V. Farm-Scale Cycling and Flows

NUTRIENT CYCLING, TRANSFORMATIONS,AND FLOWS



39



concentrates &

other on-farmfeeds

N2



I



f



I



/



farm

boundarv’



to other fields



/

/



I

bedding



offifann manures and

other organic residues,

fertilizers, lime,

atmospheric deposirion



NZ



Figure 7 Managed system nutrient cycle and flows with a boundary around the farm.



field. An expansion of Fig. 7 to include all the farm’s fields would be needed to examine the details of flows within the farm. For example, nutrients may be “cycling” on the farm, but in a very uneven transfer pattern that depletes nutrients in

some fields and builds up nutrients to excessive levels in others. This type of nutrient redistribution can occur when, for the convenience of the farmer, manure is

applied only to fields near the barn.



A. WITHIN-FARM

NUTRIENT

FLOWS

Within-farm nutrient flows represent the allocation of the nutrient stocks available to the farmer. Although there can be within-farm flows on farms producing

only crops, with residues from one field removed and applied to another field, there



40



FRED MAGDOFF ET AL.



is a much greater ability to transfer nutrients on farms based on animal enterprises. On an integrated crop and animal farm these flows reflect the movement of nutrients from crop fields to animal facilities and application of manure to the crop

fields. These within-farm flows have not been routinely characterized. Bacon et al.

(1990) measured the flow of materials and calculated the nutrient balances for all

the fields on a Pennsylvania dairy farm. The balances were influenced by crop type

and management activities. There was little similarity in balances in P for adjacent

fields. There was also little relationship between distance from the dairy manure

storage and the nutrient balances. This farm may have been small enough and the

goal to use manure across the farm important enough that such a pattern did not

develop. Others (Bouldin et al., 1984;White and Safley, 1984) have suggested that

manure applications are usually greater close to the animal facility.



B. NUTRIENT

FLOWSTO AND



FROM



FARMS



Nutrient inputs to contemporary farms come in a variety of sources such as commercial fertilizers, lime and other inorganic soil amendments, organic amendments, biologically fixed N, and purchased feed. Manure produced on the farm is

not a source of nutrients for the farm, although it can be a significant source for individual fields. It represents a material to redistribute nutrients within the space of

the farm and to link the elements of the trophic pyramid. Nutrients contained in

the manure either came from farm-grown feeds or bedding or from purchased

feeds, bedding, and minerals. The particular farm inputs and outputs and the resulting nutrient loading depend on the strategic direction the farmer has taken. The

socioeconomic and political influences of society play an important role in the

farmer’s decisions regarding nutrient flows.

Fertilizer inputs are generally related to the goal of enhanced crop production.

However, biological N, fixation is most closely associated with farms that also

produce ruminant livestock. This occurs because the most productive legumes are

typically forages. Purchased feeds are common for most farms that include animal

production, but they are particularly significant for nonruminant animal production. Adams and McAllister (1975) measured the nutrient balance of a group of

farms in Northern Ireland. They observed that the P and K nutrient balances for

the farms with ruminant livestock tended to be lower than the balances for farms

with nonruminant animals, especially hogs. This reflects the differential emphasis

on on-farm feeds for the contrasting animal types.

For a farm producing ruminant animals, taking advantage of biologically fixed

N may no longer be an important consideration in the management organization

of the farm. Westphal ef al. (1989) calculated that when there was a limited crop

area used to grow corn relative to alfalfa, there could be a significant limit to dairy

herd size if biologically fixed N was fully accounted for and manure was used only



NUTRIENT CYCLING, TRANSFORMATIONS,AND FLOWS



41



to supply the N need of corn. With this limit in herd size came an estimated decrease in the potential net farm income. A cash crop farm in a southeastern Minnesota study had the lowest excess N balance per cropland hectare, whereas a dairy

farm had the highest (Legg et al., 1989). On the other hand, it is possible to use

animal manures on forage legumes without adverse environmental consequences.

In this situation, however, the N,-fixing capability of the legume is not fully utilized (Daliparthy et al., 1994).

Kaffka and Koepf (1989) measured nutrient balances on a mixed crop and livestock farm in southern Germany over a 30-year period. They calculated greater

losses of P and K in farm outputs than additions in farm purchases, but a high degree of self-sufficiency for N because of the production of legume crops (54% of

the fields in a legume plus undersown clover on an additional 23%). Nutrients contained in purchased feedstuffs and bedding for livestock were essential to the maintenance of soil fertility. Nolte and Werner (1994) measured N in the products of a

cropanimal organic farm in Germany as 183% of the inputs, without accounting

for biological N, fixation. Most of the nutrients left the farm in crop sales, with

less than 40% in animal products. The negative N balance of the farm contrasted

with the observations of Kaffka and Koepf (1 989). However, this farm had only

29% of the area in legumes as contrasted with the almost 75% on the livestockbased farm of Kaftka and Koepf (1989). Granstedt (1992) suggested that a farm

needs a minimum of one-third of the area in legumes to do without supplemental

N fertilizer. A minimum animal enterprise may also be necessary to utilize the forages produced and to contribute to maintenance of the soil nutrient stocks.

Feed inputs from off the farm generally increase with the animal density. Granstedt (1995) observed such an increase on three Swedish farms. Nutrient export

from these farms did not increase in the same proportion as the nutrient inputs with

feeds so that farm balances of each nutrient increased with animal density. There

was little relationship between N fertilizer purchased and N balance and there was

a negative relationship between P fertilizer inputs and P balance on these farms.

Granstedt ( 1995) concluded that potential nutrient losses from enlarged nutrient

stocks are strongly related to the intensity of animal production and the extent of

use of purchased feed, whereas fertilizer use is greatest in areas without livestock.

Westphal et al. (1989) found that purchasing feed could make dairy herd increases feasible compared to the herd size supported by on-farm crop production while

still balancing crop nutrient needs with the available farm stocks. However, if the

farm performance goal of balancing soil P was eliminated, more feed could be purchased, the herd size increased, and the net returns to the farm increase again. Purchased inputs can also affect soil organic matter stocks. Kaffka and Koepf (1989)

observed increases soil organic matter (and N) in the last period of a long-term

study after the purchased feed increased on a farm in Germany.

As animal production intensifies, the efficiency of nutrient use as measured by

the fraction of the inputs exported in the animal products often decreases. Van der



42



FRED MAGDOFF ETAL.



Werff et al. (1995) calculated nutrients in the produce from three organic mixed

crop-dairy farms in The Netherlands as 3 1, 83, and 27% for N, P, and K, respectively, of the inputs, whereas a conventional farm was projected to yield only 12,

29, and 15%, respectively. The major difference between the farms was the greater

stocking density on the conventional farm (2.4 vs 1.25 cows ha-') that was supported by feed and fertilizer purchases. Frink (1969) estimated that as the dairy

cow numbers (density) increased on a farm, the N balance would increase to the

point where nitrate losses from the field-applied manure would be significant.

The long-term effects of large annual manure applications can be significant.

Kingery et al. (1994) observed that total N increased to 30 cm, P and K increased

to depths of 60 cm, and nitrate N levels were greater to or near bedrock after 15-28

years of broiler litter application in Alabama. Concentrations of Cu and Zn, common additives to poultry feeds, increased to 45 cm. Long-term studies with cattle

and swine manure have measured increases in P and heavy metals (Chang et al.,

1991; King et al., 1990).

Reliance on biological-based nutrient sources does not necessarily lead to decreased nutrient loss to the environment. Nitrate leaching has been identified as a

potential problem on organic farms in Germany where Nolte and Werner (1995)

estimated that losses could be 25 kg ha-' year-' for a case study farm. Leaching

of nitrate appears to be within the current EC limit for drinking water on three organic mixed crop-dairy farms in The Netherlands, but will not meet the future recommendation of only 50% of the existing standard (Van der Werff et al., 1995).

Nguyen et al. (1995) estimated the N, P, and S budgets for three pairs of conventional or organichiodynamic farms in New Zealand. The marketed outputs from

the alternative farms were only 5 I % of the N inputs (largely biologically fixed N)

compared to approximately 90% for the conventional farms. Grain yields were

lower on the alternative farms and N concentrations in the grain on the alternative

farms tended to be lower. With more N inputs than the conventional farms, but limited crop performance, more N is probably lost through other pathways than crop

export.

Nutrient stock mining on farms that try to minimize the use of off-farm nutrient

sources may make them unsustainable over the long run. The alternative farms in

the New Zealand study of Nguyen et al. (1995) marketed considerably greater

fractions of the soil P stock compared to the conventional farms, even though the

outputs from the alternative farms were less. Lockeretz et al. (1980) found more

P and K removed by organic corn production than under conventional approaches on farms in the United States. Net Pand K losses have been measured for an organic farm in Germany (Nolte and Werner, 1994). This contrasted with the typical conventional farm in Germany, which gained nutrients at 5-10 times as much

per hectare as this organic farm lost. In some situations the export of nutrients may

equal the value of the product sold so that wealth of the soil is marketed through

soil mining and not the value added due to crop growth (de Wit er al., 1995).



NUTRIENT CYCLING, TRANSFORMATIONS,AND FLOWS



100



43



XEX



1 . 4 more ha than needed to produce all feed

2.0.8 more ha than needed to produce all feed

3. just the right area to produce all feed



80



4. all grain imported (corn and soybeans)

6. all grain plus halfthe forage imported

6. all grain plus 3/4 of the forage imported



20



0



-20



0



1



2



3



4



5



hectares cow1

Figure 8 Implications of various crop areas available per dairy cow on P accumulation/depletion

(see text).



An example of estimated net imports or exports of P on a dairy farm under various animal density scenarios is given in Fig. 8. Calculations were made with the

following assumptions: (i) The diet consists of corn silage, haylage, corn grain,

and soybean meal diet and is 0.4% P; (ii) total dry matter needs for a 636-kg (1400Ib) cow for an entire year and lactation (with production of 8172 kg milk) plus dry

period is 7530 kg; (iii) yields of corn silage, haylage, corn grain, and soybeans are

12.I , 6.7,6, and 2.5 tons ha-', respectively; (iv) P export in milk and meat is 9.1

kg cow- I year-'; (v) all land in excess of needs for the animal will be devoted to

haylage for sale; and (vi) no nutrients enter the farm except as animal feeds. When

there is much more land than required to produce feed for animals (a total of 5.2

ha cow I ) , forage crops are exported in addition to animal products and there is a

net loss of I8 kg P ha- I year- I from the farm if no fertilizer P is applied. If all the

feed is produced on the farm, there is a slight negative balance as exports exceed

imports as P minerals to supplement the diet. If all the grain and half the forage requirements are imported, there is a positive balance of approximately 35 kg P hayear I .

-



-



C. NUTRIENT

FLOWSBETWEEN FARMS

Modern farm specialization has separated the elements of the classic natural

ecosystem so that new linkages have developed among the farms. The emerging



44



FRED MAGDOFF ETAL.



pattern of nutrient flow in a specializing agricultural sector is that nutrients in fertilizer inputs to cash crop farms are transferred in the crops produced to animal operations (Lanyon, 1995). Thus, much of the nutrient input for animal enterprisebased farms is not as fertilizer. Approximately 8 5 6 5 , and 95% of the N, P, and K

inputs to a Pennsylvania dairy farm were in materials going directly for the animal enterprise as feeds, bedding, and minerals (Bacon er al., 1990). Nutrient flows

are increasingly among these specialized farms and the pattern of flow in many

cases is not a local land-based cycle at all. This pattern of flow should be recognized because it disrupts the spatial integrity of the trophic relationship in which

the waste products of one element in the relationship were the inputs for another.

A very high percentage of the U.S. cropland is used to produce grain for animal

consumption on other farms. Of the approximately 176 million hectares (435 million acres) of cropland harvested in the United States in 1992, corn grain, soybeans, and sorghum were grown on approximately 3 1 % (55 million hectares, or

137 million acres) (U.S. Department of Commerce, 1992). Although some of this

grain is certainly used on farm, typically more than 55% of the corn produced is

not used on the farms where it is grown (Watson, 1977) and approximately 25%

is exported abroad (NCGA, 1996).



D. PATTERNS

OF FARMNUTRIENT

FLOWS

After observing nutrient flow reports for a large number of natural areas Kelly

and Levin (1986) noted that the “requirement” for recycling nutrients diminishes

as the inputs increase. They expressed the requirement as the potential nutrient uptake divided by the inputs. Since the evolution of the fertilizer industry following

WW 11, nutrient inputs for production are not as scarce as they once were. The result is that the “biological necessity” for recycling has been eliminated. Now it is

possible for farms to be organized in different ways along the continuum of the requirement for recycling.

Each of the patterns of farm-scale nutrient flows have very different implications. The main patterns discussed in the following sections along with potential

implications at different scales are summarized in Table 111.

1. Farm Nutrient Exports > Imports

This is a mining process whereby nutrients contained in soil organic matter or

associated with minerals are being depleted. This pattern is more common on

farms exclusively producing crops than for mixed livestock-crop farms where animal products are a significant component of sales. If soil organic matter is plentiful (as it was in the virgin tallgrass prairie soil of the corn belt) or if the minerals

are geologically young and easily weathered (as is the case for K minerals in many

soils in the northern Great Plains), the mining of various nutrients may go on for



Table IIl

Potential Implications of Different Nutrient Flow Patterns

Geographic extent

Nutrient flow

pattern"



Plantkoil

and field



Farm



Implications if occurs on

individual fields or whole farm

Export > import



Decreasing fertility



Yields unsustainable



Export < import



Increasing fertility



High pollution potential;

system unsustainable



Export = import



Maintaining fertility



Goal, but may not be

most profitable under

politicaVeconomic

environment



"Assume starting with sufficient, but not excessive available stocks.



Regiodnation



Earth



Implications if occurs on

majority of farms

Decreased agricultural

production

Water pollution; depletion

of nonrenewable resources

(fuel, K and P deposits):

enhanced market

position

Many changes needed

in rural and urban areas

developed and developing

countries; animal production

limited



Limitation to

human poputaion

Depletion of

nonrenewable

resources

(fuel, K and P

deposits)

Moderated impact

of agriculture



46



FRED MAGDOFF ET AL.



decades. However, Crews et al. (1991) suggest that soil fertility should not be

mined in an ecologically based sustainable agriculture. To sustain soil productivity it will eventually be necessary to either export fewer nutrients or import more

nutrients. By introducing, or increasing, a livestock component, fewer nutrients

may be exported.



2. Farm Nutrient Exports < Imports

The accumulation of nutrients with this pattern is both wasteful of nutrients and

a potential environmental hazard as N and P accumulate. This is primarily a problem on livestock farms with relatively high numbers of animal units relative to

cropland and a reliance on purchased feed. To bring import and export more into

balance, fewer animals or more cropland may be needed. Another way to deal with

the problem at the farm level is to export manure. Some farms are already disposing of manure on neighboring farms or producing composted manure for sale. Although this may solve the problem of oversupply of nutrients for the individual

farm, when there are many farms with the same pattern of nutrient flow in a given region the opportunities for local export of manure may be limited and long distance export may be very costly (Young et af., 1985).Kloen and Vereijken (1995)

observed that the soil P and K reserves on a group of organic farms in The Netherlands were greater than the agronomic requirement. These farms would need to

pursue a management approach with greater nutrient outputs than inputs in order

to reduce the soil reserves. This could complicate the supply of N to the crops because only biologically fixed N would meet crop needs without additional P or K

as in the commonly used animal manures. The authors promote the concept of setting annual input/output balances in conjunction with the use of “acceptable”

sources. On crop farms where excess use of fertilizers has caused the buildup of

nutrient stocks, these stocks can be drawn down over a period of time by ceasing

imports as fertilizer and maintaining sales of crops.



3. Farm Nutrient Exports = Imports with Regard to One or More

Nutrients, but Not with Regard to Others

Some farms follow a philosophy of trying to depend on nutrient cycles with little or no importation of nutrients. Biological N, fixation is relied on to provide a

significant portion of the N inputs, but few other nutrients are used. In this situation, exports and imports of N may be balanced, but there may be a net export of

other nutrients.



4. Farm Nutrient Exports = Imports

As long as nutrient stocks are not too low, causing a yield sacrifice, or too high,

causing environmental problems, farms should strive to have this pattern as a long-



NUTRIENT CYCLING, TRANSFORMATIONS, AND FLOWS



47



term goal. Crop productivity is maintained and the threat of future environmental

degradation is minimized. However, animal production may need to be limited under this scenario to avoid buildup of nutrients. This pattern may not be economically sustainable for many specialized farmers who rely on animal production under the current political and economic climate.



VI. WATERSHED, REGIONAL, AND GLOBAL ISSUES

Although the specialization of agriculture and various incentives have broadened the scale of nutrient flow in agriculture, there are few studies that have extrapolated from small plot or whole farm sustainable production tactics to other

geographic units. There have been studies of different land uses in relation to nutrient flow, but these have been largely conventional practices on spatial scales

larger than small plots.



A. WATERSHEDS

The fate of inputs to a watershed is sometimes difficult to determine. A southeastern U.S. watershed appeared to retain 40% of the N, 58% of the P, and 63% of

the annual inputs to the watershed (Woods et al., 1983). Fertilizer accounted for

the majority of the inputs. Because of the mixed nature of the land uses (45% rowcrop agriculture, 13% pasture, and 30% forested), both soil storage and riparian

forests were suggested as significant sinks. However, the researchers note that storage capacity is certainly limited. In subsequent studies of this area, Lowrance et

al. ( 1986) found positive correlations between fertilizer inputs and harvested nutrients as well as the balance (unaccounted for nutrients). They concluded that the

field, farm, or landscape response was dominated by the removal of material in the

harvest but the watershed response was due to the interactions among the ecosystem components. Fluck et al. (1992) estimated that 73% of the P inputs to the

600,000 ha Lake Okeechobee watershed in Florida were from fertilizer and 20%

from dairy and beef feeds. Outputs from the watershed were approximately 18%

of the inputs and the yearly loading to the lake was only approximately 8% of the

inputs. The remainder of the inputs were retained in the watershed P stock.

Nutrient losses to the environment usually are greater from agriculture than

from forests or other undisturbed vegetation. Correll (1983) reported 5-10 times

greater losses of total N and P from a watershed dominated by conventional corn

production than a forested one. Losses from a pastured watershed were intermediate. However, the fraction of the N inputs lost in the drainage was comparable

at about 10% for the two agricultural land uses. The fraction of P lost was greater

for the cornfield (1 3%) than for the pasture (5%). Highest nitrate discharges oc-



48



FRED MAGDOFF ETAL.



curred in the winter for both agricultural land uses. Nitrate losses from southeastem U.S. agricultural watersheds tended to be greater than losses from forested

ones, whereas P loads were similar or slightly less than those coming from the agricultural watersheds (Lowrance et al., 1986). Kilmer et al. (1974) measured consistently higher N and P concentrations in discharge from a more heavily fertilized

watershed than one with lower fertilizer inputs. Hallberg et al. (1983) tracked increases in nitrate-N draining from an agricultural watershed in northeastern Iowa

as agricultural practices changed following the introduction of N fertilizer. However, Thomas and Crutchfield (1974) observed that the "background" nitrate levels from small Kentucky watersheds with crop, pasture, and forested land uses

were highly variable. A strong relationship existed between the geology of the watersheds and the P contents of the stream flow rather than the land use. Stream P

levels corresponded closely to those compiled 50 years earlier when few nutrient

inputs were available to agriculture.

The aerial extent of intensive agricultural operations has been related to nitrate

observed in groundwater (Beck et af., 1985; Pionke and Urban, 1985) and surface

water (Ritter, 1984). Patterns of elevated nitrate in a southwest Georgia study were

coincident with an area of intensive cropland (Beck et af., 1985). Nitrates in

groundwater in a Pennsylvania watershed were approximately four times greater

in wells under cropland than under forest (Pionke and Urban, 1985). The mixing

of recharge from the two areas diluted the nitrate concentration before the flow entered the surface water stream. Nitrate concentrations in drainage from watersheds

in Delaware with >60% cropland were more than 30% greater than those with

<60% cropland (Ritter, 1984). Klepper (1978) determined that both N fertilizer

use for corn and the fraction of watersheds in row crops (primarily corn and soybeans) were significant factors in explaining the variation in nitrate concentration

in surface waters from a central Illinois watershed.

The watershed studies suggest that making meaningful observations of nutrient

cycling in a sustainable agriculture at aggregated geographic scales is difficult.

Lowrance et af. (1986) suggest that watershed-level studies are essential in relating management to external environmental impacts. However, the relationships

may not be simple nor sufficiently sensitive for most decision making. Small plot

and field results do not seem to aggregate well to the higher levels of spatial resolution where the land use patterns are mixed and complex. Furthermore, there must

still be some connection between the activities and the responsibilities (for both

the farmer and the beneficiaries) before changes can be promoted. The appropriate level for observation and for responsibility remains a question. Background

concentrations and differential, as well as relatively slow nutrient transfer rates

may limit the usefulness of observations at the real or perceived scale of the problem. Perhaps an indicator at another level of resolution will be useful. The German guidelines for organic agriculture have selected the farm as the unit to limit

off-farm inputs of both fertilizers and feeds as a means to meet the expectations

they associate with organic agriculture (Nolte and Werner, 1994).