6 Heavy Metal And Radionuclide Contaminants In Phosphate Fertilizers



1) National Fertilizer and Environmental Research Center, Tennessee Valley Authority, Muscle Shoals, AL35660 , USA. (new address see list of participants)
Potash and Phosphate Institute of Canada, CN Tower, Saskatoon, Canada S7K 1J5


Phosphate rock (PR) contains various metals and radionuclides as minor constituents in the ores. Varying amounts of these elements are transferred to phosphate fertilizers in production processes, and later are applied to soils with these fertilizers. Cadmium (Cd) is the heavy metal of most interest because it is potentially harmful to human health, and much attention is being given to its avenues of entry into the human food chain. Among these avenues is the application of Cd to soil with fertilizers, and the subsequent uptake by vegetable and grain crops.

Concentrations of other heavy metal and radionuclide contaminants in P fertilizers vary considerably, depending on the PR source. Some metals of possible significance are: arsenic (As), chromium (Cr), lead (Pb), mercury (Hg), nickel (Ni), and vanadium (V). However, these metals are of less concern than Cd, either because they are not as readily absorbed by plants from P-fertilised soils or their apparent relative effects on human health are less than that of Cd. The main radionuclide contaminants in PR are uranium (U), radium (Ra), and thorium (Th).


Heavy metal concentrations in phosphate rock deposits

Most known PR deposits have been assayed for their heavy metal contents (Kongshaug et al., 1992) (Table 1). Concentrations vary considerably with metal (highest with Cr and lowest with Hg) and with region.

Converting the average heavy metal concentrations in PR to a basis relative to the P concentration allows an estimation of the average heavy metal application rate to soil at a given P rate (Table 1). Rates range from 0.01 to 25 g ha-1y-1 for Hg and Cr, respectively, when P fertilizers from these PR deposits are applied at a rate of 20 kg P ha-1. These rates would be somewhat lower if more refined P fertilizers, such as triple superphosphate (TSP) or diammonium phosphate (DAP), produced from the same PR sources were applied to soil.

Both Cu and Zn are micronutrients required for crop production. Average application rates for Cu- and Zn-deficient soils range from 1 to 5 kg ha-1, which are 100 to 1,000 times higher than the rates listed in Table 1. Some countries have set tolerable limits for heavy metal additions to soil such as the values set for German soils shown in Table 1 (Finck, 1992). Such limits generally are set for the surface plow layer (20-30 cm) of soil where most root activity exists.

The years required for a heavy metal contaminant in P fertilizers to reach the tolerable limit can be estimated as in the following example: total soil Cd generally ranges from 0.1 to 1 mg kg-1. Assuming an average of 0.5 mg kg-1, the total amount of Cd in the surface 20 cm is 1.5 kg ha-1. The tolerable limit for Cd is 2 mg kg-1 which is equivalent to 6 kg ha-1. Thus, Cd applications could total 4.5 kg ha-1 to reach the tolerable limit. From the data in Table 1, it would take 1,300 y of P applications at 20 kg P ha-1 to reach the tolerable Cd limit (4.5 kg ha-1 / 3.3 g ha-1). This calculation ignores other possible Cd inputs, such as sewage sludges and aerial deposition, as well as Cd removal by crops.


Table 1. Average heavy metal concentrations in phosphate rock (PR) deposits and estimated input to soil by P fertilizers (Kongshaug et al., 1992)


Heavy metal concentration
 PR Deposit  As  Cd  Cr  Cu  Pb  Hg  Ni  V  Zn

----------------------- mg kg-1 of PR ----------------------------
Russia (Kola) 1 0.1 13 30 3 0.01 2 100 19
USA 12 11 109 23 12 0.05 37 82 204
South Africa 6 0.2 1 130 35 0.06 35 3 6
Morocco 11 30 225 22 7 0.04 26 87 261
Other N. Africa 15 60 105 45 6 0.05 33 300 420
Middle East 6 9 129 43 4 0.05 29 122 315
Avg. of 91%

----------------------- mg kg-1 of PR ----------------------------
of PR reserves 11 25 188 32 10 0.05 29 88 239

---------------------------- mg kg-1 of P ----------------------------
  71 165 1226 209 66 0.29 189 578 1561
Applied with

----------------------------g ha-1 ------------------------------
20 kg P ha-1 1 3.3 25 4 1 0.01 4 12 31
Tolerable limit

-------------------------- mg kg-1 of soil --------------------------
(Finck, 1992) - 2 100 100 100 2 50 50 300



Transfer of heavy metals in phosphate rock to phosphate FERTILIZERs


Varying proportions of heavy metal contaminants in PR will be transferred into P fertilizers, depending on the manufacturing process. Ordinary superphosphate (SSP) is produced by reacting H2SO4 with PR. The resulting product will contain all of the heavy metal constituents found in the PR. A large percentage of today's P fertilizers is produced from wet-process H3PO4. By-product phosphogypsum also will contain a fraction of the heavy metals in the PR.

Most of the studies of such partitioning have been conducted with Cd. Wakefield (1980) reported that TSP contained 60-70% of the Cd in PR. The chemical form of Cd in TSP and DAP was reported to be Cd(H2PO4)2 or CdHPO4, or a mixture of these salts, which are the Cd analogs of the Ca compounds in these two P fertilizers (Mortvedt and Osborn, 1982). Williams and David (1973) found a close relationship between concentrations of P and Cd in superphosphates and their respective PR sources in Australia. While there is little data on the relative transfer of other heavy metals from PR to P fertilizers, it would be expected that the transfer coefficients would be similar for most of these metals, depending on chemical reactions occurring during the production .

Several chemical processes to remove Cd from H3PO4 before it is converted to P fertilizers have been studied. Examples are the extraction of wet-process H3PO4 with amines (Stenstrom and Aly, 1985) and by anion exchange (Tjioe et al., 1988) and cation exchange (Anon., 1989). While not yet tested on a commercial scale, solvent extraction appears to be a possible commercial method (Smani, 1993). Estimated costs are US-$230 Mg-1 of P or US-$46 Mg-1 of DAP, which would add about 30% to current P fertilizer prices. Benchekroun (1992) suggested that increasing current P fertilizer prices more than 20% to pay for environmental protection could result in dramatic decreases in -already low- P fertilizer use by developing countries.

Strontium (Sr) is another contaminant found in some PR deposits. Strontium apparently substitutes for Ca in the apatite structure of PR. Results reported by Kadar (1992) showed that the Sr content of peas (Pisum sativum L.) increased with rate of P application. The P source was SSP containing 12,000 mg Sr
kg-1, which was produced from Kola PR containing 22,000 mg Sr kg-1. Thus, much of the Sr in PR was transferred to the SSP, as expected. Little data is available on the Sr content of other PR sources and no reports were available on potentially adverse effects of Sr applied with P fertilizers.


Plant uptake of heavy metal contaminants in phosphate FERTILIZERS

Heavy metal contaminants in P fertilizers may be available to plants. Because of the potentially adverse effects of Cd on human health, most of the studies have been concerned with Cd. An early report (Shroeder and Balassa, 1963) showed that high application rates of TSP containing 35 mg Cd kg-1 P increased Cd concentrations in several vegetable species.

Williams and David (1973) reported that plant species differed considerably in their ability to take up Cd. Leafy vegetables absorb more Cd than grasses, and only 12-18% of the Cd in cereal plant tops was translocated into the grain. However, soil application of CdCl2 or TSP containing Cd resulted in increased Cd concentrations in both cereal grains and the edible portions of vegetables. Topdressing pastures with TSP also resulted in increased Cd of pasture species, especially that of subterranean clover (Trifolium subterraneum L.).

Mortvedt and Giordano (1977) reported greater Cd uptake by maize (Zea mais L.) from commercial DAP fertilizers which contained from 100 to 260 mg Cd
kg-1 P than from reagent grade DAP (5 mg Cd kg-1 P) (Table 2). Plant uptake of Cr, Ni, and Pb was quite variable and was not directly related to their concentrations in P fertilizers. Reuss et al. (1978) also found greater Cd uptake by radish (Raphanus sativus L.), lettuce (Latuca sativa L.), and peas from soil applications of TSP containing 870 mg Cd kg-1 P than from Ca(H2PO4)2, which is the main P compound in TSP. Uptake of Cd by all crops was much lower from a calcareous silt loam soil than from an acid sandy soil.


Table 2. Uptake of phosphorus and heavy metals by forage maize, as affected by source of P in fertilizer (Mortvedt and Giordano, 1977).

Plant uptake
Phosphorus fertilizer source






mg kg-1 of P mg pot-1

------------- µg pot-1------------------
Diammonium phosphate (18-20-0)*
Reagent grade 5 190





North Carolina 150 184





Idaho PR 250 167





Fluid fertilizer (10-15-0)
North Carolina PR 100 173





Idaho PR 260 193





Control 8





* Fertilizer grade on N-P basis.


While many of these studies have been conducted in greenhouse pot experiments, others have been done under field conditions. Mortvedt et al. (1981) applied DAP fertilizers artificially enriched to 10,370, and 765 mg Cd kg-1 P to a P-deficient soil. Concentrations of Cd in both grain and straw of winter wheat (Triticum aestivum L.) were higher only with the highest-Cd DAP applied to an acid soil (pH 5.1) but not to the same soil limed to pH 5.9. Jaakola (1977) also reported that Cd uptake by grain and straw of spring wheat was not affected by applications of P fertilizers containing between 10 and 405 mg Cd kg-1 P.

Numerous reports indicate significant differences among plant species in their ability to take up Cd and other heavy metals. Leafy vegetables are well known as Cd accumulators. Evidence for varietal effects on heavy metal uptake also is being gathered, because it may be possible to select varieties for low heavy metal accumulation in order to decrease the transfer of these metals into the human food chain.

Needs for P fertilizers are greatest on many acid, infertile soils in those countries where crop production is severely limited. Bioavailability of heavy metals, especially Cd, is greatest on acid soils. Therefore, Cd uptake may be increased in some crops fertilised by P fertilizers containing appreciable levels of Cd. Effects of increased Cd intake by those consuming such crops are unknown. However, results of USDA "market basket" surveys (Wolnik et al., 1983) have shown that Cd levels in several important USA food crops were lower than previously reported. The average weekly per capita Cd intake in the USA was estimated at about 100 µg, as compared with the maximum weekly Cd intake of 400-500 µg approved by the World Health Organization (Anon., 1989). Estimated per capita weekly Cd intake in Australia was 125-225 µg, based on a 1990 market basket survey (Anon., 1990). While Cd uptake by crops might be somewhat higher on P-fertilised acid soils, it seems doubtful that weekly Cd intake by humans will increase significantly to approach the maximum recommended levels listed above. Given the available information, concern about Cd intake is only warranted were several aggravating factors are combined. This situation could arise in regions of (1) acid soils with (2) low cation exchange and (3) low P fertility, to which (4) significant P fertilisation is applied, often (5) as low grade fertilizer or PR, particularly to (6) leafy vegetable crops, and where (7) the production is the main source of local food consumption. A combination of several or all of these factors is common in truck farming belts around population centres in the tropics. The application of contaminated sewage would further aggravate the situation.


Long-term effects of heavy metals applied in phosphate FERTILIZERS

Accumulations of Cd and other heavy metals applied to agricultural soils with P fertilizers are difficult to estimate because the mechanisms for addition and removal cannot be easily assessed. Avenues for addition, other than with P fertilizers, are atmospheric deposition, return of crop residues, and application of farmyard manure, sewage sludges, and other non-P fertilizers. After application to soils, Cd is very immobile so it tends to accumulate in the surface soil. Removal is mainly through grain, forage, and livestock products, with some removal possible from soil erosion. Levels of total Cd have been found to correlate with those of extractable P in pasture soils of South Australia, indicating that P fertilisation was the major route of Cd addition to these soils (Merry and Tiller, 1991). In contrast, the principal route of Pb addition was found to be from atmospheric additions (mainly automotive emissions) and not fertilizers. One method used to assess the net effects of heavy metal contaminants in P fertilizers applied to soils is to examine crops and soils from long-term soil fertility experiments. While heavy metal analysis may not have been made on many P fertilizers, such concentrations can be readily estimated if the source of PR used to produce these P fertilizers is known. Isermann (1982) analysed soils from 20 west European, long-term (26- to 138-y) experiments and calculated that Cd inputs from P fertilizers were of the same magnitude as the amounts recycled in the agricultural system through farmyard manure or crop residues applied to soils. Smilde and van Luit (1983) compared trends in soil Cd in P-treated and control plots in some long-term field experiments in The Netherlands. Estimated totals of 135 to 450 g Cd ha-1 applied over a period of 8 to 64 y in five experiments did not affect Cd concentrations in wheat and barley (Hordeum vulgare L.) grain, potato (Solanum tuberosum L.) tubers, sugar beet (Beta vulgaris L.) leaves, or onion (Allium cepa L.) bulbs.

Rothbaum et al. (1986) analysed soils from three long-term experiments at Rothamsted, England, and one in New Zealand. Annual application rates were equivalent to 33 kg P ha-1 and 5 g Cd ha-1 for 95 y in England and to 37 kg P
ha-1 and 20 g Cd ha-1 for 30 y in New Zealand. Very little Cd accumulated in the surface soil of two experiments, but about half of the applied Cd was found in the surface layer of a grassland soil in England and in the New Zealand soil. Mortvedt (1987) conducted a similar study on nine long-term (>50 y) experiments in the USA. Annual Cd applications were estimated to range from 0.3 to 1.2 g ha-1. Analyses of leaves or grain from maize, soybean (Glycine max L. Merr.) and wheat, and timothy (Phleum pratense L.) forage indicated that uptake of Cd contaminants in P fertilizers was negligible. Nickel concentrations in both maize and soybean grain were not affected by TSP applications for >50 y at the Morrow Plots of the University of Illinois (Table 3). Concentrations of Cd and Ni in TSP were estimated at 50 and 80 mg kg-1 P, respectively.

While data from long-term field experiments are needed to study the long-term effects of various practices on crop production and possible environmental consequences, such data may not give as definitive results as expected. Soil movement from plot to plot caused by tillage, soil fauna, wind, and/or water over a period of time may result in experimental errors. Such movement could negate differences between crop response from control and adjacent treated plots over time. Sibbesen (1986) reviewed research in which a soil movement model was fitted to soil P data from two 90-y field experiments in Denmark. This model helped explain why soil P concentrations in control plots did not decrease with time. Crop removal of soil P in these plots had been more than compensated by soil exchange between the control- and P-treated adjacent plots. Model simulations showed that only 28% of the surface soil layer presently in the center quarter of each plot was present at the initiation of these experiments. If soil movement in a given long-term experiment is severe, data from such experiments could be wrongly interpreted.

Size of individual plots as well as the size of the harvested portion also can affect interpretation of the data. While the magnitude of soil movement in the Morrow Plots is unknown, examination of the data in Table 3 shows a significant crop field response to P in the manured and P-treated plots. Clearly, design and management of long-term field experiments must be considered in interpretation of such data.

Vegetable and fruit crops may be more heavily fertilised than field crops, so Cd accumulations would be greater on these soils. Mulla et al. (1980) cropped soil from a 36-y citrus field experiment to barley (Hordeum vulgare L.) in a greenhouse pot experiment. They reported that Cd concentrations in grain and leaves grown on soil fertilised annually with 175 kg P ha-1 as a high-Cd TSP were similar to those grown on unfertilised soil. The average Cd concentration in TSP was estimated at 870 mg kg-1 P, so total Cd applied to soil was about 5.5 kg ha-1 (155 g ha-1 annually). When soil from both plots was cropped to Swiss chard (Beta vulgaris), a heavy metal accumulator crop, Cd concentrations were significantly higher from the P-treated soil.

In another study, 618 kg P ha-1 was applied annually for 9 y to several vegetable crops in New York (Mortvedt, 1984). The estimated annual Cd rate was 13 g ha-1, assuming 50 mg Cd kg-1 P in TSP. Concentrations of Cd were similar in fertilised and unfertilised seed of snap beans (Phaseolus vulgaris L.); blades, petioles and roots of beet (Beta vulgaris L.); heads and cores of cabbage (Brassica oleracea L.); and leaves and grain of sweet corn.

The accumulation of Cd in New Zealand agricultural systems has been recognised as a potential health problem as a result of unacceptably high Cd levels in some meat products (Bramley, 1990). The extensive use of P fertilizers on pastures has been related to increased Cd concentrations in kidneys of sheep and cattle. The maximum permissible Cd concentrations in kidneys for human consumption is 1 mg kg-1. Cadmium tends to accumulate in such tissue with time, so kidneys in all older animals may exceed this level.


Table 3. Concentrations of some heavy metals in maize and soybean grain after long-term applications of farmyard manure (FYM) and/or superphosphate (TSP) fertilizer, Morrow Plots, Illinois (Mortvedt, 1987).












  kg ha-1

mg kg-1
kg ha-1

mg kg-1
Control 3,700 0.020 2.5 2,300 0.062 5.8
FYM 6,200 0.011 2.7 3,200 0.032 3.9
TSP 10,200 0.028 1.4 3,400 0.035 3.4
FYM + TSP 10,500 0.016 3.3 3,500 0.032 2.9


Heavy metals in the human food chain

During the past 25 y, increasing attention has been given to the fate of heavy metals contained in municipal wastes and in commercial fertilizers applied to agricultural soils, with most of the attention paid to Cd. The medical community has been trying to determine relationships of various human maladies to heavy metal levels in the diet. Increasing interest has been given to various avenues of entry of Cd and other heavy metals into the human food chain, including the role of heavy metal contaminants in fertilizers (Table 4).

Availability of fertilizer Cd to plants is related to its solubility in the P fertilizer as well as to soil factors such as pH, cation exchange capacity, and clay and organic matter contents. Plants vary in both their capacities to absorb and to translocate Cd from vegetative tissues to grain.

Retention of absorbed Cd by humans or animals consuming plant tissues or grain varies considerably, but generally is considered to be low. Most of the retained Cd is found in kidney, liver and other organ tissues rather than in muscle tissue. Influences of retained Cd on human health are not well understood, although long-term exposures of a relatively high-Cd diet among individuals of poor physical condition would be likely to result in the most adverse effects. Paths in the human food chain should be somewhat similar for other heavy metal contaminants in P fertilizers applied to soil, although transfer coefficients could vary considerably with each metal.


Table 4. Probable route of fertilizer Cd in the human food chain and influencing factors and conditions (Gunnarsson, 1983).

Food chain component Influencing factors and conditions
Fertilizers Content of Cd and its solubility,
Content of other elements,
Influence of pH in soils.
 Soil  Natural content of Cd and its availability,
pH, CEC, clay content, organic matter,
Inputs from the air and via precipitation,
Leaching and losses by suspended particles,
Recirculation in plant residues and manure,
Accumulation in the soil
Plants Genetic variation,
Differences among plant parts,
Direct uptake from air and precipitation,
Content in harvested parts
Husbandry Animals  Retention, accumulation in certain tissues
Transfer to meat and milk
Food processing Separation of different parts and cooking
Breakdown of structure,
Changing of biochemical properties
Intake by man  Retention: 2-5% of ingestion
Other inputs: breathing, smoking, water, drink
Occupational exposure
Influences on health Total daily retention
Long term exposure
Diets: proteins, vitamins, other elements
Genetic variation among individuals
General physical condition

Using a model for the transfer of Cd through a New Zealand agricultural system, it was shown that careful grazing management on pastures fertilised with SSP made from Nauru and Christmas Island PR was required to avoid excessive Cd intake by grazing animals. Careful management includes avoiding excessive direct animal intake of fertilised soil or granular fertilizers topdressed on pastures. The model showed that excessive Cd levels in kidneys results with normal grazing management practices. The offal from animals grazing under these practices would accumulate Cd concentrations in excess of the maximum permissible levels for humans, but much of such offal is or could be marketed in pet food, thereby mitigating economic loss.

Not all of the heavy metal inputs to the human food chain are through P fertilizers and municipal waste applications to soil. The majority of environmental Cd inputs are anthropogenic in origin. Natural Cd levels in the atmosphere, in agricultural soils, and in surface or groundwaters generally are low. However, localised areas near certain metal smelters, mines, or other Cd-using industries could have considerably higher Cd levels in crops due to airborne depositions and water or soil pollution.


Regulations concerning heavy metal concentra-tions in phosphate FERTILIZERs

Because of the above concerns regarding heavy metal effects in the human food chain, various governments have introduced controls on heavy metal concentrations in municipal wastes (sewage sludges). During the early 1980's, a commission of the European Community (EC) recommended maximum total and annual loading rates of various metals to agricultural lands. For Cd, these were 2.4 and 8.4 kg ha-1 for recommended and mandatory cumulative maximum limits, respectively, and 0.1 and 0.15 kg Cd ha-1y-1 for recommended and mandatory maximum annual loadings, respectively (Webber et al., 1984). After The Netherlands introduced controls on Cd in sewage sludge used as fertilizer, the EC issued a directive specifying recommended and mandatory limits of 20 and 40 mg Cd kg-1 (dry weight basis), respectively, for sewage sludge applied to agricultural lands (Anon., 1989). Another EC directive in 1986 limited Cd concentrations in industrial effluents discharged into the Rhine River. A comprehensive report on sources, human exposure, and environmental impact of Cd in EC countries was prepared by Hutton (1982).

More recently, attention has been directed towards regulation of maximum Cd concentrations permitted in fertilizers (specifically P-containing fertilizers). It was estimated in 1987 that the annual Cd input to soil at average fertilizer application rates was 3.5 g ha-1 in EC countries, as compared with Cd inputs from sewage sludge of 60 to 167 g ha-1 in EC countries and 500 g ha-1 in the USA (Anon. 1989).

The Dutch Government proposed regulations in 1987 limiting the maximum Cd concentration in P fertilizers to 35 mg Cd kg-1 P (15 mg Cd kg-1 P2O5) by the mid-1990s (Table 5). The Danish Government announced proposals to gradually reduce the maximum Cd concentration in P fertilizers in four phases from 1987 to 1997. This proposal was later revised to slightly higher allowable Cd levels by 1998 as shown in Table 5 (Anon. 1989). Germany has a Cd limit on a voluntary basis. One reason for varying limitations of Cd in P fertilizers among countries (Table 5) is because agricultural practices vary significantly. For example, P application rates generally are much higher in Europe than in Australia. It should be noted that such regulations are based on Cd concentrations related to P concentrations in fertilizers, rather than a unit weight basis. Therefore, P application rates will dictate Cd inputs to soil, regardless of the P concentration in a fertilizer.


Table 5. Proposed or implemented Cd limits in P fertilizer in various countries.

Cd limit

Effective Year

mg kg-1 P
Australia 450 In effect
  350 1995
  300 2000
Austria 275 In effect
Denmark 200 In effect
  150 1995
  110 1998
Finland 100 In effect
Germany 200 Voluntary
Japan 343 In effect
Norway 100 In effect
  50 1995
Sweden 100 Uncertain
Switzerland 50 1996
The Netherlands 35 1995


Perhaps the main reason for regulating Cd limits in P fertilizers is that some surveys are showing increasing average Cd contents of surface soils and some plant tissues. In a comprehensive study using archived samples collected at the Rothamsted Experiment Station in England since about 1850, Johnston and Jones (1992) reported that the net annual Cd input by atmospheric deposition, which had averaged 3.2 g ha-1 over a 100-y period, increased to 14 g ha-1 during the past 20 y. In comparison, annual Cd inputs would be 6.9 g ha-1, with SSP applied at a rate of 20 kg P ha-1. Cadmium accumulated in acid soils containing about 5% organic matter but not in neutral soils low in organic matter. While Cd did not accumulate in subsoils low in organic matter, it was not confirmed if Cd losses by leaching had occurred. In addition to proposing controls on industrial Cd emissions, limits on Cd inputs from P fertilizers have been proposed in England.

In contrast, the amounts of Cd added to soils via aerial deposition in rural agricultural soils in the USA are on the order of few g ha-1, or about equal to those added to soil from normal P fertilisation programmes (Page et al., 1987). No limits on industrial Cd emissions or Cd concentrations in P fertilizers have been proposed in the USA at this time.

It seems probable that once limits on Cd in P fertilizers are in place in some countries, others will consider similar legislation. While there is less interest in placing limits on other heavy metals in P fertilizers at this time, they may be included even though there is much less data on which to base any regulations. Australia has a limit of 500 mg of Pb and 5 mg of Hg kg-1 of product for all fertilizers at present. More research is needed to determine the fate of heavy metal contaminants in phosphate fertilizers applied to soil. At this time, it appears that there may be some justification to consider limiting Cd concentrations in P fertilizers, especially in those countries where P application rates are rather high, or where soil conditions and production and consumption patterns are favourable to the transfer of Cd through the human food chain. However, there is very little evidence to justify limiting other heavy metals in P fertilizers.


Radionuclides in phosphate rock and transfer to phosphate FERTILIZERs

Phosphate rock varies considerably in content of U, Ra, and Th, depending on the geographical area from which it was mined. Typical levels of gamma radiation from the PR deposits listed in Table 1 ranged from 5 to 30 kBq kg-1 P (Kongshaug et al., 1992). The median contents of PR in the USA were reported as 59 mg kg-1 of U, 8 mg kg-1 of Th, and 18 mg kg-1 of Ra (Menzel, 1968). It was estimated that 3.7 x 107 Bq each of U and Ra and 1.4 x 106 Bq of Th had been applied per hectare in some potato fields in Maine over a 45-y period. Such additions of U and Ra were nearly equal to the total amounts naturally occurring in the plow layer (surface 15 cm) in soils (Talibudeen, 1964), thus potentially doubling the load. However, the rate of Th addition was much lower than the naturally occurring amount. Most crops are not as heavily fertilised with P as potatoes, although tobacco (Nicotiana tabacum L.) soils usually are highly fertilised with P.

Uranium in PR deposits throughout the world ranges from 3 to 400 mg kg-1, with those in the USA ranging from 50 to 200 mg kg-1 (Guimond, 1977). It has been estimated that Florida PR contains about 3.7 x 106 Bq (100 µCi) Mg-1 each of U and 226Ra and 1.5 x 105 Bq Mg-1 of 230Th (Menzel, 1968). Some of these elements are retained in the H3PO4 and the remainder are transferred to by-products during fertilizer manufacture. Guimond and Windham (1975) estimated that 60% of the radioactivity in mined PR in Florida remained with slime (phosphatic clay) and sand tailings during the beneficiation process.

In a comprehensive study of PR from four deposits in the USA, Wakefield (1980) reported that about 33% of the U in beneficiated PR concentrate, the feedstock for acidulation by the wet-process, was found in the phosphogypsum by-product. The remainder of the U was found mainly in the H3PO4, which subsequently is processed to several types of P fertilizers. Some companies remove the U from the H3PO4 and sell the "yellow cake" for purification as a nuclear power plant fuel.

Mustonen (1985) analysed 81 samples of NPK fertilizers in Finland for 238U, 235U, 226Ra, and 228Th. These samples represented 28 grades of fertilizers used for agriculture, forests, and gardens, with a mean N-P-K grade of 16.5/6.2/11.3. The mean activities of radionuclides per unit weight of P varied among fertilizer plants (Table 6). The average radionuclide concentration in these fertilizers was calculated from the relative quantities of fertilizers produced by each factory. The radionuclide rates were estimated at an annual P application rate of 30 kg ha-1. It was estimated that the annual contribution of 238U in P fertilizers was about 0.25% of the total U naturally occurring in the surface 10 cm layer. Because applied P eventually is mixed with greater depths of soil (probably a 25-cm layer), the annual U contribution would be less than the above percentage. More (1977) estimated that P fertilisation would cause an annual increase of about 0.04% of the total Ra concentration in tilled agricultural soils in Sweden. Rothbaum et al. (1979) reported that most of the U applied in SSP or TSP in long-term field experiments at Rothamsted, England and in New Zealand remained in the surface 23-cm layer of soil.


Table 6. Mean concentrations of radionuclides in compound fertilizers from various fertilizer factories in Finland (Mustonen, 1985).







---------------------------Bq kg-1 P --------------------------
I 2900 160 830 87 210
II 980 42 290 59 140
III 2400 130 770 95 220
IV 7400 420 1500 62 200
V 4100 230 2000 100 160
Weighted mean 3800 210 1100 78 190

Few papers have been published on uptake of the radionuclides in P fertilizers by agricultural crops. Kirchmann et al. (1978) analysed winter wheat grain and straw for 226Ra after plots were fertilised annually with P fertilizers for 11 y in Belgium. Total P application rates for this period were 150 and 600 kg ha-1, which would provide mean annual P rates of 14 and 56 kg ha-1, respectively. They reported no statistical differences in 226Ra concentrations of either grain or straw with both P application rates. A zero-P rate was not included in this study for comparison purposes, but non-fertilised soil and fertilised soil were analysed for 226Ra contaminants in P fertilizers added to soil at the low and high P rates represented only 0.24 and 0.96% respectively, of the total 226Ra in the upper 20 cm layer of soil. They concluded that the application of P fertilizers, even at the high rate, did not significantly affect the 226Ra content of wheat grown on this soil.

Plant tissues and soil samples were obtained from nine long-term (>50 y) soil fertility plots in the USA, and analysed for 226Ra, U, and Th. The TSP used for these studies, made from Florida PR, had been applied at a P rate of about 30 kg ha-1 annually. There were no differences in U, Ra, or Th concentrations in corn leaves or grain, wheat grain or straw, soybean leaves or grain, or timothy forage grown on non-fertilised or TSP-fertilised soil (Table 7). There also were no differences in concentrations of these elements in soil samples from the fertilised or non-fertilised plots, although concentrations in soil were much higher than those in grain, leaves, or forage tissue. These data suggest that annual applications of P fertilizers at recommended rates (15 to 30 kg ha-1) did not result in increased levels of radionuclides in grain and leaves of field crops. Considering these data, there seems to be little evidence for the entry of harmful levels of radionuclides into the human food chain from P fertilizers.

Table 7. Concentrations of radionuclides in plant tissues and soils from long-term soil fertility experiments in the USA (Mortvedt, 1991).








Type of sample control TSP control TSP control TSP
   --- mg kg-1 ---  --- Bg kg-1 ---  --- mg kg-1 ---
Morrow Plots, University of Illinois
Corn grain 0.01 0.01 0.7 0.4 0.01 0.01
Soil 1.40 1.30 22.6 31.1 2.90 2.80
Sanborn Plots, University of Missouri
Corn grain 0.01 0.01 0.4 0.4 0.00 0.01
Wheat grain 0.00 0.01 0.4 0.0 0.00 0.00
Timothy forage 0.04 0.06 1.5 2.6 0.02 0.00
Soil 1.70 1.50 22.6 20.7 4.60 4.20
Magruder Plots, Oklahoma State University
Wheat grain 0.01 0.00 0.4 0.4 0.01 0.02
Wheat straw 0.02 0.03 0.7 1.1 0.01 0.01
Soil 1.30 0.60 16.7 17.0 4.00 2.10
Auburn University, Alabama    
Corn grain 0.01 0.01 0.7 0.4 0.01 0.01
Corn leaves 0.01 0.01 1.1 1.1 0.01 0.01
Soybean grain 0.01 0.01 0.7 1.3 0.01 0.01
Soybean leaves 0.05 0.05 0.7 0.7 0.01 0.01
Soil 0.01 1.10 13.0 18.5 3.04 3.03

Phosphorus in the Global Environment.

Edited by H. Tiessen

© 1995 SCOPE. Published in 1995 by John Wiley & Sons Ltd.


Last updated: 12.07.2001