140 F.P. Vinther, E.S. Jensen Agriculture, Ecosystems and Environment 78 2000 139–147
Estimates of SNF are influenced by the method of measurement used. The soil is typically labelled with
15
N by adding labelled nitrate or ammonium to the standing crop grass-clover crop. Inaccuracies in
15
N isotope dilution estimates arise from poor matching of
reference and legume crop, coupled with the decline in
15
N enrichments of plant available soil N Witty, 1983; Witty and Ritz, 1984; Giller and Witty, 1987.
Stabilising the
15
N enrichment of the plant available N over time could potentially contribute to improved
estimates of fixation. The primary objective of this work was to esti-
mate the input of symbiotically fixed N
2
into an or- ganic cropping system via a grass-clover mixture and
to evaluate its importance for the N supply of crops in the rotation. The study included an evaluation of
15
N isotope dilution methods for estimating SNF in grass-clover mixtures.
2. Materials and methods
2.1. Soil and cropping system The experiment was carried out at the experimental
farm Foulumgaard at Research Centre Foulum in Den- mark 9
◦
35
′
E; 56
◦
30
′
N. The soil is a sandy loam with 7.5 clay 2
m
m, 9.5 silt 2–20
m
m, 45.5 fine sand 20–200
m
m, and 34.5 coarse sand 200
m
m, 1.6 total C, 0.14 total N, and pHCaCl
2
of 5.6. The cropping system was a 6 year crop rotation
with fodder beets Betae vulgaris L. cv. ‘Magnum’, spring barley Hordeum vulgare L. cv. mix. ‘Alexis’,
‘Ariel’, ‘Lenka’ and ‘Etna’ undersown with a mix- ture of perennial ryegrass Lolium perenne L. cv. mix.
‘Sisu’, ‘Merlinda’, ‘Borvi’ and ‘Tivoli’ and white clover Trifolium repens L. cv. ‘Milkanova’, first year
grass-clover, second year grass-clover, barley and pea Pisum sativum L. var Odin in mixture, and winter
wheat Triticum aestivum L. var. Obelisk. The 1- and 2-year grass-clover fields were grazed by young cattle
at a stocking rate of 10–12 animals ha
− 1
. Each year one cut for silage was removed before cattle-grazing.
The undersown grass-clover mixture was grazed dur- ing the autumn after harvest of the barley crop. Plots
for SNF measurements were placed in an area of the fields excluded from cattle grazing.
The crop rotation was established in 1986 and has been managed according to the principles for organic
farming, which implies that pesticides and industrially synthesised fertilizers have not been used over a 10
years period. Nitrogen and other nutrients are applied in farmyard manure or slurry and in deep litter. Two
levels of animal manure were used: 0.8 and 1.2 animal units a.u.. The only N
2
fixing crop that received ma- nure was grass-clover undersown to barley. Shortly be-
fore sowing, animal manure and deep litter was applied and incorporated into the soil in amounts correspond-
ing to approximately 150 and 225 kg total N ha
− 1
for the two fertilisation levels, respectively. The experi-
mental fields were irrigated at a water deficit corre- sponding to 60 of plant available water. In 1994,
1995, and 1996 barley with undersown grass-clover re- ceived 80, 70, and 40 mm, and the first and the second
year grass-clover 135, 97, and 96 mm, respectively.
2.2. Methods of
15
N labelling for determination of N
2
fixation Two methods for
15
N labelling of the soil were used for determining N
2
fixation in the grass-clover mixture from 1994 to 1996. In 1994 plots of 1 m
2
were estab- lished for
15
N labelling the standing crops of 1 and 2 year perennial ryegrass-clover mixtures. On 20 April,
the plots were supplied with 1 g N m
− 2
as ammonium sulphate [
15
NH
4 2
SO
4
] with 3 at.
15
N excess. The plots with undersown ryegrass-white clover mixture
were labelled by applying
15
N together with sucrose in order to immobilise the
15
N to obtain a more stable
15
N enrichment of the soil N pool over time, arising from remineralisation of immobilised
15
N. Also on 20 April, two days after sowing grass-white clover un-
dersown to barley, plots of 1 m
2
were established to which 1.5 g N m
− 2
was applied as
15
NH
4 2
SO
4
with 30 at.
15
N excess together with 15 g sucrose-C m
− 2
solution distributed evenly over the entire micro-plot area. In 1995 labelled plots were prepared in a sim-
ilar way, but ammonium sulphate with 6.2 at.
15
N excess solution was also applied to swards of one and
2-year ryegrass-white clover. In 1995 the ryegrass-white clover undersown in bar-
ley was labelled as in 1994 30 at.
15
NH
4 2
SO
4
+ sucrose. In 1996 the 1-year ryegrass-white clover
sward was labelled with 6.2 at.
15
N excess. All plots were replicated four times at each of
the two fertilisation levels. The above ground plant material was harvested several times during the
F.P. Vinther, E.S. Jensen Agriculture, Ecosystems and Environment 78 2000 139–147 141
growing seasons to a height of 2–3 cm above soil surface, separated into grass and clover, and dried at
80
◦
C overnight. 2.3. Nitrogen incorporated into the soil
Shortly before the second year grass-clover sward was ploughed under in the spring of 1995, 1996 and
1997, samples were taken for determination of dry matter and total N. Four turves from each of the two
animal manure treatments were taken. Each turf mea- sured 20 cm × 20 cm and included soil to approxi-
mately 20 cm depth.
Soil was washed from the turves, which were then separated into above-ground plant material and roots.
It was not possible to separate grass and clover roots. The dry matter production and total N content was
determined by elemental analysis after drying the fractions at 80
◦
C overnight. 2.4. Analysis and calculations
SNF was determined using the
15
N isotope dilu- tion method with ryegrass from the grass-clover mix-
ture as the reference plant. The content of total N and
15
N in the dried plant material was determined on a N-analyzer coupled online to an isotope ratio mass
spectrometer Carlo Erba-Finnigan as described by Jensen 1991. The
15
N natural abundance of the soil was 0.3663 at.
15
N and this value was used for cor- recting the
15
N abundance of plant samples for back- ground. The proportion of N derived from the atmo-
sphere was calculated using the following equation Fried and Middelboe, 1977:
pNdfa = 1 − atom
15
N excess
clover
atom
15
N excess
grass
Atom
15
N excess =
atom
15
N
clover or grass
− 0.3663
The amount of N
2
fixation was estimated as SNF = total clover N × pNdfa.
Analysis of variance was carried out using the GLM procedure in SAS SAS Institute, 1989. Least signif-
icant difference LSD was used to compare means, when differences were found to be significant at the
P = 0.05 level.
Table 1 Soil temperature and precipitation during the 3 years of the inves-
tigation Months
Soil temperature at Precipitation
10 cm
◦
C mm
1994 1995
1996 1994
1995 1996
January 2.3
1.3 0.5
82.1 72.0
2.7 February
0.9 3.2
0.1 21.0
77.6 13.8
March 2.5
3.4 0.3
74.6 62.8
0.7 April
7.2 6.6
5.7 20.1
35.9 5.3
May 11.3
10.8 9.5
10.2 49.8
51.7 June
13.4 14.8
14.8 58.5
45.0 14.8
July 18.8
17.4 15.4
10.2 18.4
27.6 August
17.3 18.6
17.0 133.0
14.1 48.2
September 13.5
14.1 12.2
171.8 83.4
58.7 October
9.2 11.3
9.8 43.9
28.6 62.6
November 7.1
5.6 5.7
55.0 50.5
99.7 December
4.6 1.9
1.9 73.9
5.0 34.7
Mean Total
9.0 9.1
7.7 754.3
543.1 420.5
3. Results
