A. Granstedt Agriculture, Ecosystems and Environment 80 2000 169–185 171
Fig. 1. Inputs of fertilizer nitrogen, phosphorus and potassium kg ha
− 1
per year and outputs of nitrogen, phosphorus and potassium kg ha
− 1
in the form of animal- and plant-based food products a in 1995 in Swedish agriculture and b in 1993 in Finnish agriculture. Data for the period 1940–1995 and 1940–1993 for Sweden and Finland, respectively.
1800 and 1950, i.e., before the large-scale introduc- tion of artificial fertilizers and pesticides, agriculture in
Sweden underwent a technical and biological revolu- tion aimed at meeting the increased need for food by the
Swedish population which had grown from ca. 2 mil- lion to 7 million during this period Granstedt, 1998.
Animal production was relatively equally distributed among farms, and the number of animals on each farm
was determined by the farm’s fodder-production ca- pacity. In addition, the manure produced on a given
farm was returned to the soil on the same farm. In other words, animal and crop production was integrated
on each farm. The agricultural system was also self- sufficient with regard to fodder for horses. The export
and losses of nitrogen were compensated through nitro- gen fixation in leys with clover and other nitrogen-fixing
crops in a crop rotation tailored to the needs and condi- tions on each farm. In this highly developed agricultural
system the importation of external resources was very low, and by reconstructing plant nutrient balances from
1951 it was shown that nitrogen losses were about 50 lower than compared with today SCB, 1997. In prin-
ciple, the same description is applicable to Finland, the difference being that the agricultural revolution came
about 50 years later there Granstedt, 1999, and also to other countries in Western Europe. Kjaergaard 1994
has described in detail how agricultural production in Denmark was increased substantially by the introduc-
tion of the legume clover as early as the 18th century.
The nutrient flows on various types of farms in dif- ferent regions of Sweden have already been determined
for 1990 Granstedt, 1995. The aims of this paper are i to determine the nutrient flows in different regions
of Sweden in 1995, ii to compare the situation in 1995 with that in 1990, iii to explain the strong de-
pendence on artificial fertilizers and the high losses of plant nutrients in agriculture today, and iv to describe
how, with the technological resources available today plant nutrients could be handled more effectively at
both the production and consumption levels, thereby minimizing losses of nitrogen and phosphorus to the
environment.
Although the results and discussion presented here mainly relate to the situation in Sweden, the general
conclusions reached are generally applicable to the whole Baltic Sea drainage area and also to agriculture
in other parts of Europe with excessive surpluses of ni- trogen and phosphorus. The ultimate aim is to create
the knowledge base needed for developing a strategy for changing Baltic agriculture systems so that losses of
nitrogen and phosphorus can be reduced by 50 hope- fully, only the first stage of a successive reduction in
the countries around the Baltic Sea.
2. Methods
2.1. Introduction of the NPK-flow model used Several methods for quantifying and modeling plant
nutrient flows have been developed earlier Frissel, 1977; Jansson, 1983; Kaffka, 1984; Biermann, 1995;
Bleken and Backen, 1997. Here an NPK-flow model developed by the author is presented which is based
on case studies made on individual conventional and organic farms Granstedt, 1990, 1992a. These basic
studies were combined with field studies for 7 years 1981–1987 in which soil nitrogen mineralization,
172 A. Granstedt Agriculture, Ecosystems and Environment 80 2000 169–185
nutrient uptake in crops, harvested yield and recycling with crop residues, manure and urine on each field were
combined with data on inputs in the form of fertilizers, nitrogen fixation, atmospheric pollution, imported fod-
der and the export of food products. Mean values over all years were used to calculate a total balance for the
farm, in kg N ha
− 1
per year. Thereafter the model was simplified so that it predicted the flows of N, P, and
K without field studies Granstedt, 1995. This sim- plified model was based on statistical data from each
farm or statistical data allowing the model to be used for a group of farms at the municipality, landscape or
Fig. 2. Plant nutrient N, P and K flows, in kg ha
− 1
per year, calculated for a farm in Skaraborg County that specialized in cereal production in 1990.
country levels. Such studies of NPK flows in Sweden were made in both 1990 Granstedt, 1995 and 1995
this paper.
2.2. Plant nutrient pools in the agriculture–community ecosystem
Data on plant nutrients are presented here in kg ha
− 1
per year for arable land Fig. 2. Arable land is defined in this report as the area used for crop production within
a farm, county or country. To illustrate the circulation of plant nutrients, the ecological system including farms
A. Granstedt Agriculture, Ecosystems and Environment 80 2000 169–185 173
and the community can be visualized as being com- posed of different pools: one soil pool, one plant pool,
one domestic animal pool and one human pool. Plant nutrient flows between these different pools can be cal-
culated, as can the budget.
This study is based on available Swedish official statistics SCB, 1996a, b and, at the farm level, also
on information obtained from accounting records. Plant nutrient balances have been calculated at the farm,
regional and county levels. In these calculations flows related to the production and human consumption of
agricultural foodstuffs within the agriculture–commu- nity ecosystem have been analyzed. This ecosystem is
not closed, since there is a constant input of plant nu- trients, e.g., in purchased fertilizers, feed, atmospheric
deposition and biological nitrogen fixation. At the same time nutrients are exported from the system through
the sales of crop and animal products, losses to the at- mosphere and water, etc. Plant nutrient stores within
the different pools have not been estimated, nor has long-term variation been considered here. All calcu-
lations are based on statistics from two representa- tive years, 1990 study described earlier and on those
obtained in the 1995 study.
In this study, priority was given to studying animal production and crop production together as compo-
nents of the agricultural system as a whole. Manure from animals is in this context viewed as a transforma-
tion of nutrients within the system, and it is important to understand how each farm utilizes its own plant nu-
trient resources. On farms or within counties or coun- tries with insufficient home-produced fodder for the
animals there is an input of nutrients in purchased fod- der. In other studies in Europe Brouwer et al., 1995
and Sweden Hoffman, 1999 the animal production with purchase of fodder are excluded. In these stud-
ies estimates have been made of the amount of manure used in agriculture as one component in the calculation
of the nutrient input to agriculture.
2.3. Import of plant nutrients to the agro-ecosystem Nitrogen is imported to the agro-ecosystem through
bacterial nitrogen fixation by legumes, atmospheric deposition, the application of fertilizers and the pur-
chase of fodder. Phosphorus and potassium can be made available for plant uptake through the application of
fertilizers, the purchase of fodder or soil weathering. In estimating the contribution of nitrogen fixation it
was assumed that 25 of the ley area SCB, 1996a in all counties is in the form of a first-year ley with
legumes producing 100 kg fixed nitrogen per hectare. Fodder peas and beans were assumed to fix 50 kg ni-
trogen per hectare.
The figures for nitrogen deposition are based on measurements of wet and dry deposition made by the
Swedish Environmental Research Institute IVL for the period 1990–1994 SCB, 1996b, and the value used
represents the net effect after evaporation of ammonia to the atmosphere from crops and the soil surface.
For counties, the import of plant nutrients in the form of purchased fodder was calculated as the difference
between the total fodder requirements of the animals and the total amount of nutrients in fodder produced in
each respective area.
For counties, the import of plant nutrients in food- stuffs was calculated as the difference between the
production of the ecosystem and consumption by the population, based on average 1 year consumption and
the number of inhabitants. Calculations based on food consumption statistics SCB, 1996a show that the
human community met 60 of its nitrogen and phos- phorus requirements and 40 of its potassium require-
ments by consuming animal products, while the rest came from plant products.
To calculate the import of nutrients in form of purchased fodder and foodstuffs for the country as a
whole, the official trade balance for agricultural prod- ucts was used SCB, 1996a. For plant nutrients in an-
imal food products, it was calculated that there was a balance between imports and exports and, for cereal
products, that there was an export surplus for the year in question.
Estimates of the amounts of crop nutrients removed at harvest are based on the total calculated amounts of
plant nutrients in fodder and cash crops produced in each respective area. Values for the nutrient contents
of agricultural products are based on Swedish statis- tics Swedish National Food Administration, 1993 and
results of our own studies on nitrogen supply in con- ventional and ecological agriculture Granstedt, 1990.
Amounts of plant nutrients taken up by roots were as- sumed to constitute 25 of the total uptake by the
crop Hansson, 1987; Granstedt, 1992a. It was also assumed that 40 of the dry matter DM production
in ley crops was returned to the soil as residues.
174 A. Granstedt Agriculture, Ecosystems and Environment 80 2000 169–185
2.4. Transfer of plant nutrients within the system In this study annual mineralization was assumed to
correspond to the amount of organic nitrogen supplied every year through the decomposition of crop residues
and animal manure. For the country as a whole, it is difficult to determine whether there has been a net in-
crease or a net loss of organically bound nitrogen in the soil. Of course, there are also regional differences.
In the NPK-balance for the whole of Sweden we con- sidered statistics for 1995 with regard to amounts of
nutrients provided to the fodder industry in the form of slaughter waste and the spreading of wastes from com-
munity sewage treatment plants on agricultural fields Granstedt, 1992a.
2.5. Export Significant losses of organically bound phosphorus
occur when animal manure is spread under unsuit- able conditions. Losses of phosphorus and potassium
through leaching and runoff were assumed to be around 10 and 5, respectively, of the amounts supplied to
the soil in the form of animal manure. Phosphorus in manure is mostly in organic form, and only a small
part is soluble. It is assumed that about 30–40 of the nitrogen is lost through volatilization during storage
and that 10–20 is lost in the field, although this fig- ure will vary depending on the species of animals and
manure-handling techniques Lundin, 1988; Claesson and Steineck, 1991. Calculation factors used are given
in Figs. 3 and 4.
Soil nitrogen can disappear from the agro- ecosystem through denitrification, volatilization and
leaching. Soil denitrification is higher in clay soils than in sandy soils. However, the proportion of nitrogen lost
through leaching is larger in sandy soils than in clay soil Gustavsson, 1996. Recently it was discovered
that substantial losses of nitrogen can occur from wilt- ing organic material Whitehead et al., 1988. These
losses were included when calculating nitrogen losses from the soil. Losses of nitrogen were calculated as
the sum of total nitrogen supplied to the soil in crop residues and manure and mineral nitrogen supplied
in fertilizer minus the nitrogen taken up by the plants.
Losses of plant nutrients in the form of wastes cre- ated when processing food were calculated for ani-
mal products. The animal parts not used for human consumption wastes from the slaughterhouses were
considered as lost. One part is used by the fodder- processing industry; i.e. the nutrients are recirculated
in fodder. This recirculation is taken into account in the figures for Swedish counties Figs. 5 and 6 and
for Sweden as a whole Fig. 7. These wastes from slaughterhouses are, like the wastes from the human
population, a potential agricultural resource. N, P, and K contents of whole animal bodies are 2.6, 0.5 and
0.2 for pig and 2.5, 0.7 and 0.2 for cow, respec- tively Kirchmann and Witter, 1991; Fagerberg and
Salomon, 1992. Based on the relation between ani- mal production and actual consumption of the animal
products SCB, 1994, it was estimated that only 50 of the nitrogen, 57 of the phosphorus and 64 of
the potassium are present in the consumed parts of the animal body. The rest is waste.
Figures describing plant nutrient flows in the country as a whole are based on the calculated nutrient budgets
and nutrient flows in each county complemented with data on the export and import of agricultural products
in 1990 and 1995. Data used in a calculation program are given in Fig. 7.
In the figures one animal unit a.u. corresponds to one dairy cow, two young cows, three sows, 10 fatten-
ing pigs or 100 hens.
3. Results and discussion
