very low, ll-low, III- average, IV-higher, V-high, VI-very high and fertilizer rates determined ac- cordingly. Potassium, extracted by the appropriate reagent, appears to be the sum of water-soluble, exchangeable and some less readily soluble K, which constitutes an immediate soil reserve for plants. Often in practice these forms are replaced by the term ((exchangeable potassium . The effectiveness of potassium fertilizer increases as exchangeable potas- sium content in soil decreases. The correlation is most evident in sod-podzolic soils of fine texture and peats (Kulakovskaya andBogdevitch [1971]; Kulakov- skaya [1978j; Kulakovskaya, Knashis, Bogdevitch et al [1984]). K fertilizers improve nutrition of farm crops and considerably increase yield. Exchangeable potassium content is not always sufficient to predict potash fertilizer effectiveness. Then, additionally one determines that part of non- exchangeable potassium content, considered to be a potential nutritional reserve for crops. Extractants used are 2 N HCI (Pchelkin method [19661) deducting the quantity extracted by 0.2 N HCI or I N CO 3COONH 4 . Non- exchangeable potassium is also determined in 10o solution of hot HCI (Gedroyts [1955/), this extracts nonexchangeable potassium of soil colloids and fertilizer potassium fixed in the soil. In sodic-podzolic loamy and sandy loam soils of Byelorussia watersoluble potassium amounts to 5-15%, ex- changeable to 30-41% and nonexchangeable to 47-65%. Recently, as well as the capacity indices mentioned above K intensity in the soil solution is also measured (pK) or potassium potential (pK-0.5 pCa) (Korzun, Skoropanova and Melnitchenko [19841; Skoropanova [1988]). However, most practical recommendations are based upon exchangeable potassium content in soil. 3. Conditions for effective fertilizer application According to the results of numerous short-term trials by zonal stations responses in yield of grain crops to applied potassium fertilizer range within wide limits 0.03-0.45 t/ha with their participation in response from a full mineral fertilizer 6.6-31.9% and a return of 1.2-7.0 kg of grain per kg K20. Potassium accounts for 22.1-36.9% of response by potatoes to complete fer- tilizer (Derzhavin [1984]). In long-term trials response to K fertilizer increases over time. Thus, on the sodic-podzolic soil (of the Moscow region) with the uniform application of manure, lime, nitrogen and phosphorus fertilizers, average response of all crops in the first rotation was 5.6 kg grain equivalent per kg K20 applied and in the second 8.8 (Sobachkin [1984]). A higher crop 336 Table 1. Grouping of soils according to the content of mobile potassium (mg K20/kg of soil) According to Kyrsanoff BSSR According to According to According to According to According to Groups USSR Mineral Peat-boggy Maslova Egner-Riem- Egner-Riem- Chyrikoff Machygin soils soils Latvia [1985] Domyngo 1 0- 40 0- 40 0- 150 0- 50 - 0- 50 0- 20 0-100 ll 41- 80 41- 80 151- 250 51-100 0- 80 51-100 21- 40 101-200 11 81-120 81-140 251- 500 101-150 81-160 101-150 41- 80 201-300 IV 121-170 141-200 501- 800 151-200 161-200 151-200 81-120 301-400 V 171-250 201-300 801-1200 201-300 >200 >200 121-180 401-600 VI >250 >300 >1200 >300 >180 >600 -J return from potassium fertilizers is observed in the northern and western parts of the country, where sodic-podzolic and peat soils poor in potassium are predominant. According to the data of 950 field trials with winter wheat of the geographi- cal network the average return for 1 kg K20 on chernozems was about 2 kg of grain, on gray forest soils - 3.0 and on sodic-podzolic soils - 3.2. Liming of acid soils considerably improves the efficiency of fertilizer (pota- toes - by 6-24%, winter rye - by 20-40%, barley - by 50-80%, winter wheat- twice), therefore potassium rates are increased as pH (KCI) incrases up to the optimum level. The quantitative dependence of crop returns upon the exchangeable potas- sium content, soil acidity (pH), texture and rate of potassium fertilizer, estab- lished in more than 1000 trials were used for the purpose of working out im-. proved recommendations by computer in Byelorussia (Bogdevitch, Shatalova et al. [1983]; Detkovskaya, Bogdevitch, Kulakovskaya et al. 1986]). The aim is to regulate nutrient balance aiming for target yields and to increase soil fertility. Similar principles in determination of optimal fertilizer rates are adopted in other Republics. Long-term field trials with increasing rates of potassium fertilizer on different levels of soil potassium content have been carried out on soils of various texture. Three zones of fertilizer effect have been found: - a high and economic response; optimum exchangeable potassium content with minimum return; the zone of yield depression from excessive potas- sium content in soil. Introduction of intensive technology of grain cropping (split application of nitrogen, fungicides etc.) raised the return from potas- sium fertilizers and effective responses were obtained with the exchangeable K20 content in soil of 250 to 300 mg/kg of soil (Figure 1). The general relationships between the productivity of crop rotations and ex- changeable potassium content in different textured sod-podsolic soils are shown in Figure 2. The optimum content of exchangeable potassium greatly depends on clay content and CEC and is lower on sandy soils than on soils of heavier texture. If soil K content is close to optimum, the potash fertilizer recommendation is calculated so as to replace potassium removed in targeted crop yields. For low K soils, the recommendations are increased by 120 to 200% of crop removal (Table 2). Soils with above optimum K content should receive less potash fertilizer. All crops respond well to potash in Byelorussia. Therefore, K20 application per hectare of agricultural land increased rapidly from 17 kg in 1966 to 88 kg in 1988. Still more potash fertilizer is used on arable land, where the average rate of application reached 102 kg K20 (Table 3). Beginning from 1967 the sum of potassium applied with organic manures and fertilizers exceeded the crop removal and leaching. The exchangeable potassium level in soil has risen almost three times; a stable increase in arable land productivity is observed. This must be taken into account since crop response per kg K20 applied is less on high K soils (Table 4). Maximum response to fertilizers can be 338 achieved by applying calculated optimum rates of potash but all other nutrients must also be applied in the correct ratio. The results of changes in soil fertility on all fields, farms and areas are accumulated, stored and analysed by computer. There has been a tendency for variation in soil K level between farms to decrease (increase on low K soils, decrease on high) as a result of advice (Figure 3). 6.0- 1987 5.5" _5.0- Appl 2of4.5- potash fertilizers: 4.0 - Profitable C (35 1982-1984 Unprofitable 't, o- 3.0 2.5 10 150 200 250 360 350 460 K20 (mg/kg soil) Figure I. The yield of barley and potash fertilizer effectiveness depending on available potassium content in sodic-podzolic sandy loam soils. 7.0- /" - I -" T Loamy soil 6.0- I Sandy loam on moraine .C 5.0- 4.012004 3.0 K20 (mg/kg soil) Figure 2. Productivity of rotations as related to exchangeable potassium content in sodic- podzolic soils. 339 300- 1980 250- 42 1985 398 'S 200- E150- 14 Farms (% of total) o KH 0 50- M32 21 Figure 3. Variation of available potassium content in soil according to thegroups of farms in BSSR (1980-1985). Table 2. K fertilizer recommendations in % of crop removal (clayey loam soils) Crop Yield K20 content (mg/kg soil) (t/ha) < 80 81-140 141-200 201-300 > 300 Winter > 5.0 120 110 90 70 40 rye 4.1-5.0 140 130 100 80 50 3.1-4.0 160 140 110 90 - <3.0 180 160 120 100 - Flax 1.1-1.5 i8O 160 140 120 60 0.9-1.0 200 180 160 140 70 0.7-0.9 220 200 180 150 80 <0.7 240 220 200 180 90 Potatoes 31-50 100 80 70 60 30 Sugar 26-30 120 110 100 80 40 beet 21-25 130 120 110 90 50 <20 150 130 120 100 60 340 Table 3. Potassium balance in amble soils in BSSR Indexes Years 1966- 1971- 1976- 1981- 1986- 1970 1975 1980 1985 1988 K20 applied to soil (kg/ha per year) - with fertilizers ...... 39 72 101 95 102 - with organic manure ...... 20 26 33 37 42 Total ...... 59 98 134 132 144 K20 consumption according to yield removal and leaching (kg/ha) ..... 51 66 76 90 113 Balance (kg/ha) ...... 8 32 58 42 31 K20 content in soil (mg/kg) ...... 67 101 137 156 171 Change compared with the preced- ing period (mg/kg) ...... 10 34 36 19 Is K20 expenditure from fertilizers by 10 mg/kg of soil (kg) ...... 40 47 81 Il 103 Productivity of arable land (t/ha of fd unit) ...... 2.15 2.59 3.01 3.38 4.23 Table 4. Yield response to K fertilizers (kg per kg K20 applied) on sodic-podzolic sandy loam soils Crop Fertilizer/ Potassium Yield response crop price application Potassium content (mg K20/kg soil) ratio (kg K20/ha) 100-150 200-250 260-300 310-400 Barley 1.2 60 4.7 5.0 3.4 1.2 90 4.4 3.7 1.5 0.9 120 4.2 2.9 0.9 0.1 Potatoes 0.7 60 47 27 10 8 120 25 19 10 0 180 18 5 4 2 Perennial 3.3 60 8.3 8.8 5.5 5.0 grasses 90 10.0 10.0 0.3 1.8 (hay) 120 7.2 8.0 0 0.8 Results of soil analysis and the results of trials on the effects of potash are used to work out soil fertility models (Shishoff et aL [1987], Bogdevitch [19881; Ivanova [1989]). In working out the models, a significant number of indicators determining fertilizer effectiveness: physico-chemical properties of soils, biological peculiarities of mineral nutrition of agricultural crops, weather conditions, agrotechnical factors, balance of plant food elements and others are taken into consideration. The models seem to be a good basis for constructing computer programs for calculation of rates, dates and methods of potash fertilizer usage in combination with other kinds of fertilizer, lime etc. As examples of such programs one may mention those elaborated by 341 the Union Research Institute for Fertilizers and Agrology, the scientific and production system of agronomist-technologist (ARM-agronomist), the sys- tem of elaborating fertilizer usage plans in Russia - 4. Conclusion Data on crop response to potash fertilizer for natural and economic zones of the country have been established; the optimum rates, dates, methods of application and also conditions for increasing their efficiency have been deter- mined. Stabilization of the level achieved in potash fertilizer usage and increase in their effectiveness are only to be expected in the very near future in connec- tion with farming intensification and soil cultivation. As the available K con- tent of soils increases to the optimal level, potash fertilizer recommendations will be designed only to compensate for potassium removal by crops. The most important and complicated problem appears to be practical application of differentiated fertilizer rates, calculated by computer for each field with adjustment for soil type according to soil analysis with the aim of achieving the optimum level. 5. References Bogdevich, L M., Shatalova, T F and Ochkovskaya, L. R: Determination of the optimal dose of fertilizers containing potassium for grain cultivation on sod-podzolic soils. Poch- vovedenie i agrochimiya, 19, 142-156 (1983) Bogdevich, L M.: Model for increased fertility reproduction on sod-podzolic soils. Vestnik s-ch. nauki, No. 11,40-44 (1988) Derzhavin, L. M.: Influence of agrochemical conditions of the soil on the effectiveness of fertilizers containing potassium. Bjul. VIUA, No. 70, 6-13 (1984) Detkovskaya, L. P., Bogdevich, L M., Kulakovskaya, T N. et aL: Methodology to work out a scheme for the application of fertilizers on agricultural used soil. (vUrozhaj>) Mn., Gosagroprom BSSR, 92 pp., 1986 Gedrojc, K. K.: Chemical analysis of the soil. lzbrannye sochinenija, Vol. 2, Moskva, Sel- 'chozizdat, 615 pp., 1955 Ivanova, T.L: Prognostication of the effectiveness of fertilizers by the aid of mathematical models. Agropromizdat, 235 pp., 1989 Korzun, A. G., Skoropanova, L. S. and Mel'nichenko, E L: Valuation of the indicators of the activity, of the thermodynamic potentials and of the forms of potassium in the soil for the diagnosis of nutrition with potassium in frequently cut ryegrass. Poch- vovedenie i agrochimija, Minsk, Uradzhaj, Vyp. 20, 94-101 (1984) 342 Kulakovskaya, TN. and Bogdevich, L M.: The correlation of agrochemical soil quality with the crop yield and the effectiveness of fertilizers. Omnibus volume: Pochvovedenie i agrochimija, Mn.: Uradzhaj, Vyp. 8, 98-110 (1971) Kulakovskaya, T N.: Basis of the agrochemistry of the soil for the production of high crop yields. Mn.: Uradzhaj, 271 pp., 1978 Kulakovskaya, TN., Knashis V. Ju., Bogdevich, L M. el aL: Moskva o(Kolos>, edited by the member of the VASCHNIL academy TN. Kulakovskaya, 271 pp., 1984 Pchelkin, V.U.: Soil potassium and fertilizers containing potassium. Moskva ' 343 Coordinator's Report on the 4th Working Session Dr. Ch. Pieri, Sous-Directeur des Ressources Naturelles, CIRAD/IRAT, BP 5035, 34032 Montpellier Cedex I, France; member of the Scientific Board of the International Potash Institute Many speakers in this session clearly stated that a cropped soil should be consi- dered as a system. The functioning of such a system, the result of which is crop production, depends upon the interactions of its component parts, that is the soil constituents and the plant - more particularly the root system. The intensity of these interactions and the resulting yield are directly related to the supply of energy and other inputs such as water and nutrients. This implies that the response of a crop to fertilizer is not directly connected with the nutrient status of the soil but rather with the ability of the soil-plant system to operate adequately in processing the inputs. This means that the experimental approach remains, until now, the surest way of assessing fer- tilizer response at field level, by taking into consideration, through natural integration, site factors, technical and agronomical conditions (cropping sys- tems, cultural practices... ) which affect the physico-chemical and biological processes determining the yield. This complex situation is not easily accessible to system analysis and modelling. The main paper wisely reminded us that farmers who invest in fertilizer should receive good advice. Similarly, the field experiments in which agronomists invest their time are always costly and must therefore be genera- lised to other conditions. This implies that much care should be taken in designing experiments so as to provide the maximum of information. How can we reach that objective? First by clearly identifying the target and the simple questions the field experiment should answer: annual fertilization assessment including correc- tive maintenance dressings, soil dynamics studies ... require generally differ- ent experiment design as well as number of replicates. Similarly the nature of the experimental (and non experimental) variables monitored for the dura- tion of the field experiment is largely dependent upon the final aim. Secondly, the discussions made it clear that the o 345 different soil tests has been discussed. Exchangeable K is the most widely used indicator of potash requirement though, for soils rich in interlayer K with resultant high K-fixing power, other methods such as extraction by HNO 3 or KNaTPB appear more reliable in predicting crop response. Plant analysis is also extensively used. However it has been pointed out that the time required for the laboratory processing does not permit immedi- ate corrective action in the field, should a nutrient deficiency or imbalance be detected. Computerized systems are used to process field experimental data, with the aim of implementing a system of fertilizer recommendations based upon a more comprehensive approach. Several examples have been described. By using geographical information systems it has been shown that better assessment of K fertilizer requirement could be obtained by considering not only soil but site factors including geo- logical and climatic information (precipitation and temperature). This ap- proach leads to the concept of «site-related fertilizer recommendation > as used in the FRG by the KALIPROG software. More empirical approaches are used in France and Italy where «IPAS >and «DEMETRA 1) computerized systems are used. These systems as well as environmental and soil factors take into considera- tion the management conditions under which the crop is grown which greatly influence the plant's response to fertilizer. Similar systems are also under test in Eastern Europe and the USSR. However two points were emphasized in most of the papers. Firstly, whatever system is considered, computerized or not, no valid results will be obtained without a network of long term experiments set up on representatives sites. Secondly, there must be competent advisors with field experience in constant touch with the farmers who can adapt recommenda- tions to local conditions (climatic variations, crop rotation with demanding and less demanding crop, equipment availability ... ) also taking care to in- clude the farmers experience into the body of information available. Finally there have been interesting discussions about the ultimate aim of fertilizer recommendation which ideally would take into consideration not only short term profitability criteria, but also long term objectives. It has been shown that with the perspective of a sustainable agricultural development fertilizer recommendation may lead to a progressive building up of soil fertility. Better yield and more consistent responses to fertilizer are then experienced. On the contrary examples have been indicated where soil nutrients depletion occurs, inducing through physical and biological degrada- tions, a decrease of fertility which means a decrease in the ability of the soil- plant system to transform efficiently into biomass the energy and the inputs provided. It has to be stressed that long-term experiments are needed, as are compe- tent agronomists capable of convincing decision makers and equipped with adequate data and the words needed to put their message across - that long- term may mean «tomorrow). 346 Chairman of the 5th Session Prof. Dr. K. Mengel, Institute of Plant Nutri- tion, Justus-Liebig-University, Sadanlage 6, D-6300 Giessen, Federal Republic of Germany; member of the Scientific Board of the Interna- tional Potash Institute 5th Session Implementation of Fertilizer Recommendations with Special Reference to Potash Fertilizers 347 Channels to Reach the Farmer P Hotsma* Summary The functioning of the Dutch Agricultural Advisory Service is discussed. Attention is paid to thestructure of the service and the contribution made to its operation by the government, by farmers' organizations and by the private sector. The pathways of communication are dealt with, and soil testing in the Netherlands is introduced as an example of the contribu- tion farmers' organizations can make to the optimalization of agricultural productivity. Brief mention is made of impending alterations in the role the advisory service will play in future agriculture in the Netherlands. 1. Introduction The degree of success of an agricultural advisory service depends not only on the quality and efficiency of the services rendered but also, and perhaps more, on the degree of receptiveness of the farmers and growers involved. Farmers are in general only inclined to accept advice offered by an advisory service when: 1. they can recognize the recommended action and/or product as an essential part of their entire farming operation, and 2. they have good reasons to believe that, when the advice is followed, it will improve the profitability of their farming and will, therefore, raise their income. The knowledgeability of the average Dutch farmer is generally high. Most of them received at least three years of vocational training following primary school, but a sizeable number obtained an agricultural college education. On the average, a Dutch farmer runs a family farm together with his wife and perhaps a son or daughter. The farm is either owned by the farmer, or he operates it on a long-term lease. Farm sizes vary widely, but the income of a farmer is determined more by the intensity of the operation than by the size of the holding. Hired labor is rarely employed, except for labor-intensive activities. A vegetable grower with one hectare of land usually employs more hired labor than an arable farmer on a 60 ha farm. *P. Holsma, Information and Knowledge Center, P.O. Box 474,6710 BL Ede, Netherlands 349 Thirty years ago, in order to stay in business many small-scale farmers were forced to intensify their farming operations. Those who lacked the skill and inventiveness to adapt their farming practices had to abandon farming. The same holds for many older farmers lacking a successor in the family. As a result, on a population of 14.4 mio. people, the Netherlands has only 135 000 farm holdings, which is only 50076 of the number of holdings existing 25 years ago. Part time farming never became very popular: only 10% of the farmers draw more than 50% of their income from off-farm employment. Mixed farm- ing which traditionally was widespread in the Netherlands has practically dis- appeared. One of the reasons is that investments in specialized machinery equipment are so high that nobody can afford to purchase equipment needed for both mechanized arable and dairy farming. Farming in the Netherlands can be profitable, provided the farmer displays a great deal of dedication, business instinct, inventiveness, and skill. In many respects the Agricultural Advisory Service can help him or her to maintain or improve the profitability of his or her enterprise. In view of the fact that 2507o of the total value of Dutch export is derived from agricultural commodi- ties on a world-wide basis, making the country the second largest agricultural exporter, the Government has every reason to uphold the quality and effective- ness of the Agricultural Advisory Service. 2. The Agricultural Advisory Service Until very recently, this organization was entirely government-operated, and almost all services were free of charge. Depending on his type of farming oper- ation, each farmer can count on the services of a regional advisory office, mainly occupied by advisory officers specialized in his type of farming, e.g. vegetable growing, dairy farming, etc. Such officers, usually around 20 per office, are required to know all new developments pertaining to their com- modity. For instance, an officer specialized in intensive pig production should be well versed in pig nutrition, pig breeding, stable construction, ventilation techniques, manure handling and storage, economics, etc. To assist these advi- sory officers inkeeping abreast of all new developments, each regional advi- sory office employs around 5 specialists. Such persons are specialized in a certain discipline, such as fertilizer use or animal nutrition. They do not con- fine themselves to one crop or animal species, but cover, for example, the fertilization of cereals, vegetables as well as bulbs, or the nutrition of cattle, pigs as well as poultry. These regional specialists are, in turn, supported by national specialists, who are usually stationed at one of the governmental Research Institutes or Experimental Stations, and who serve as liaison officers. Their activities are not limited to one region, but cover the whole country. In addition to one agricultural University, the Netherlands has 19 Research Institutes involved in fundamental and applied research, mainly discipline- or problem-oriented, e.g. land reclamation or soil fertility and plant nutrition. 350 The results of research conducted at these institutes are of great importance for the development of agriculture, but individual farmers will feel its effect only indirectly. Research at these institutes used to be fully financed by the Government, but in the last decade funding by the private sector, e.g. indus- trial corporations, has gained importance. More practical, commodity-oriented research is conducted at II Ex- perimental Stations. Every segment of agriculture, e.g. dairy farming, glass- house vegetable growing, pomology, has such an experimental station. Fifty percent of the operation costs are paid for by farmers' organizations, which implies that farmers have a strong influence in deciding which type of research will be conducted at these stations. Each experimental station has at least two experimental farms. On these farms, new methods are tested at farm-scale level, with much emphasis placed on the economical feasibility of such new findings. Each farmer pays regular or occasional visits to such experimental farms, where he or she can discuss the results of the experimental work with the farm supervisor, the advisory officers and one or more liaison officers. 3. Farmers' organizations Each farmer in the Netherlands is a member of one of the three Farmers' Unions. These unions employ socio-economic advisory officers giving infor- mation on all non-technical matters, such as contracts, legal matters, book- keeping, taxes, inheritance, termination of farming, etc. Such service is free of charge for all member farmers and is subsidized by 500o0 by the Government. Thus, up till 1990 the Government absorbed all running costs of the above discussed technical advisory service, and 50076 of the costs of the socio- economic advisory service. Between 1993 and 2003, the Government will gradually lower its contribution to the technical advisory service to 50% of the total costs. Therefore, by the year 2003 the Government and the farmers' unions will each absorb half of the costs of both the technical and the socio- economic advisory system. 4. The private sector For a long time, the advisory officers of the Government and the farmers' organizations dominated the image of the Dutch agricultural advisory system. The division of tasks was clear: the fields and stables were the government advisory officer's working area, and the farmhouse was the domain of the socio-economic advisory officer. Farmers' cooperatives and private enterprises have traditionally made modest contributions to the total package of advice rendered to a farmer. There are indications, however, that in the coming years the contribution made by the private sector will increase. 351 Each Dutch farmer is a member of at least one cooperative supplying inputs such as fertilizers, feeds and pesticides. These cooperatives also employ advi- sors trained to supply information on the use of the inputs sold by their cooperative. Large producers and distributors of farm inputs may also employ agricul- turists trained to advise farmers on the use of the products sold by their em- ployers. In the period between 1935 and 1975 many farmers in the Netherlands became aware of the beneficial effects of fertilizers during visits paid to demonstration fields run by fertilizer companies. Since nowadays no farmer has to become convinced any more of the role of fertilizers in high-production farming, most of these demonstration fields were abandoned. Some large fertilizer manufacturing companies used to have their own research stations where applied and also fundamental plant nutrition research was conducted, sometimes in cooperation with Governmental Research Insti- tutes or the Agricultural University. Nowadays, Dutch farmers can still learn about the functioning of potassium in plant nutrition by visiting the Research Stations of the German and French Potash Mining Companies, both of which have sales organizations in the Netherlands. Agronomists employed by these organizations form the link between their companies and researchers in the Netherlands, Germany and France, on the one hand, and advisory officers and farmers in the Netherlands, on the other hand. About ten years ago, the fertilizer producing and distributing companies in the Netherlands joined forces in establishing the Netherlands Fertilizer In- stitute (NMI). This institute employs agronomists who are stationed mainly at government research institutes and experimental stations. Apart from pay- ing a basic fee for the regular advisory work of the NMI, each member company can place orders for specific research to be carried out by NMI researchers. The reports of these investigations are confidential, but are all approved by thedirector of the host institute or experimental station. In this way, the objec- tivity and scientific standard of the work are guaranteed. Through the NMI channels, the results of these investigations can be spread among advisory officers and farmers. The activities of this NMI are an example of cooperative efforts of private- sector researchers and governmental advisory officers directed towards use of an input commodity, namely fertilizers. An example of such cooperation regarding an output commodity, namely a crop, is to be found in the so-called Institute of Rational Sugar Production (IRS). The Netherlands has two sugar- manufacturing companies, a private one and a cooperative one. Together they finance the IRS, involved in both research and advisory work on sugar beet production. In order to avoid situations in which conflicting messages reach the farmer, the advisory service of this Institute works in close cooperation with the governmental advisory service. When opinions are found to differ, the two organizations meet for discussions aimed at drafting joint statements. A similar situation exists with respect to the above discussed NMI. There is always the risk of industrially supplied advice being biased, but the same can be said of governmentally supplied advice, which might be in- 352 fluenced by political motives. Advice is most convincing and effective when the messages coming from governmental and industrial sources are uniform. The best guarantee for such uniformity is equally strong and equally capable organizations showing a willingness to come to terms with each other. The last form of private-sector advisory work to be discussed here is the rather recent phenomenon of individual, independent advisors operating predominantly in glasshouse horticulture. The amount of money involved in growing a crop in a glasshouse is so large that growers are willing to pay for the advice of a specialist contracted to visit the holding every week or every two weeks. During such visits the advisor may act as a sounding board and a discussion partner for the grower who himself is an expert, but who has a desire to share his feelings and worries with a fellow expert. Due to the intensity and frequency of such contacts, each advisor can serve no more than 20-40 growers. The income of the grower is large enough to enable him to pay 507o of the specialist's salary. 5. Pathways of communication Pathways along which research findings reach the farmer can differ in length, as illustrated in the following scheme: a. researcher - farmer b. researcher - regional advisory officer(s) - farmer c. researcher - national liaison officer - regional advisory officer - farmer d. farmer - farmer Ad a. Many pathways exist along which Dutch farmers obtain printed infor- mation on farming. Farmers can subscribe to a daily newspaper aiming espe- cially at the farming community. Most of them subscribe to a weekly magazine on farming, and farmers' organizations and cooperatives provide their mem- bers with periodicals containing information on the latest developments in technology, agricultural policies and marketing trends. In addition, many brochures and folders arrive by mail. In the newspaper and periodicals, researchers can address themselves directly to the farmer. A researcher can also be invited to give a lecture on a topic considered of interest to a group of farmers. Such researchers might be staff members of Research Institutes or of the Agricultural University, but they might also be the supervisors of experimental farms, talking about the results of recent experiments on their farms. Every day at lunchtime, special weather forecasts for farmers are broadcast over the radio, followed by information on the likelihood of certain pests and diseases affecting certain crops. Systems such as Viditel enable a farmer to have the latest information on market prices on his television screen at the touch of a button. Furthermore, considerable effort is made by governmental services, farmers' organizations and commercial enterprises to introduce com- 353 puter systems through which farmers can receive help with the day-to-day management of their farms. Ad b. Regional advisory officers, both the commodity-oriented and the discipline-oriented ones, may have more time than the farmer to absorb the information supplied by the abovementioned communication media. The farmer may therefore prefer to discuss his problems and ideas with one or more of the advisory officers in his region. Alternatively, the advisory officer reads the findings published by the researcher and transfers the information to the farmer by means of a lecture during a meeting or in the form of an adapted report. Adc. Being stationed at a Research Institute or an Experimental Station, the discipline-oriented liaison officer is in an excellent position to learn about the latest development in his discipline. He or she organizes meetings with groups of regional advisors of the same discipline, or these regional advisors are brought up to date through reports written by the liaison officers. Ad d. One of the latest developments in the Dutch agricultural advisory world is the appearance of many farmers' and growers' study clubs. By now, about one thousand of these clubs exist. In each one of them, 10-40 farmers meet regularly. During these meetings, with often an advisory officer present, the participants compare results and talk about the possible reasons behind differences in financial returns. They organize study tours, and invite experts to discuss certain topics of common interest. Especially in the horticultural sector, these clubs have become very active. Undoubtedly, they have made major contributions to the rapid development noticeable in this sector. Particularly in these study clubs, it becomes evident that agricultural advis- ory work is most effective when the flow of information is a two-way, rather than a one-way affair. When a regional advisory officer finds it difficult to answer a question raised by a farmer, he may take the question to his colleague specialized in the discipline to which the question pertains. This advisory officer, in turn, may bring up the question during the next meeting he has with his national liaison officer. From there, the question may be taken to a researcher working in the Institute in which the liaison officer has his office. This bottom-up approach is a very effective alternative to the top-down approach which may dominate in hierarchical societies with farming com- munities consisting mainly of farmers and growers who were not trained well enough to serve as discussion partners for advisory officers. 6. Soil testing in the Netherlands Traditionally the three earlier mentioned Farmers Unions have together ac- cepted the responsibility of operating the Dutch soil testing service. There is one centrally located laboratory, handling more than 9007o of all samples taken in the Netherlands for soil testing purposes. Most employees of this laboratory are persons trained in collecting representative soil and crop sam- ples. They are stationed throughout the country, thus making sure that little 354 time elapses between the moment of sample taking and the moment at which the sample enters the drying oven in the laboratory. The system of employing professional sample collectors adds considerably to the total cost of soil test- ing, but the certainty of working with representative sample material is thought to justify the additional expenses involved. The sample collector inquires which crops the farmer plans to grow on the sampled parcel in the coming four years. With this information at hand, the outcome of the extraction procedures enables a computerized calculation to be made of the quantities of fertilizers and liming materials to be applied to each crop in the forthcoming four-year period. Many farmers subscribe to a system in which the sample collector automatically returns after four years to take a next sample. Until 1970, the results of the soil test were sent to the advisory service in the farmer's region and an advisory officer would visit the farmer to discuss the results of the soil test with him. In this manner, the soil testing system was a valuable incentive for periodical contacts between advisory officers and farmers. Gradually, however, this system of advisory officers serving as inter- mediary between the soil testing laboratory and the farmer proved to be too time-consuming and therefore had to be discontinued. Nowadays, the farmer receives the pertinent information on quantities of fertilizers and lime to be applied on a printed sheet. If he wishes to discuss the recommendations made with an expert, he can call upon the advisory officer in his region, who is specialized in matters of soils and fertilizers. 7. Assessment of fertilizer-N requirement A discussion of chemical extractants used in soil testing and of calibration methods employed to convert laboratory data into recommendations for quantities of fertilizer to be applied is considered to lie beyond the scope of this article. An exception will be made for the relatively new procedure of determining immediately available N still present in the soil profile at the end of the winter. Traditionally a standard soil test used to be confined to the nutrients P, K and Mg and to measuring soil pH. Available N was never determined until about 30 years ago it was realized that the quantity of inorganic N still present in the profile in late winter might be large enough to be taken into account when fertilizer-N recommendations are made. This quantity is dependent on I. the quantity of soil organic N present in the soil, 2. the quantity of fertilizer N applied to the previous year's crop, 3. the weather conditions during the autumn, 4. the absence or presence of a second crop grown in the autumn after the main crop was harvested, 5. the weather conditions during the winter. 355 In the Netherlands, the amount of winter precipitation can vary considerably, and especially this quantity exerts a strong influence on the quantity of inor- ganic N in the profile at the end of the winter. When this quantity is known, it can be subtracted from the quantity of fertilizer N to be applied to meet the N requirement of the next crop to be grown. To determine this available quantity of N (a combination of N0 3-N and NH 4-N) soil samples are taken in late winter to depths varying from 30 cm to 100 cm, depending on the rooting depth of the arable crop to be grown. This work is again done by the sample collectors of the soil testing laboratory mentioned earlier. Little time should elapse between the moment of sample taking and the moment at which the farmer receives the outcome of the test. For the extraction, use is made of field-moist soil. The extractant can be a KCI or NaCl solution, usually of I-molar strength. The method is known as the Nmin method. With information available on the Nmin value of his soil, a farmer is in a better position to determine the quantity of fertilizer N needed for optimum crop growth, thus avoiding leaching losses of inorganic N. Such losses should be avoided not only for financial reasons, but also for environmental protection. Remaining uncertainties in determining the correct quantity of fertilizer-N to be applied are caused by insufficient insight into the quantity of inorganic N that may become available during the growing season as a result of minerali- zation of soil organic N and into the quantity of NO 3 lost due to denitrifica- tion. Both quantities are weather-dependent, and summer weather cbnditions in the Netherlands can vary strongly from year to year. 8. The future The effective functioning of the Dutch agricultural advisory service as a link between the farmer and the agricultural research institutes was highly ap- preciated by all parties involved, as long as the Netherlands itself, the member countries of the European Common Market, and the overseas continents were ready to buy and consume the ever increasing quantities of agricultural com- modities produced by the Dutch farmer. Presently, the average Dutch farmer feeds 112 persons, which is the highest value of farmer's output in the world. Recently, however, this ever increasing productivity has contributed to the overproduction of agricultural commodities, such as milk, butter, sugar and cereals in the European Common Market. Moreover, heavy inputs of fer- tilizers and pesticides have added to the contamination of soils, surface waters and deep drinking-water reservoirs in the Netherlands. Intensive livestock production resulted in quantities of manure far larger than the country can cope with. As a result, agricultural advisory officers nowadays have to tell farmers to use less instead of more fertilizer to produce less, instead of more milk or wheat. Such restrictive policies might have negative effects on the farmer's income and put a strain on the relationship between the advisory officer and 356 the farmer. The existing overproduction also explains why governmental authorities show some reluctance to maintain the present manpower of the agricultural advisory service and make efforts to share the costs of this service with the private sector. In present agricultural research much emphasis is placed on investigations which may lead to e.g. further crop diversification, better conversion of the livestock feed, more efficient recycling of nutrients, manufacturing of useful products out of liquid manures, biological control of pests and diseases, etc. It is to be expected that in the future the Agricultural Advisory Service may play a different, but equally useful role in introducing the results of this in- novative research to the farming community. 357 Adjustment of Fertilizer Dressings to Achieve High Quality and Optimum Price for Potatoes and Sugar Beet Ht. Vandendriessche, M. Geypens and J. Bries* Summary The quality of agricultural products has recently assumed greater importance. Quality is almost as important as quantity in determining the return to the farmer. Domestic con- sumers and the processing industry sometimes pay more for high quality especially when quality has an important influence on the efficiency of technological processes. Sugar facto- ries and potato processors will pay more for high quality products. Potato quality is much influenced by nitrogen and potassium fertilizers. The most impor- tant criteria are dry matter content and susceptibility to bruising. Quality requirements depend on destination of the potatoes: table, chips, starch manufacture etc. Sugar beet quality determines the cost of sugar extraction. Internal quality can bejudged by extractability index. Good quality beet have high sugar content, low a-amino N content, and low potassium and sodium contents. Nitrogen, potassium and sodium fertilizers all influence extractability. Fertilizer advice should aim to achieve the optimum financial return to the farmer and must therefore take into account effects of fertilizer dressings on quality. 1. Introduction Recently, the quality of agricultural products has become more and more im- portant. Next to quantity, quality is for many products an important criterion in price determination. Not only the consumer but also the processing indus- try is sometimes willing to pay more for agricultural products of high quality, especially when technological processes require high quality products. Pota- toes and sugar beet are typical examples of crops for which, in addition to a basic price for quantity, processing industries are willing to pay more for high quality. * H. Vandendriessche, M. Geypens and J. Bries, Soil Service of Belgium, W. de Croylaan 48, B-3030 Leuven-Heverlee, Belgium 359 2. Sugar beet 2.1 Quality requirements for sugar beet The quality requirements for sugar beet are both external quality or tare per- centage and internal quality. 2.1.1 External quality Impurities accompanying the beet to the factory include soil, crowns, leaves, weeds etc. A high tare percentage increases the cost of transport and the storage losses. The processing of sugar beet in the factory and the extraction of sugar are also adversely affected by high tare percentages. Finally, high tare percentages give rise to environmental problems with effluent water and recovery of soil. The tare percentage due to crowns decreases during the growing season be- cause the proportion beet/crown increases. The factories accept beet with a high proportion of crown but do not pay for it because the structure and the chemical compositions of the crown are such that the sugar cannot be extracted profitably. In Table I the compositions of the crown juice and of the juice of beet without crowns are compared (Kadijk [19801). The general opinion is that processing of the crown yields as much sugar as the damage it causes. It is necessary to remove leaves and petioles, because they have a strongly negative influence on the quality of the juice. Table I. Composition of the juice of crowns and of beet without crowns (Kadijk [19801) Juice of Juice of beet crowns without crowns Percentage sugar in dry matter ...... 81.4 95.2 Colouring matter ...... 7.4 0.8 Potassium ...... 2.4 0.9 Sodium ...... 0.5 0.07 Nitrogen ...... 0.9 0.3 2.1.2 Internal quality The internal quality of the sugar beet is mainly determined by its sugar con- tent. The higher the sugar content, the higher is the internal quality of the beet. Secondly, internal quality is a function of extractability of the juice which is defined as the percentage of total sugar in the beet that can be ex- tracted as white sugar. The remainder of the sugar is lost in the pulp, the lime-sludge and the molasses. As is generally known, the quantities of potassium and sodium are of ut- most importance for the white sugar yield. The content of nitrogen is of in- direct importance because the negative influence of a high alpha-amino- 360 nitrogen content has to be eliminated in the production process by adding alkali. The alkali is added in the form of NaOH which leads to an increase in molasse sugar (van Geijn et al. [1983]). Low juice purity causes various problems in beet processing, such as corro- sion in the evaporators, the formation of more invert sugar in the thick juice, a lowering of the pH, more sugar lost in the molasses and the Maillard reaction (colouring of the thick juice). Various equations have been formulated to give a concrete form to the inter- nal quality of sugar beet. Together with the results of sugar beet analysis such equations are very important tools in sugar processing. The most interesting equations are those supplying information on the alkalinity coefficient (A.C.) and on the total loss of sugar contained in molasses (Sm). To characterize the rawjuice the A.C. is defined according to Wieninger and Kubadinow [19711 as: A.C. = (K + Na)/a-N where K, Na and a-N are expressed in milli-equivalents per 100 g beet. If the A.C. is high, there is no need to add alkali (NaOH). If the A.C. is low, addition of alkali is necessary to avoid the formation of invert sugar and the occurrence of corrosion with accompanying losses of sugar in molasses. The lowest acceptable limit of A.C. lies between 1.8 and 2.3, depending on the factory (Devillers [1983] and van Geijn et aL [19831). To calculate the quantity of molasse sugar (Sm), various equations are described in the literature. In the equations of Wieninger and Kubadinow, Akyar, and American Crystal, the nutrients K, Na and a-N are expressed in me/100 g beet and in the equation of van Geijn in me/100 g sugar: Wieninger and Kubadinow [19711 Sm= 0.349 (K + Na) for A.C. > 1.8 Sm=0.628 a-N for A.C. To compare the different ways in which the Sm value is calculated, the 4 equa- tions mentioned above were applied to data obtained in a field trial on a loamy soil in Ath (Belgium) in 1987. The field trial was a factorial design in which the influence of varying combinations of N, P and K applications on the levels of Na, K and a-amino N in sugar beet was examined. The field trial of Ath was a long term experiment from 1965 until 1987 (Vanderdeelen et aL [1985]). A review of the application of the 4 equations is given in Tables 2, 3 and 4 respectively for the factors nitrogen, potassium and phosphorus. The extractability index (P) is calculated as: P= 100-Sm 361 According to Dutch criteria (PAGV [1989]), the extractability index is evalu- ated as follows: P>88: very good P 88-85: good P 84-80: moderate P<80: poor Table 2. Composition of sugar beet and calculation of the % molasse sugar for a field trial on a loamy soil in Belgium (Ath, 1987). The data are means of four replica- tions (basal fertilization 160 kg P and 240 kg K/ha) N-ferti- me/100 g beet 0 Molasse sugar (01o)(Sm) lization K Na a-N sugar Wieninger van Geijn Akyar Am. Crystal (kg/ha) content 0 5.64 0.41 1.19 15.92 2.11 2.07 1.43 1.19 120 5.70 0.47 1.79 15.89 2.15 2.11 1.62 1.34 180 6.04 0.45 2.32 15.51 2.27 2.22 1.83 1.51 240 6.24 0.50 2.84 15.06 2.35 2.45 2.02 1.66 300 6.23 0.55 3.11 14.54 2.37 2.65 2.10 1.72 Table 3. Composition of sugar beet and calculation of the C/omolasse sugar for a field trial on a loamy soil in Belgium (Ath, 1987). The data are means of four replica- tions (basal fertilization 150 kg N and 160 kg P/ha) K-ferti- me/l00 g beet CIO Molasse sugar (%) (Sm) lization K Na a-N sugar Wieninger van Geijn Akyar Am. Crystal (kg/ha) content 0 3.83 0.79 2.39 14.64 1.61 1.58 1.51 1.22 120 4.93 0.54 2.31 15.33 1.91 1.87 1.64 1.34 240 6.04 0.45 2.32 15.51 2.27 2.22 1.83 1.51 360 6.63 0.46 2.33 15.42 2.47 2.42 1.94 1.60 480 7.06 0.42 2.27 15.40 2.61 2.56 2.00 1.65 Table 4. Composition of sugar beet and calculation of the Olcmolasse sugar for a field trial on a loamy soil in Belgium (Ath, 1987). The data are means of four replica- tions (basal fertilization 150 kg N and 240 kg K/ha) P-ferti- me/100 g beet C70 Molasse sugar (07c)(Sm) lization K Na a-N sugar Wieninger van Gelin Akyar Am. Crystal (kg/ha) content 0 6.04 0.48 2.12 15.38 2.28 2.23 1.78 1.47 80 6.08 0.46 2.21 15.53 2.28 2.24 1.81 1.49 160 6.04 0.45 2.32 15.51 2.27 2.22 1.83 1.51 240 6.11 0.52 2.23 15.20 2.31 2.27 1.83 1.51 320 5.81 0.49 2.38 15.54 2.20 2.15 1.81 1.49 362 2.2 Quality as a criterion for price determination The external and internal qualities of sugar beets are very important for the processing industry because they determine the (supplementary) costs of juice extraction. Consequently the processing industries give a financial bonus for good quality in addition to a basic price for quantity. Because of an over-production of sugar in the European Common Market a quota is laid down for every EC-country. This arrangement sets limits to the maximum production of sugar in each country. For example, the Nether- lands has a contingent of 872 000 tons A + B sugar and Belgium a contingent of 826 000 tons A + B sugar. The sugar produced above this contingent is called C-sugar and must be sold at the world price. The price received by the farmer is based on the quota and eventually on the world price for sugar if C-sugar is produced, but the real price the farmer receives is influenced by the bonus (or penalty) that the factory pays for good (or poor) quality. 2.2.1 Price determination in Belgium Farmers are paid on the basis of net beet delivery. In consequence the tare percentage has an influence on the financial return for sugar beet because the gross beet delivery has to be reduced for the tare percentage. The influence of the sugar content on the price is regulated as follows. The farmers receive 1785 Belgian francs (BEF) per ton sugar beet with 16076 sugar. For every 0.1% sugar more or less a financial bonus or penalty is calculated according to the following rules: Sugar content (%) Financial bonus or penalty/0.1% sugar 16 or more +0.9070 15 -15.9 -0.9% 14 -14.9 - 1.0% 13.5-13.9 - 1.2% 13.4 or less - 2.0% Up to the time of writing, juice purity did not affect the financial return. 22.2 Price determination in the Netherlands (PAG V [1989/) The tare percentage has an important influence on the gross financial result. According to the Dutch arrangements (Sugar Union (SU) and Central Sugar Company (CSM)) a 4% reduction in tare percentage results in a financial advantage of 60-180 Dutch guilders (about 30 to 90 US$) per ha (Figure 1). For beets with 16% sugar the farmer receives a basic price of 110 Dutch guilders per ton. For every percent sugar more (positive) or less (negative), 9% of the basic price more or less is paid. 363 CSM regulation a Tare 1500- Costs for tare 40% 40% 1000 35% 30% 500 25% 20% 10% 40 50 60 70 80 Net beet yield (t/ha) SU regulation b Tare 50% 1500- Costs for tare - (Dfl/ha) 45% 1000- 40% 35% 30% 500 25% 20% 10% _ __. 6% 40 50 60 70 80 Net beet yield (t/ha) Figure I. Influence of the tare percentage on the price paid to farmers in Dutch guilders (Dfl) per ha in the Netherlands, following the CSM- and the SU regulations. 364 Depending on the factory, a settlement is made to incorporate extractability of the juice in the sugar beet price. The extractability index is taken into con- sideration as 8% of the calculation for sugar content per % for every point of the extractability index (Figure 2). Extractability index + 500- Calculation for juice purity ) 93 +400- +300- 9 89 +200- + 100- +100 878 0 86 0- 85 " 84 -100- 83 -200- -300- 79 -400- -450- 77 40 50 60 70 80 Net beet yield (t/ha) Figure 2. Influence of juice purity on the price paid to farmers for beets, in Dutch guilders per ha. Basic price: Dfl 110 per t net beet with 16% sugar. Calculation for sugar content: 9% of the basic price per 07o. Calculation for juice purity: 8% of the calculation for sugar content per /Ol for every point of extractability. The conclusion can be drawn that depending on country and factory, the gross financial result of sugar beet growing depends on: - root production; - sugar content; - tare percentage; - extractability index; - quota; - world price for C sugar, if produced. 365 Table 5 compares the 2 price systems (The Netherlands, Belgium) for a beet production of 50 tons/ha with a tare percentage of 20%, a sugar content of 16076 and a juice purity of 86 according to the equation of van Geijn. Table 5. Calculation of the financial return of I ha sugar beet with a production of 50 t/ha, a tare percentage of 20%, a sugar content of 16% and a juice purity of 86 according to the equation of van Geijn. The comparison is made for 2 different price systems Price system Price determinants Belgium The Netherlands Gross production (1) 50 t/ha 50 t/ha Net production (2) 40 t/ha 40 t/ha Basic price 1785 BEF/t I10 Dfl/t A+B sugar, 16% (3) Deduction for tare (2) x (3) = 71400 BEF 160 Dfi Deduction for extractability 48 Dfl Compensation for pulp (2) x 130 F/t =5200 BEF - Financial return* 76 600 BEF 4 192 Dfl (*) I Dfl=18.85 BEF (5/4/1990) 2.3 Effect of fertilization on sugar beet quality The quality of sugar beets is influenced by weather conditions, soil type, fer- tilization, etc. Fertilization has important effects on internal quality (van der Beek [1989]); Boon and Vanstallen [1983]). The results of Tables 2 and 3 indi- cate that this is especially true for nitrogen and potassium. On the basis of results obtained in the field trial in Ath (Belgium) (see above), the influence of varying fertilizer combinations on the production and quality of a number of crops was calculated and expressed as index values (Table 6). The results of Table 6 are based on results obtained in the 1972-1986 period. In that period sugar beets were grown twice. Field tests with liquid manure indicate that the occurrence of too much mineral nitrogen in the subsoil (30-60 and 60-90 cm) in February-March has a negative influence on the sugar content of the beet. Experiments of the Soil Service of Belgium indicate that for each 100 kg of mineral nitrogen reserve in the soil (0-90 cm) the sugar content is lowered by 1.67%. The sugar content is lowered by only 0.55% for each 100 kg of fertilizer nitrogen applied at sowing time. For making recommendations on fertilizer N to be applied to sugar beet the N-index method is used in Belgium and the N-min method in the Nether- lands. In both methods it is customary to measure mineral nitrogen present 366 in the soil to a 60 cm depth in February-March. The fertilizer nitrogen to be applied for the highest financial return is calculated as follows: - on the basis of the N-index method for sandy, loamy, clay and lime soils: N rate= A-0.85 x N-index with A=257 for harvest before I October. A=264 for harvest between 1-15 October. A=271 for harvest between 16-31 October. A=277 for harvest after 31 October. - on the basis of the N-min method for sandy, loamy and clay soils: N rate= 220-1.7 x N-min The N-index is a function of several factors, such as the mineral nitrogen in the soil (0-60 cm), the soil carbon content (depending on soil type) and the quantity of organic manures applied. Table 6. Yields without N, P or K fertilizer, relative to those on the normally fertilized plots (N2PIK 2), on a loamy soil (field trial of Ath). Production on the (N2P2K 2) plots= 1000. Soil Service of Belgium 1972-1986 Nutrient not applied Frequency of cultivation in N P K 15 years Potatoes: tubers ...... <40% 70-79% 50-59% 3 leaves ...... <40% 60-69% 60-69% Sugar beet: roots ...... 60-69% 95-99% 60-69% 2 sugar ...... 60-69% 95-99% 60-69% crowns ...... 70-79% 10001o 70-79% Winter wheat: grain ...... 50-59% 95-99% 100% 5 straw ...... 40-49% 95-99% 90-94% Winter barley: grain ...... 40-49% 95-99% 95-99% 5 straw ...... <40% 90-94% 85-87% 3. Potatoes 3.1 Quality requirements for potatoes The consumer as well as the potato processing industry appreciate potatoes of a good external and internal quality. The meanings of these terms are given below. 3.1I External quality The following properties are important: 367 - the potatoes must be free of soil and sprouts - good looking potatoes are: - tubers free of Phytophthora, Fusarium and scab; - fresh tubers, without bruises and other damage; - regularly shaped; - uniform in size. 3.1.2 Internal quality Good quality potatoes have: - low susceptibility to bruising; - no subcutaneous discolorations; - no discoloration after cooking or baking; - no tendency to waxiness; - no hollow hearts, and crinkle; - a favourable dry matter content; - a good taste; - a limited content of reducing sugars. The dry matter content of potatoes fluctuates in general between 18 and 2707o. The dry matter content determines the cooking characteristics which are very important for the normal consumer. However also for the chips-, the French fried- and the starch industry, dry matter content of potatoes is an important quality factor, because it exerts a strong influence on the profitability of the production process and on the quality of the final product. For example, for chips processing a high dry matter content means a lower oil content in the chips. Moreover less water has to be evaporated off. The chips industry prefers potatoes with a dry matter content of 22-2501o. French fried potatoes have a dry matter content preferably lying between 20 and 2407o. The content of reducing sugars in potatoes must be kept low to avoid dis- coloration during baking and drying. The discolorations are due to the Mail- lard reaction between reducing sugars and amino acids. The sugar content is dependent on variety, on temperature during storage, and on storage dura- tion. For chips the best storage temperature range is 7-10'C, whereas for French fried potatoes and potatoes for the starch industry the best storage temperature is lower (5-8°C). As indicated in Table 7, applications of nitrogen, potassium and chloride can have positive and negative influences on the quality of potatoes (van Loon [19891). 368 Table 7. Quality aspects of potatoes which are positively (+) or negatively (-) influenced by ample application of nitrogen (N), potassium (K) or chloride (Cl) (van Loon [1989]) + I Several types of secondary growth ...... N (e.g. elongated tubers, bottlenecks). 2 Dry matter content ...... N, K, Cl 3 Tuber size (>50 mm diameter) ...... N N 4 Hollow hearts ...... N 5 Bruise susceptibility ...... N , K, Cl 6 Baking colour ...... N 7 Texture of French fried ...... N, K, Cl 8 Nitrate content ...... N , K 3.2 Quality as a criterion for determining potato prices Prices for potatoes are in general determined by the law of supply and de- mand. However, many farmers have a contract with the processing industry so that the (minimum) price is more or less fixed (guaranteed) for a certain quantity. This is especially the case for chips-, French fried- and starch pota- toes. Prices depend strongly on industry-farmer agreements, and are difficult to describe. In general the price is based on quantity, provided the potatoes meet a certain quality requirement. In the Netherlands, prices for table potatoes are determined by two arrange- ments - the AMV and the PF3 (De Jong [1985]). The AMV sets the price to be paid to the farmer for potatoes on the basis of their quality. Potatoes of low external quality due to green color, damage, etc. are set aside and calculated as tare percentage. The AMV pays attention to the following internal quality properties: bruise susceptibility, subcutane- ous discolorations, dry matter content, the percentage of waxy potatoes, hol- low heart and crinkle, and calculates the quality index. The PF3 quality evaluation for French-fried potatoes, is based on the same quality aspects, but is somewhat easier to calculate, especially for the percen- tage of waxy potatoes (De Jong [1985]). In Belgium, no general quality payment system exists. Efforts were made in 1989 to create a quality label for potatoes under the supervision of a promo- tional organization (N.D.A.L.T.P./O.N.D.A.H.). In 1989-1990 the so called «A-label> (Figure 3) was commercialized for the first time, but only for the variety Bintje. Until February 1990, 650 t A-label potatoes had been sold. A first evaluation of the A-label potato indicates that the labelling was only a limited success, both for the interested farmers and for the distributors. A stable and guaranteed price seems necessary to remunerate the farmers for the extra efforts made (Wullepit [1990]). In Germany the nitrate content of potatoes can be a quality aspect. 369 BELG______ Figure 3. Promotion of the M 4. Conclusion Technological processes require high quality basic products (sugar beet with a high sugar content and a high extractability or potatoes with a good external and internal quality) to result in high quality end-products with an optimal financial return. Processing industries give (sometimes) a financial bonus for good quality in addition to a basic price for quantity (yield). The question how fast farmers respond to price incentives and adjust their organic and mineral fertilizer application can, in most cases, be answered by the law of diminishing returns. Therefore, fertilizer recommendations must result in the highest financial return. They have to take into consideration the basic price for quantity and the financial bonus for good quality or the financial penalty for low quality products. 370 5. References Bosch, H. and De Jonge, P?: Handboek voor de Akkerbouw en de Groenteteelt in de vollegrond 1989. PAGV Publicatie nr. 47, 72-83 (1989) Boon, R. and Vanstallen, R.: Avis de fumure azote pour betteraves sucri&res sur base de lanalyse de terre. IIRB-Symposium «Nitrogen and sugar beet), Brussels (1983) De Jong, J.A.: De teelt van aardappelen. Reidingweg 5, Drachten, 320 pp., 1985 Devillers, P: R~le des composes azot~s dans la fabrication industrielle du sucre de better- ayes. IIRB-Symposium <(Nitrogen and sugar beet,>, Brussels (1983) Hobbis, J., Kysilka, M. and Halle, M.: Sucrerie Beige 101, 49-58 (1982) Kadijk. E. J.: De suikerbiet. Dronten, 1980. N.D.A.L.TPIO.N.D.A.H.: Aardappel, groente op z'n best. N.D.A.L.T.P., Brussels, 1989 Van Geijn, N.J., Gi/jam, L. C and DeNie, L. H.: Alfa-amino-nitrogen in sugar processing. IIRB-Symposium 371 Need of Improved Livestock Production from Farm Produced Feed and the Consequences for Fertilizer Application TF Gately and WE. Murphy* Summary The production of ruminants from grassland must become much more efficient if it is to continue to occupy 67% of the agricultural land of the world while the human population continues to grow. The potential for increasing output of milk and meat from grass is enor- mous. It is estimated that, within a few years, temperate grass production should average 15 000 kg DM ha -' year - on well managed farms. Milk or beef/mutton output from grass with- out supplementation should reach 12 5000 and 950 kg ha - ' respectively. The rates of fertilizer N, P and K used to sustain high grass yields for grazing and conser- vation must allow for the careful return of all animal effluents to the conserved grass areas in order to optimise the use of their nutrient content and protect the environment. It is calculated that the amount of N, P and K required, in kg ha- ', on an all grass farming system will be circa 325 N, 15 P and 55 K for grazing and 190 N, 15 P and 55 K for conserva- tion. Legislation to protect the environment may further reduce the amount of N and P which can be used. The proportion of animal products in the diet is a major factor in the amount of land needed to support a given human population. This is because livestock are very inefficient converters of food. The world has a finite amount of avail- able land and at present over 65 percent of it is used for grazing livestock. If people were satisfied to use a largely vegetarian diet, enough food could be grown to meet the needs of the world's population for the foreseeable future. Although man can meet his needs for energy and protein using an almost vegetarian diet, animal products are much sought after and it is difficult to resist the temptation to eat them when one can afford them. In the developed countries, pigs and poultry, which once were scavengers, are now fed cereal based diets i.e. on foods suitable for human consumption. Most cows, cattle and sheep are also fed some cereals or food suitable for humans in their diet. With the world population expected to increase for almost another 100 years, from about 5000 to 12 000 million, it seems unlikely that the luxury of using human food for animal production will be tolerated particularly when over 600 million people do not get enough to eat at present. * Dr. T E Gately and Mr. W E. Murphy, Teagasc, Johnstown Castle, Research and Devel- opment Centre, Wexford, Ireland 373 The advantage of ruminant livestock to man is that they can feed on herbage that people cannot consume directly. In addition to supplying milk and meat, ruminants also produce skins and wool. Animal versus crop production A comparison of animal and crop production up to the farm gate in terms of «support) energy - i.e. energy which is not derived from solar radiation Table 1. Efficiency of energy use (Leach [1976]; Spedding et a. [1976) Product MJ of energy in product per MJ of support energy used Wheat at farm gate ...... 3.2 Bread white, sliced, wrapped ...... 0.5 Milk at farm gate ...... 0.65 Milk bottled and delivered ...... 0.595 Table 2. The amount of crude protein, gross energy and the number of people who could be supported by the production from one hectare of land (Spedding et aL [1981]) No. of people whose annual requirement could be met Crude protein Gross energy (kg ha) (MJ ha- ) Proteina Energyb Crop: Cabbage ...... 816 105000 34 23 Field beans ...... 613 43466 26 9 Peas ...... 566 40805 24 9 Potato ...... 522 102080 22 22 Wheat ...... 469 69534 20 15 Sugar beet ...... 416 152469 17 33 Maize ...... 392 75905 16 17 Rice ...... 375 87768 16 19 Barley ...... 350 59274 15 13 Animal: Beef ...... 65 4796 3 I Lamb ...... 65 7486 3 2 Bacon ...... 105 14438 4 3 Rabbit ...... 292 13251 12 3 Chicken ...... 135 7056 6 2 Eggs ...... 74 4118 3 1 M ilk ...... 118 8770 5 2 a Taking protein requirement as 65 g per day i.e. 24 kg per year (assuming adequate protein quality). Taking energy requirement as 12.6 MJ per day i.e. 4599 MJ per year (assuming all the energy is available e.g. digestible). 374 but is obtained mainly from fossil fuels always shows up the inefficiency of animal production (Tables I and 2). The ((support> energy cost of plant production is incurred mainly as fertilizer, machinery and fuel but there are additional losses in the conversion processes involved in animal production. In developed countries, where 70 to 80 percent of food is now processed before consumption involving packaging, storage and distribution costs and where most of it is cooked, a farm gate comparison of the < Ruminant production The production of ruminants from grass and forage crops will have to become much more efficient if they are to continue to occupy 67% of the world's agricultural land (Table 3) while the human population continues to grow. No doubt, the possibilities for improvement are enormous. Table 3. Proportion of the world's agricultural land occupied by grass (FA.. [1974]) Agricultural land 106 ha 070 Arable land and land under permanent crops ...... 1475 33 Permanent grassland and meadows ...... 3005 67 Dairying Almost all the genetic improvement in dairy animals has been achieved within the last 25 years. Indeed, the modern dairy cow is a remarkable phenomenon. She can be regarded as a member of a herd whose average weight is 600 kg and lactation yield in 305 days is 7000 kg (Brooster et aL. [1987]). Within the herd some of her sisters will be yielding at least 50 percent more. In contrast, the average yield per cow of the less improved dairy herd, fed mainly on a forage diet, is more likely to be less than 3500 kg. Milk production can be increased very dramatically if the economic rewards for the producer are worthwhile. In Ireland, milk production almost doubled in a 15 years period (Table 4) accelerated by our accession to the EEC in 1973 which increased the value of milk and dairy products. This increase was despite the fact that there was a 50 percent reduction in the number of sup- 375 pliers and the land area devoted to dairying decreased by 10 percent (Mac Carthy [1988]). No doubt, this rapid increase in milk production, approaching approximately 5 percent per annum, would have continued for many years but for the imposition of milk quotas by the EEC in 1984. Table 4. Milk production and value in Ireland, 1970-1985 (MacCarthy 11988]) Volume of milk output Value of milk and (million tonnes) dairy products Year Total Manufacturing Liquid (Ir £ million) 1970 ...... 3.2 2.6 0.6 80 1975 ...... 3.5 2.9 0.6 236 1980 ...... 4.7 4.0 0.7 540 1985 ...... 5.8 5.1 0.7 965 Cattle and sheep Meat production, unlike milk, involves slaughter of the animal. As the animal gets older there is an increased maintenance burden that is not accompanied by proportionate increases in growth rate. Thus, meat producing animals be- come less efficient in the use of feed as they get older. Despite this inherent inefficiency there is enormous scope for improving the productivity of the ruminant meat producing animal. North [1987] estimated that the rate of genetic improvement in cattle could approach 2 percent per annum with the adoption and use of multiple ovula- tion and embryo transplant techniques, together with the possibility of sex determination. Table 5 shows his forecast of potential productivities from the application of new technologies mainly on the more productive farms in the year 2015 or just a quarter of a century from now. His forecasts do not consider Table 5. Potential production levels in 2015 (North 11987]) M ilk ...... 8900 kg cow Beef production - Number of calves ...... 1.8 cow -' year ' - Liveweight gain ...... + 20% Sheep production - Number of lambs ...... + 10% - Liveweight gain ...... + 15% Grassland utilisation - D airying ...... + 100% - Beef and sheep lowland ...... + 50% - Beef and sheep in less favoured areas ...... +20% 376 the use of bovine growth hormones, such as somatotrophin. The use of these hormones would increase milk yields per cow by 15 to 20 percent and also improve the liveweight gain and allow higher slaughter weights in beef animals and sheep. Taylor [1986] maintained that the overall food efficieny of beef production could be increased by up to 25 percent by the use of sex control to produce an (all female line. In Ireland, O'Keily (pers. comm.) estimated the beef production potential of well managed grassland, without the input of concentrates, to be circa 950 kg ha - ' year- '. Grass production The enormous potential for increasing milk and meat production will be fully exploited to provide these much desired animal products for a rapidly expand- ing world population if, and only if, the cows, cattle, sheep and goats of the future are fed on a mainly grassland diet. This is the challenge facing agricul- tural and animal scientists of the future. The emphasis will be on efficiency of utilisation of pastures to provide the complete or almost complete diet of the high performing animals. It is likely that complementary forage crops such as maize and fodder beet, will have a role to play in improving the seasonal supply of energy feeds and offering diet variety which may improve dry matter intake and reduce metabolic disturbances. Grass is only half as expensive as compound foods e.g. cereal based concen- trates, per unit of metabolisable energy (M.E.) Brooster et al. [1987] said that the approximate ratio of costs of M.E. is 4:2:1 for compound foods: conserved forage: grass. Dry matter (DM) yields of more than 40 t ha- ' year- , have been recorded commonly for C4 tropical grasses with 85 t ha-, year-' reported for Napier grass (Pennisetum purpureum) (Vincent-Chandler et al. [1959/). Such data compare with yields of C3 temperate grasses which rarely exceed about 20 t DM ha - ' year - ' (Wright [19781). In England, Thomas[1980/ found that over 75 percent of the energy require- ments of cows yielding 5800 kg milk year- came from grass and grass silage. Research on ICI farms between 1981 and 1984 (Reeve [1987]) showed the present potential of an all grass system (grazing plus silage) for milk produc- tion (Table 6). January/February calving cows were fed only high quality silage ad libitum from calving to turnout to grass in late-April. For the re- mainder of lactation the cows were grazed on perennial ryegrass (Lolium perenne) swards. The cows consumed an average of 12.6 kg silage DM per day whilst indoors and milk yields averaged 21.1 kg for cows and 16.0 kg for first lactation heifers. The milk was of satisfactory composition and there were no problems with cow health and fertility. Full lactation yields averaged 4681 kg for cows and 4006 kg for heifers. The lactation period was 277 days and it is likely that if this was 300 days the average yield for the cows sould have approached 5000 kg on grass plus silage only. 377 Table 6. Performance of cows on grass and silage alone (Reeve [1987]) Calving to turnout period (87 days) Silage intake (kg DM day - ') ...... 12.6 Milk yield (kg day- ') - co w ...... 2 1.1 - heifers ...... 16.0 Milk composition (%) - fat ...... 3.9 2 - protein ...... 2.96 - lactose ...... 4.7 1 Full lactation (277 days) Milk yields (kg) - cow s ...... 468 1 - heifers ...... 4008 Milk composition (07) - fa t ...... 3 .96 - p rotein ...... 3.16 - lactose ...... 4 .66 In Ireland, it was found that output per hectare could be incrased by about 7 percent with mixed stocking of cows and sheep or cattle and sheep. The grazing behaviour of the animals complement each other (O'Riordan pers. comm.). Dry matter (DM) The main factors which influence grass dry matter (DM) production are spe- cies, cultivar, temperature, rainfall, frequency of defoliation and of course fertilizer and animal manure use. Grazing animals can also affect yields through the effects of excreta, treading and differential grazing. Utilisation: For the efficient use of grazing swards tillers must be defoliated whilst they are still highly digestible in order to maintain good quality herbage. Researchers disagree on the merits of continuously versus rotationally grazed swards. In Ireland, O'Sullivan [19841 found a definite advantage in favour of rotational grazing at high stocking rates which was due to the shorter dis- tance the animals walked, thus expending less energy in attempting to main- tain intake where feed was limited. Stocking rate is used to match feed supply and animal requirement over the grazing season. The movement from pad- dock to paddock is based on sward height (Carton et al. [1989]). Conservation: The amount of grass conserved for indoor feeding and its quality is very important to the achievement of maximum animal output on intensive grass based systems in Western Europe but not so important in say New Zealand or Australia where an indoor period may not be necessary. The 378 technology is now available to make quality silage where the aim must be to reduce energy losses, which are inevitable during silage making, to less than 20 percent. Indeed, losses can be reduced still further with recycling of the effluent through the animal and reducing aerobic losses at the feeding face. Effects of fertilizer We have seen that in intensive grassland use in the future one can expect to see highly selected animals giving large yields of milk or meat. These will be grazed at high stocking rates on much improved pastures which produce large yields of a highly digestible feed for grazing or conservation. Fertilizers, especially nitrogen (N), posphorus (P) and potassium (K) will play a key role in affecting the yield, seasonality of production and quality of these high yielding swards. Indeed, without adequate inputs of these essential nutrients it will be impossible to fully exploit the potential of the animals and grass/ forage crops of the future. In that case the price of food to the consumer will increase and farmer incomes will decline. Nutrient uptake: Estimates of N, P and K uptakes in grazed and conserved swards giving a combined total of 15000 kg DM hat'year-', a yield which should be readily attainable in the near future, are given in Table 7. All but a small amount of these nutrient uptakes will have to come from fertilizers and animal manures (in the case of N, some may come from biological fixa- tion) if the soil reserves are not to be depleted. Estimates of the N, P and K removed in milk and calves are shown in Table 8 for a stocking rate of 2.5 cows ha (Agricultural Research Council[1965). The milk yield is assumed to be 5000 kg cow -on an all grass or grass/forage system. Table 7. Estimate of N, P and K uptake in a grazed and conserved grass sward at a yield of 15 t DM ha- year- Yield kg ha-' - (t DM ha ' year') N (3.0%) P (0.350Wo) K (2.5%) Grazed 9 ...... 270 31.5 225 Conserved 6 ...... 180 21.0 150 Total 15 ...... 450 52.5 375 Table 8. Estimate of the N, P and K removed per year at a stocking rate of 2.5 cows ha - (Agricultural Research Council [1965]) Yield kg ha- (kg ha-') N P K M ilk 12500 ...... 69 13 18 Calves 125 ...... 1.5 1 0.3 Total ...... 70.5 14 18.3 379 Nutrient availability: Estimates of the N, P and K available per year for a 15 t DM ha - grass sward yield used for grazing and conservation, due to animal excreta (outdoor), animal manures (indoor), atmospheric deposi- tions and free-living N fixing microorganisms are shown in Table 9. In the future, animal manures, properly stored and drilled into the areas set aside for conservation at the most favourable growth period to reduce run-off, leaching, volatilisation and denitrification losses and to optimise production may be expected to make available for growth at least 60 percent of their N and 95 and 9007o of the P and K respectively. It is assumed that N (and S) depositions in rainfall and atmospheric dust will decrease dramatically due to environmental restrictions on agricultural and industrial emissions. Indeed, because of these restrictions the need for S fertilizer use will assume even greater importance in the future. Table 9. Estimate of N, P and K available per year for a 15 t ha- ' grass yield due to animal excreta on grazed area, animal manures on conserved area and at- mospheric depositions (Table 7-Table 8) Yield kg ha-< - (t DM ha ' year') N P K Grazing - 215 days - in grass ...... 270 31.5 225 - in milk and calves ...... 41.5 8.3 10.8 Difference ...... 228.5 23.2 214.2 Available for grazing...... 45.7 (20%0)+6' 22 (950o) 193 (90%0) Winter feed - 150 days - in silage ...... 180 21 150 - in milk and calves ...... 29 5.8 7.5 Difference ...... 151 15.2 142.5 Available for conserved area* .... 90.6 (60070)+4' 14 (9507a) 128 (900o) 'Assumptions: Grazing - only one-fifth of area covered with excreta each year, about 8007o of N is lost. There is some loss of N, P and K on roadways at milking times and some K loss due to leaching. Loss of 8007o, N, 507o P and 1007 K. Conservation - with bandspreading or drilling at least 6001o of the N should be available. There is some loss of N, P and K on roadways and some K loss due to leaching. Loss of 40% N, 5% P and 10% K. 'Atmospheric depositions and non-symbiotic fixation of nitrogen (50070 available) Fertilizer requirements: The N, P and K fertilizer requirements for a 15 t DM ha- ' year grass sward for grazing and conservation assuming 2.5 livestock units ha - and an indoor period of 150 days are given in Table 10. This shows a total requirement, in kg ha - ' of 303 N, 16.5 P and 54 K, respec- tively for an intensive dairying system. Thus, approximately 6701o, 310 and 14076 of the total N, P and K requirement must be supplied as fertilizer. The 380 rates required for grazing and conservation, in kg ha- ', are 363 N, 16.5 P and 54 K and 212 N, 16.5 P and 54 K respectively. Table 10. Estimate of fertilizer N, P and K required for a 15 t DM ha year' grass crop used for grazing and conservation (Table 7-Table 9) Yield kg ha' (t DM ha-I year' N P K Grazed 9 ...... 363 16.5 54 Conserved 6 ...... 212 16.5 54 Total 15 ...... 303 16.5 54 It is clear from these estimates that intensive livestock production in the future, which will be grass/forage based, will require considerable inputs of fertilizer to optimise the use of solar radiation. This will be so despite the need to make maximum use of animal manures if only for the protection of the environment. It is likely that the emphasis will remain on intensive rather than on extensive systems of livestock production to make good use of all resources. Since the amount of animal manures and excreta depends on the stocking rate, which in turn is influenced by the annual dry matter production, it seems that fertilizer dressings of circa. 300 N, 15 P and 55 K in kg ha I will be necessary even where stocking rates and yields differ considerably from 2.5 livestock units and 15000 kg DM ha respectively. Of course, the N, P and K requirements will be much less on extensive systems of livestock production but these are unlikely to persist on arable land because of the de- mand for food in the future. Where there is sheep or mixed grazing of cattle and sheep the requirement for nitrogen will be less as there is a more even distribution of excreta and less loss of nitrogen from this source. Legumes The use of forage legumes in farming systems e.g. grass/white clover, is more suited to beef/sheep than to milk production. The mean annual nitrogen out- put of a white clover ley has averaged 185 kg N ha - (Sheldrick et al [19871). Since the presence of clover increases intake and digestibility, it has the capac- ity to make a worthwhile contribution to pasture improvement for sheep and beef systems, particularly where there is a demand for more herbage in the summer and autumn. In these legume based systems the need for fertilizer nitrogen would not exceed 80 kg N ha- ' i.e. nitrogen for first grazing and for conservation. The requirements for P and K could be larger in grass/clover systems than in grass only systems because of the inability of the clover roots to forage as successfully for nutrients as the roots of grasses. 381 Concentrates As demand for grain for human consumption increases there will be a move away from the present practice of high concentrate feeding of ruminants. If, in the short-term, some concentrates are fed at say 0.5 t livestock unit - ' this will reduce the total fertilizer N, P and K requirement in kg ha- ' by about 20 N, 5 P and 15 K respectively. Future use of fertilizer The aim in ruminant farming will be to maximise production from grass and forage crops. Fertilizers will be used to replace nutrient losses from the farm and to optimise the production of highly digestible herbage for grazing and conservation. The optimum economic amount of K will be used as it has no adverse effect on the environment. In the case of N and P inputs, it is likely that they will be used at about 90 percent of the optimum in order to reduce the amount of run-off, leaching, volatilisation or denitrification. This would suggest that the rates of N, P and K used in kg ha' will be circa 325 N, 15 P and 55 K for grazing and 190 N, 15 P and 55 K for conservation. To protect the ozone layer, reduce the acid rain problem and eutrophication legis- lation may be enacted which would further restrict the amounts of N and P which can be applied. This, of course, would reduce total output unless plants can be bred to utilize these nutrients more efficiently. Timing of N, P and animal manure applications will be critical in order to match, as precisely as possible, uptakes with inputs so that there is rapid removal of available nutrients from the soil. References 1. Agricultural Research Council., The Nutrient Requirements of Farm Livestock, No. 2. London; HMSO (1965) 2. Brooster, W4.H., Johnson. C. L. and Phipps, R. H.: Feeding the dairy cow in the '90's. J. of the Royal Society of England, 148, 168 (1987) 3. Carton, 0. T, Brereton, A. J., O'Keeffe, W. F and Keane, G. P: Effect of turnout date and grazing severity in a rotationally grazed reproductive sward. It. J. Agric. Res. 28, No. 2, 153 (1989) 4. F.A.D.: Production Yearbook 1973. 27, FA.O. Rome (1974) 5. Leach, G.: Energy and Food Production. IPC Science and Technology Press (London) (1976) 6. MacCarthy. D.: The role of grassland in animal production in Ireland. Irish Grassland and Animal Production Association Journal, 22 (1988) 7. North, J. J: Land use for food production. J. of the Royal Society of England, 148, 16 (1987) 8. O'Sullivan, M.: Tenth Richards-Orpen Memorial Lecture. Irish Grassland and Animal Production Association, Journal, 104 (1984) 382 9. Reeve, A.: Maximising production from forage. Milk Check Conference. January, I.C.L Jealotts Hill Research Station, U.K. (1987) 10. Sheldrick, R., Thomson, D. andNewman, G.: Legumes for milk and meat. Chalcombe Publications. Bucks, England (1987) II. Spedding, C. R. W and Walsingham, J M.: The production and use of energy in agriculture. J. agric. Econ., 27, 1, 19 (1976) 12. Spedding, C. R. W., Walsingham, J. M. and Hoxey, A. M.: Biological Efficiency in Agriculture, Academic Press, London (1981) 13. Spedding, C.R. W.: Energy use in the food chain. Span 27, No. 3, 116 (1984) 14. Taylor, St. C S.: Breeding for efficient use of food by beef cattle. SPAN 29, No. 3, 114 (1986) 15. Thomas, C: Factors influencing the potential of silage for milk production. Br. Grass- land Soc. Natn. Silage Competition Conf. National Agricultural Centre, Stoneleigh, April (1980) 16. Vincent-Chandler, J. Silva, S. and Figarella, J.: The effect of nitrogen fertilization and frequency of cutting on the yield and composition of three tropical grasses. Agron. J., 51, 202 (1959) 17. Wright, C. E.: Maximising herbage production. Joint Br. Grassland Soc. Br. Vet. Assoc. Conf. Intensive Grassland Use and Livestock Health. Berkshire College of Agriculture, February, pp. 11-23 (1978) 383 The Range of Various K-Fertilizers and their Agricultural Use H. Vis* Summary There are three main types of potash fertilizer: potassium chloride, sulphate and nitrate. These are available in various grades giving the farmer a wide choice. Some contain sodium and magnesium and are especially valuable for natrophilic crops like sugar beet and for improving the mineral balance and feeding value of fodders. Potassium sulphate has a low salt index, important if there is a salinity hazard and for chloride sensitive crops. Con- centration (06 K20 content) and physical condition in relation to application technique are also important to the farmer. Methods of fertilizer application (placement and timing) and treatment of problem (low K availability or high K fixation capacity) soils are discussed. 1. Introduction In following fertilizer recommendations, the farmer will choose the potash material which best suits his conditions (crop, rotation, spreading equipment, etc.). He may choose a straight potash fertilizer or a multinutrient (PK, NK or NPK) type, solid or liquid, chloride or non-chloride. Physical form (granu- lar) and packaging may be dictated by his equipment. Most important is price per unit K20 content. 2. Classification of potash fertilizers 2.1 According to the accompanying anion If we exclude potassium metaphosphate, potassium bicarbonate and potas- sium silicates which are little used in agriculture, potassium fertilizers com- prise 3 main types: - potassium chloride KCI (94.5% of world K-consumption) - potassium sulphate K2SO 4 (507o) - potassium nitrate KNO3 (0.507o). * H. Vis, International Department, Extension Service, SCPA, 2 Place du G6ral de Gaulle, BP. 1170, 68053 Mulhouse Cedex, France 385 All are water soluble. Potassium chloride. The CI anion is not an important nutrient, except for oil palms and coconuts which have a real chloride requirement (von Uexkiill [1971]). Excess of chloride proves detrimental to all crops, and damaging to specific ones. However, under normal climatic and soil conditions, KCI fertili- zation does not lead to hazardous levels of Cl. Potassium sulphate combines 2 major nutrients K and S in the sulphate form (SO4) which is readily available to the plant in the soil solution. In areas where sulphur dioxide air pollution has been drastically reduced and where highly concentrated N and P fertilizers no longer contain S, deficiencies in this nutrient tend to be widespread, especially in sensitive crops. Potassium nitrate is also a real binutrient fertilizer, associating 2 of the 3 major plant nutrients. It is the most expensive item among these three K-fer- tilizers. Its use is restricted to intensive horticulture in open fields or, often linked to hydroponics, in greenhouses. 2.2 According to associated auxiliary nutrients (Na, Mg) As potash fertilizers are recovered from salt deposits or brines (salt lakes), potassium is often associated with other cations of real value for agriculture: Mg, Na. Virtually, all possible combinations among K, Mg, Na in chloride or sul- phate form can be found. In Europe particularly, 4007o or 50% K2 0 potassium chloride, magnesiakainit, kieserite, double sulphate of potassium and mag- nesium have become popular grades for specific crops (natrophilic plants and grassland). - Substitution of K+ by Na on sugar beet: Though all physiological processes involved have not been elucidated, it seems that potassium may be replaced by sodium in its osmo- and turgor- regulative functions. There is evidence that Na stimulates leaf growth whilst K + improves growth of the storage root (Lindhauer[1989).Substitu- tion of up to half the K * by Na ' in experiments has had a positive effect on plant dry weight and sucrose content in the storage tissue (Mengel [1987]). But in osmotic functions, inorganics can also be substituted by or- ganic solutes (sugars, organic acids). The intensity and duration of photo- synthesis, i.e. the weather conditions, may override such nutritional effects. Field experiments and surveys in Great Britain show that on high yielding sugar beet dressed with 300 kg K20/ha, the application of 200-250 kg Na/ha improves sugar percentage by 0.307o (17% instead of 16.7%) and the sugar beet yield by 2.3 t/ha (Draycott et al. [1976]). European sugar beet growers have acknowledged the value of such experimen- tal results and have therefore adopted sodium chloride or sodium containing potash fertilizers in their fertilizer programme: for instance in U.K. the area 386 of beet receiving Na has increased from 23% in 1955 to 60% in 1985. In soils low in Mg, sugar beet benefits from application of Na-Mg K fertilizers. In the Federal Republic of Germany, a KCI fertilizer containing 40I7oK20, 6% [1987]).MgO and 8% Na2O is the standard K-fertilizer for sugar beet (Beringer Other natrophilic species (cabbage, spinach, tomato, oats) can accomodate Na containing potash fertilizers (Mengel [1989). - For grassland, the combined application of K with Na and Mg is recom- mended. This results in herbage with a well balanced mineral content (proportions of K, Mg and Na) with improved palatability and beneficial to animal health and fertility of livestock. Mean incidence Incidence of hypomagnesaemia increases of tetany I%) as Na content of herbage falls 7.0- 6.0- 5.0- 4.0 3.0- 2.0 1.0 0.0 0.00 0.05 Na 0.10content of0.15 pasture 0.20 0.25 0.30 0.45 (% dry weight) Figure 1. The relation between tetany incidence and pasture sodium (from Butler [1963). Ideally, the herbage should contain 0.18% Mg and 0.13% Na. Mineral balance is particularly important in springtime: it is considered wise to aim for a K:Na balance of 20:1 as wider ratios depress Mg digestibility (Figure 1). German trials have shown that dressing with magnesia-kainit can improve palatability and hence feed intake, especially in pastures containing species (e.g. Lolium perenne) that take up much Na (Beringer [1988]). 2.3 According to salt index Large areas of soils are affected by salinity, especially under arid or semi-arid conditions where evapotranspiration exceeding rainfall tends to accumulate salts in the upper soil layers, in the vicinity of the dense rooting zone. 387 The salt index represents the ratio of the increase in osmotic pressure of the soil solution produced by the fertilizer material to that produced by the same weight of NaNO 3 (=100). Potash fertilizers vary widely in this important criterion, as Table I shows. "rable I. Salt index of potash fertilizers (after Zehler et aL [19811) (per unit of product) Sodium nitrate (reference) ...... 100.0 Potassium chloride (60% OK20)...... 116.3 Potassium nitrate (47% K20, 14% N) ...... 83.6 Potassium sulphate (50% K20) ...... 46.1 Sulphate of potassium and magnesium (2207o K 20, 1010 MgO) ...... 43.2 Susceptibility to salinity varies with species. Table 2 shows for various crops the salt concentration - measured by electrical conductivity - at which yield is reduced by 5007o (Zehler et al [1981]). In parallel, a scale for Cl-tolerance is shown. Table 2. Salt tolerance and Cl tolerance of various crop plants Salt tolerance poor moderate good EC value for 50% yield 2 4 8 12 16 reduction (mmhos/cm) Salt content (Ol0dry soil) 0.2 0.35 0.65 - fodder plants lucerne clover fodder beet - arable crops sunflower rye cotton sugar beet barley wheat oat maize - vegetables radish potato tomato spinach beans carrot cabbage cucumber onion lettuce - fruits apple grapes cherry peach apricot oranges Cl tolerance very poor poor moderate good tobacco flax lucerne barley red clover wheat, oats sugar beet beans potato tomato cabbage melon grape onion 388 2.4 According to nutrient concentration Farmers will consider as a deciding factor in their choice the price of fertilizers delivered to the farm gate, in terms of price per unit of nutrient. The latter results from the ex-mine price and the transportation costs which depend on the distance and the mode. Accordingly potash fertilizers with low content of nutrient will prove too expensive to farmers if they have to be transported over long distances, as the following simple example shows for low and high analysis KCI: Ex mine price Costs- for transport Delivered price bulk and handling to the farm Potassium chloride /t prod. It K20 It prod. /t K20 40% K,0 ...... $40 $100 $56 $ 96 $240 60%o K,0 ...... $65 $108 $56 $121 $201.70 These costs include the bulk transportation over 2000 kms (efficient unit train), the trans- loading into a regional warehouse for bagging (the throughput, storage, bag+bagging costs) and the final distribution to farmers by trucking within a radius of 150 kms. Regarding price, the farmer will make his choice on the basis of cost per unit K20 delivered to the farm. The low grade material is much cheaper per t product but only a little cheaper per unit K20 ex-mine, but, by the time it is ready for application in the field, having incurred charges for transport, transloading, bagging and delivery to the farm, the low grade material is 19% more expensive per unit K20. The choice is obvious. 2.5 According to physical form and fertilizers distribution systems Firstly fertilizers can be applied in solid or fluid form. Wherever the industry is selling urea ammonium nitrate solutions or NP solutions based on superphosphoric acid and ammonia, NPK solutions can be prepared with highly water soluble potash grades. Another alternative con- sists in more concentrated NPK suspensions where, through addition of ben- tonite, KCI crystals are kept in suspension in a saturated solution. Liquid fer- tilization offers a quick and effective option with minimum losses of nutrients and the possibility to add pesticides and/or micronutrients. However NPK clear solutions or suspensions (more concentrated) are corrosive and suscep- tible to cold conditions (crystallization occurs between 0°C and -25°C ac- cording to grades). For special crops, fertigation or hydroponics tend to expand in high technol- ogy agriculture: both of them demand very pure potash grades (chloride, sul- phate or nitrate) virtually 100% water soluble (no insoluble impurities). Still the most widespread fertilization, especially in Europe, is applied as solid in the form of compound, bulk blend or straight fertilizers. 389 The production of PK or NPK compounds involves the powder form of potash, called standard grade. The bulk blended PKs or NPKs and the straight potash fertilizers use granular grades: farmers are increasingly particular as regards granulometry. As a result, more and more so-called screened granular (2-4 mm) is being sold in Western Europe. As Figure 2 shows, some countries are using most of their potash in granular form; but, as a whole, granular is gaining importance. One must admit that for small to medium holdings growing various crops and having different soil types - the fertility of which is increasingly monitored by soil tests - using straight fertilizers or NPK blends makes it possible to adjust optimum fertili- zation almost to each individual field. % Total K20 90 USA 89 % 80 70 60 50 40 Average EEC 38 % 30 20 0 0 IRL NL UK UEBL I FRG F E GR P DK Figure 2. Proportion of potash consumed in form of granular among EEC countries in 1987/88 (source: SCPA Statistics Dept.). Finally, the choice of a specific form of fertilizer - fluid or solid multi- nutrients, alternatively straight fertilizer - will depend much on the local facil- ities available to farmers. If the retailer has specialized in one type of fertilizer with competitive prices and good service, most farmers in his area will adopt his products. 390 3. Guidelines for the choice of K-fertilizers and for potash dressing management The composition of potassium chloride and potassium sulphate in the most commonly available grades is given in Table 3 (Zehler et al [19811). Table 3. Composition of usual potash fertilizers* 070 Kainit Potassium chloride Potassium Sulphate of sulphate potash magnesia 40% 50% 60% K20 15 38-42 48-52 57-61 50 (47-52) 28 (22-30) MgO 15 3.5 2.5 1.5 2 8-12 Na2O 24-29 10-18 5-8.5 1-4 - CI 39-43 41-53 42-50 47-54 0.5-2.5 0.5-6 S 3-9 3 2 I 18 16-22 * The Table states ranges of content in materials from different mines. Setting aside potassium nitrate, a specific NK bi-nutrient fertilizer reserved for intensive special crops, the criteria that should guide the choice among potash fertilizers can be summarized as in Table 4. If crop quality is important, potassium sulphate is advised. Besides tobacco for which chlorine-free fertilizers are strictly necessary, a number of quality crops requiring high K dressing cannot tolerate a high Cl content without having their quality (for fresh consumption or processing) impaired. This applies to potato, tomato, melon, strawberry, citrus, grapes and glasshouse grown vegetables for which potassium sulphate is highly advisable (Loud [1978] and Vis [1989]). In Western Europe, the wide range of potash products available on the market helps meet a multifaceted demand. The structure of this potash demand and its evolution can be seen from Table 5. Whilst the ratio between chloride and sulphate has remained static, a his- torical trend in favour of increasing use of granular MOP is outstanding. Standard procedure is to broadcast potash before or during initial cultiva- tion. There are, however, some conditions under which it is advisable to adopt a different procedure and, here, practice especially as regards placement and timing has as much effect on crop yield and profit as does the rate of applica- tion. These situations are reviewed in Table 6. 391 Table 4. Criteria for choosing K fertilizers (after von Braunschweig [1986]) Salt and Request for High requirement for Thrget of Crop quality Spraying Distance chloride high improved (high value solution from mine tolerance analysis Mg Na S nutritional crops) for foliar to user fertilizers quality application Low High (fodder) Short Long Kainit X X X X X Potassium chloride 40% K2O X X* X X 5007o KO X 6007o K20 X X X Potassium sulphate X X X X X X 50% K20 Sulphate of potassium X X X X and magnesium If derived from Mg containing salt deposit such as kainit. Otherwise (sylvinite), no. Table 5. Structure of the West-European potash demand (in %) Grades 1978 1982 1988 Up to 40% K O ...... 9.7 10.0 5.9 KCI standard (40-60% K20) ...... 64.5 62.0 57.2 KCI granular (40-6007o K20) ...... 12.4 15.0 22.7 Sulphates (50% K20) ...... 8.3 8.0 9.0 Industrial grade ...... 5.1 5.0 5.2 Source: SCPA Statistics Department Table 6. Specific potash dressing management (after Gros [1974] and Mengel [1987) ANNUAL ARABLE CROPS Soil and climatic Soil with low available K or high Low characteristics K-fixation capacity rainfall area Potash placement Band applica- OR Strip ap- Additional ap- tion (I or 2 plication plication for Deep bands 5-10 cm (10-15cm fast growing placement below the seed wide and high K row and 5- strips) on requiring 10cm to the the sur- crops:side side) face of dressing the soil Application time At planting Incorporation Flowering stage At plowing with tillage or during operations vegetative growth INTENSIVE FODDER- PLANTATION CROPS (fruit trees, grapes) Soil and climatic Good rainfall or irrigation Fine textured (clay) deep soils characteristics Potash placement Broadcast Deep placement (30-50 cm according to species and soils) Application time Split: first application in early Abundant dressing before spring planting followed by regular second application after maintenance dressings the first cut * Besides this top dressing, a good K dressing ploughed in before planting is essential, especially for deep-rooting species (lucerne) 393 4. Conclusion Potash fertilizers are often available in a large range of products. Moreover, the multiplicity of existing fertilization systems (fluid, solid, straight or blends/compounds in bulk or in bags) that are encountered in Europe and the high yields that can be achieved with adequate management, demonstrate that farmers can choose convenient and effective options for their potash fer- tilization. 5. References Beringer, H.: Recent data about P-, K- and Na fertilizer application to sugar beet in Central Europe. Proceedings 50th I.I.R.B. Winter Congress, 103-133 (1987) Beringer, H.: Potassium, sodium and magnesium requirements of grazing ruminants. Potash Review, No. 2, 1-10 (1988) Braunschweig, von L. C.: Type of K-fertilizers in the K-replacement strategy. Proceedings of the Int. Potash Inst. Congress, Reims/France (1986) Butler, E. G.: The mineral element content of spring pasture in relation to the occurence of grass tetany and hypomagnesaemia. J. Agri. Sci. 60, 329 (1963) Farley, R.F. and Draycott, A. P.: Growth and yield of sugar beet in relation to potassium and sodium supply. J. Sci. Fd. Agri. 26, 385-392 (1974) Gros, A.: Engrais ( 394 Environmental Aspects of K-Fertilizers in Production, Handling and Application H.J. Scharf* Summary The salts of saline deposits are not poisonous. Still, environmental problems arise in the area of potash mines through residue salt and its removal. These residues arise in solid form and partly in solution, too. The removal of the solid residues is carried out by backfilling and deposition in heaps. When disposal is in heaps, measures have to be taken to protect the underground water. The occurrence of saline waste water should be limited to unavoidable cases. Such cases are the production of sulphates of potash and sodium, the presence of carnallite in raw salt and saline waste water from piles. The disposal of saline waste water is successful in some cases through injection into the deep subsoil. When discharged into rivers, great care has to be taken to adjust to the changing volume of water in the river in a controlled fashion and in well-measured doses, at the same time paying due attention to keeping within prescribed limits. According to the standards of present technology, there are proven procedures to get rid of dust and harmful gases and so to purify the air. In handling potash fertilizer, the fight against dust is the main problem. However, the problem can be minimized through primary measures in production (crystallisation, granu- lation) and secondary anti-dust treatment. If fertilizer recommendations are heeded, there need be no fear of disadvantageous in- fluence on the subsoil water due to the application of potash fertilizer. The necessarily high costs for the erection and maintenance of environment protection installations have to be financed by the potash industry itself. In future, these costs must be taken into the calculation of prices by all producers and world-wide. In this way, the solidarity of all potash producers is demanded in the interest of our environment. 1. Introduction World-wide, there is no disagreement about the fact that environmental pro- tection is an international task and just as important as keeping peace or solv- ing the problem of hunger. There is debate, however, on the questions of the fastest way, the best methods and, above all, of adequate financing of the measures necessary. The potash industry, too, has to ask itself what it has contributed to the protection of the environment and in which areas there is still much to be done. This is a task which, first and foremost, must be tackled out of moral responsibility for our environment - for our planet. * Dr. H.J.Scharf, Kali und Salz AG, Postfach 10 2029, D-3500 Kassel, Federal Republic of Germany 395 But in the interests of our own business, too, we must be concerned in ensur- ing that our product ( 2. Environmental aspects in production 2.1 Disposal of residues We know, at least since Paracelsus, that it solely depends on the quantity to decide when a substance is useful or harmful. This is true for eating and drink- ing, for fertilizing and especially for the environment. Danger to the environ- ment through salts is to be suspected where these are extracted and where they are transferred in great quantities. Depending on the kind of deposit, we can say that from 100 t crude salt - only between 15 and 40 tons of potash fertilizer can be produced - but between 60 and 85 tons of residue have to be disposed of. Since roughly 250 million tons of crude salt are extracted yearly from potash mines worldwide, the enormous amount of about 200 million tons of residual 396 salts has to be disposed of without damage to the environment. Without doubt, this is the main problem for the potash industry. In as far as residue salts occur in solid form, the following solutions present themselves: a) The most attractive solution of the problem would naturally be the sale of the residue salts as products. That is succesfully done only in few, region- ally limited, cases because the cost of transport to main customers (as, for example, production of chlorine by NaCl-clectrolysis and of soda-ash [Na2 C0 31) is generally much too high. There are also often quality reasons against many possible uses, because accompanying minerals of calcium, magnesium and clay occur more espe- cially in residue. Purification through dissolution and recrystallisation is uneconomic with the salt industry which already produces sufficient. b) Thus only the disposal of salt residues remains. The obvious method is backfilling into the cavities of the mines. On closer inspection, however, problems show up. The return transport of residue salts to faraway sections of the mines and the influence of residual moisture on the climate in the hotter pits increase running costs considerably. For this reason, backfilling cannot be practised by all mines. Beyond that, not all residues can be disposed of by backfilling, hence, roughly 30-50% of residues must be piled up on the surface. The reason for this is that grinding of raw salt almost doubles its volume. Also some stretches in the mines must always be kept free for transport and ventila- tion. c) Thus residues have, no matter what we do, to be piled up. The environmen- tal problem of heaps arises as a result of rain and the occurrence of salty waste water thus caused. This cannot be allowed to seep underground, be- cause drinking water and the vegetation might possibly be damaged. For this reason, the present standard of technology demands that the heaps be sealed off, for example with clay or plastic film, unless nature herself has already provided an impermeable subsoil. The salty rainwater running off at the foot of the heap, however, must also be disposed of. Moreover, we must aim at having as big a heap as possible in order to keep the area, and so the effect of rain, as limited as possible. It is irresponsible, in terms of the environment, to dissolve solid residues in water only for the purpose of cheap disposal and then to let it into the nearest river, as is done in some countries. d) In the production of potash fertilizers, however, there are cases in which residues occur unavoidably not in solid form, but as salt waste water. These are: - Carnallite (KCI - MgCl2 • 6 H20) in crude salt - Production of sulphate of potash through the reaction of potassium chloride with other sulphates - and the salty rainwater of piles already mentioned. 397 Canalisation into rivers has to be regulated and controlled. That means that by adjusting the waste water to the changing quantity of water in the river, the limit allowed is not violated. Large water-tight dump basins for the waste water are necessary with this system. In some cases, the geological situation makes it possible to inject salt waste water into the deep subsoil, so-called ((deep well disposab: Here it must be ensured that the injection is only carried out in those formations * which are sealed off from the useful subsoil water above them. * Standard procedures in waste water technology such as purifying or filtering do not work with salt waste water. Theoretically speaking, drying by evaporation would be possible, but the energy costs for that are so high 'that it is cheaper to close the plant. 2.2 Purification of the air In several procedural steps during production, salt dust appears. This is not dangerous for human beings. That can be shown very convincingly by the fact that miners in potash mines do not, although they work in dusty sur- roundings, contract vocational diseases. But salt dust can damage the vegetation if it continually falls on trees. Effec- tive removal of dust, then, should be a matter of course for every potash mine, especially since technology offers suitable procedures. These are: 1. Stage: zyklon dust removal 2. Stage: electric filtering 3. Stage: textile-tissue filtering or wet scrubbing. With a combination of these 3 steps, the dust emission of drying installations can, as has been clearly proven, be kept under, for example, the presently offi- cial limits of 50 mg/m 3 in West Germany. Salt dust can be thrust out of the shafts with escaping air, too, and so affect the environment. One precaution against this is to install stretches with ex- tended cross-sections in the mine in front of the shaft. Here, the dust has the chance to settle at a lower air speed. In potash mines with magnesium compounds in the raw salt, a special problem arises: By the decomposition of magnesium chloride at the tempera- tures of the drying process, hydrochloric acid gas is generated in the neigh- 3 bourhood of some hundred milligrammes per m . This harmful substance must be removed from the escaping gases by ab- sorption. There are two procedures for this: a) Dry absorption by adding lime dust before a textile tissue filter or b) Wet scrubbing by washing the complete escape gas stream and neutralising the cleansing water. 398 By doing so, we can achieve purification to levels under 30 mg hydrochloric 3 acid per m . It must be added, however, that such an installation costs 5-10mio. DM in investment, depending on the size of the mine. The waste gases of power plants of the potash plants, too, can represent an environmental problem, especially when they are not heated by natural gas, but by sulphurous coal or oil. In this case, to remove sulphur dioxide (SO 2) perhaps gas purification has to be planned beforehand. The other harmful substances such as the oxides of nitrogen and carbon must be kept as low as possible by modern burners and control installations. 2.3 Noise protection In general, a potash mine has no difficulty in preventing noise emission. If, nevertheless, there is some annoyance to the neighbourhood, disturbances of the orderly running of machinery are usually the cause. These noises should and can be eliminated quickly by repair. 3. Environmental aspects in handling The handling of potash fertilizers begins in the potash plants; it occurs during transport by truck, rail or ship, again in the storehouse and in loading at the distributor and ends with the farmer as the last stage in its journey from the mine to agriculture. The only environmental problem of significance which turns up in the handling is dust. What, actually, is characteristic for dust? - The particles are so small and therefore so light that they escape by the movement of the potash in handling - or by the influence of wind - and float around in the air for a long time. - The grain size of dust is below 0.1 mm. - The dust problem cannot be completely solved by screening. Firstly, be- cause the fine dust persistantly sticks to the large particles, and especially because new dust is generated again and again by the friction of the large particles during movement. It has been shown repeatedly in technology that it is more economical to re- move the causes of mistakes than the consequences afterwards. According to this philosophy, we have to try to solve the problem as far as possible by primary measures: - To ensure that as little as possible dust is generated by controlled crystalliza- tion. - Where dust occurs unavoidably in production, it should be screened off before shipping the products. The removal of particles causing dust has 399 the additional advantage that, in so doing, caking of the potash fertilizers is reduced. For it is precisely the dusty particles which are the main cause of caking. - The dust removed must either be recycled in the production process or granulated. - In order to bind the dust which arises in further handling anew through rubbing off, an anti-dust treatment is necessary of both standard and granular products. Here it is a question of adding thick, sticky and oily fluids. The dispens- ing and especially the method of distributing the dust-binding agent equally over all the particles of the products depend on the know-how of the various potash producers. It is important, however, that no substances which are dangerous for the environment and plants are mingled with the potash fertilizer for the pur- pose of anti-dust treatment. Up to this point, handling is completely the responsibility of the producer. On leaving the mine, our product is then in other hands. It is in our interests to increase the acceptance of potash fertilizer on its way to agriculture through good technical advice. Experts of the potash industry should be will- ing to help in problems which come up. Today, I can only point out some aspects in captions: - The free fall of potash fertilizer from high levels during the filling of store- houses, ships and other vehicles is to be avoided. It is advisable to use a funnel which reaches a little bit above the heap of goods and which can be raised according to the filling level. - The air escaping through the filling process contains dust and should be sucked off and purified by filters. - Storehouses must have doors and windows which are well sealed-off. On the one hand in order to prevent dust from escaping into the environment, on the other hand this is, at the same time, a method of ensuring that moisture from outside does not enter into the storehouse. 4. Environmental aspects in application The same aspects which I have mentioned before remain valid in principle in the application of potash fertilizer by farmers. The chemical and physical nature of the K-cation is fortunately so favoura- ble that sufficient nourishment of plants can be secured, and that the con- tamination of subsoil water is avoided when applied properly. We have heard from the various papers on fertilizer recommendations what proper fertilization means. If they are heeded, environmental problems are surely avoided. This is true, too, for the K-accompanying anion Cl - or S42 -, which 400 is little held in the soil by adsorption. I have no informations on harm to human beings and animals via drinking water and edible plants by potash fertilizers. Sometimes, however, an environmental problem may arise which is appar- ent rather than real, because the scientific facts are not known or have not been understood. Such a problem came up as a consequence of the reactor accident in Cher- nobyl. Many states demanded a certificate which attested that the potash fer- tilizers were free from radioactivity before letting them enter the country. Now since the beginning of the earth's existence, a portion of the element potas- sium, namely the isotope 4 K, is in fact radioactive. This radioactivity, however, is of natural origin. This can be illustrated by the fact that, as a result of the K-content in his blood, each human being has a radioactivity of 4000 becquerel. You also go through 13000 becquerel when you swim in I m3 sea water - without anything happening to you. That proves more than words do that the natural radioactivity of potassium presents no risk whatsoever for health. Risks and dangers are to be found in artificial radioactivity as it is produced by nuclear fusion. This harmful radioactivity is not at all increased by potash fertilizer. One cannot point to these facts often enough so that the application of potash fertilizer will not one day be rejected out of ignorance. 5. Concluding comments I have tried to show you the most important environmental aspects in produc- tion, handling and application. Technological methods suitable for the solu- tion of the problems are at our disposal. But the necessary investments are not the only cost factors. After they are built, the environmental installations must be run and maintained. That means additional operating costs for environment protection. These investments and working costs must be financed by the potash producers and so earned in advance. In the case of the Kali und Salz AG, for example, more than a quarter of all investment is made for environment protection; in the meantime, the running costs of environment protection in- stallations have reached the level of 100 mio. DM per annum. As already experienced in western countries, the potash industry in other countries can no longer count on the financial support of the state. This means consequently, that the costs for environmental protection based on standard technology have to be a routine part of calculating the prices for potash fertilizer. Public awareness of the environment has grown to the extent that the consumer should also be willing to pay the extra price - so to speak as an insurance for a healthy environment. However, it will be neces- sary to increase the willingness to do so worldwide through in formation which people can understand. 401 Competition between producers to win over customers is necessary and a good thing because the value of the products is then secured and innovation is promoted. However, if competition has the effect that such low prices develop in the potash market that vital environment protection measures can no longer be carried out, competition becomes irresponsible. Such a situation demands the solidarity of all producers. 402 Coordinator's Report on the 5th Working Session Prof. Dr. K. Mengel, Institute of Plant Nutrition, Justus-Liebig-University, Sfldanlage 6, D-6300 Giessen, Federal Republic of Germany; member of the Scientific Board of the International Potash Institute The main paper of this session was presented by Hoisma. He gave an informa- tive and clear picture of how in the Netherlands scientific knoweldge is trans- ferred from the scientists via the extension service to the farmer. The system is a pyramid with the scientists at the top, the farmers at the base and the experts of the extension service in between. Knowledge is transferred in both directions, from the top to the base and reverse. There is an intense dialogue between the farmers and the experts of the extension service which brings about rapid implementation of new knowledge relevant for practical farming. The high standard of Netherland's agriculture proves that this system of chanelling knowledge to the farmer is very efficient. Its prerequisite is high motivation and educational level of farmers. Mrs. Vandendriessche presented a convincing example how farmers can be motivated to apply fertilizers and particularly potash by showing the effect of fertilizers on sugar beet quality. The main conclusion of her paper is that the farmer must be shown that the application of potash is profitable; this is a prerequisite for persuading the farmer to apply new knowledge. Gately pointed out that it is still possible to increase live-stock production from farm-produced forage, as is desirable in Ireland, by using new techniques for grazing and forage production, including higher fertilizer rates. He em- phasised that the removal of potassium in animal farm products is low so that rather low potassium fertilizer rates are required for maintaining a suffi- cient potassium status in the soil. This statement was not universally accepted by the audience and it was suggested that high leaching losses of potassium may occur on pastures. The final two contributions of the session were not completely focused on the main topic of the session, nethertheless they provided valuable data ac- cepted by the audience with much interest. Vis gave an impressive survey on the broad spectrum of potash fertilizers available on the Western European market which covers the diverse demands of agriculture, horticulture and forestry. Scharf showed that production, handling, and application of potassium fertilizers can be associated with environmental problems. He presented good examples of how such problems can be overcome. From the whole session one may draw the conclusion that there is already much knowledge regarding potassium fertilizer application but much of this 403 still needs to be implemented in practical farming. Financial support by the potash fertilizer producers is desirable in order to obtain further research data which may foster potash fertilizer application. 404 Closing Address Dr. N. Celio, former President of the Confederation of Switzerland, President of the International Potash Institute, Bern/Switzerland Ladies and gentlemen. The last two items on our agenda are the reports of the coordinators and the closing of the Colloquium by the President. Concerning the scientific mat- ter of the Colloquium, you, the Coordinators have just presented your con- siderations and reflexions. It would be outside the competence of a lawyer to speak about the plant root, nutrient balances and the adjustment of fer- tilizer recommendations and this is a reason why I am grateful to you for presenting the conclusions of this meeting. Needless to say, I am also most grateful to all who have presented papers and taken part in the discussions. It remains for me, in one or two words, to pass overall judgment on the conduct of this Colloquium. Everything was perfect, of a very high standard and we are all well satisfied. We have all to offer thanks to our Russian friends for the organisation of the scientific part of the Colloquium, for the technical and cultural excursions and last but not least for the rich food with which you have regaled us during our stay in Soligorsk. I wish to embrace all in our gratitude; everyone has been so kind and considerate, from the Director General to all his cooperators, including the hotel and restaurants, everywhere we have met the same cordiality. We use to say in Europe that «all are equal but some are more equal than others>. Nevertheless I would like to single out the following for special men- tion. First of all, Director General Podlesny who welcomed us here and did us the honour to be with us for the whole week. I am also grateful for the unstinted support of Byeloruskali; our thanks to Mr. Sychevsky for the excel- lent organisation in Soligorsk and for the very interesting excursions to the mines. Mr. Buksha of Leningrad is a friend who has supported our coopera- tion right from the outset. Mr. Prokoshev was responsible for the scientific part and for the Russian version of the preprints. At this point, I must express our thanks to all the translators for the highly professional work they have done. Mr. Bogdevitch, we thank you for the demonstration of how to dance the Russian polka, and even more for the organisation of the visit to the kolkhoz, and with him I thank the Chairman of the Lenin Kolkhoz, Mr. Wasilevitch. I must not forget to include in our thanks the coordinators and all the eminent scientists, professors and chiefs of research stations for their excellent presen- tations; the members of the Preparatory Committee and last but not least 405 my friend Director von Peter and his collaborators who have made such a large contribution to the success of the Colloquium. We are all delighted with our stay in Soligorsk and we leave with some nostalgia. Destiny has prescribed for each of us his place in human society and, beautiful as it may be here, we have to go back to our work. Our lives have been enriched by a new experience and we have made new friends. I wish you all a happy return home and you, my Russian friends, all the best for your families and for your country. 406