Content uploaded by Patrick Galler
Author content
All content in this area was uploaded by Patrick Galler on May 11, 2015
Content may be subject to copyright.
ORIGINAL PAPER
Identification of Marchfeld asparagus using Sr isotope ratio
measurements by MC-ICP-MS
S. Swoboda & M. Brunner & S. F. Boulyga & P. Galler &
M. Horacek & T. Prohaska
Received: 13 July 2007 / Revised: 20 August 2007 / Accepted: 22 August 2007 / Published online: 14 September 2007
#
Springer-Verlag 2007
Abstract This work focuses on testing and application of
Sr isotope signatures for the fast and reliable authentication
and traceability of Asparagus officinalis originating from
Marchfeld, Austria, using multicollector inductively cou-
pled plasma mass spectrometry after optimised Rb/Sr
separation. The major sample pool comprises freeze-dried
and microwave-digested asparagus samples from Hungary
and Slovakia which are compared with Austrian asparagus
originating from the Marchfeld region, which is a protected
geographical indication. Additional samples from Peru, the
Netherlands and Germany were limited in number and
allowed therefore only restricted statistical evaluation.
Asparagus samples from Marchfeld were harvested within
two subsequent years in order to inves tigate the annual
variation. The results show that the Sr isotope ratio is
consistent within these 2 years of investig ation. Moreover,
the Sr isotope ratio of total Sr in soil was found to be
significantly higher than in an NH
4
NO
3
extract, reflecting
the mobile (bioavailable) phase. The isotope composition in
the latter extract corresponds well to the range found in the
asparagus samples in Marchfeld, even though the concen-
tration of Sr in asparagus shows no direct correlation to the
concentration of Sr in the mobile phase of the soil. The
major question was whether the ‘Marchfelder Spargel’ can
be distinguished from samples from the neighbouring
countries of Hung ary and Slovakia. According to our
findings, they can be clearly (100%) singled out from the
Hungarian samples and can be distinguished from the
Slovakian asparagus samples with a probability of more
than 80%.
Keywords Food authentication
.
Strontium isotope ratio
measurements
.
Multicollector inductively coupled plasma
mass spectrometry delimeter Asparagus
.
Soil
Introduction
Food traceability (‘from farm to fork‘) and food authentic-
ity have become a major concern in antifraud and consumer
protection durin g the past few years. Opening of interna-
tional markets and global competition affords companies
the possibility to offer high-quality food as well as cheap
products with low manufacturing guidelines labelled as top-
quality food. The European Union released mandatory
regulations to prevent customers from fraud. Regulation no.
178/2002 established the European Food Safety Authority
and laid down the general principles and requirements of
food law. Regulation no. 2081/92 has been developed as a
code for protection of geographical indications and desig-
nations of origin of agricultural products and foodstuffs [1].
Several methods and analytical techniques are currently
under investigation for proper determination of the geo-
graphical origin, such as sensory analysis [2], DNA-based
analysis [3, 4], spectroscopic techniques [5, 6], NMR [7]or
elemental and isotope fingerprinting methods using gas
source isotope ratio mass spectrometry (GS-IRMS) [8–12],
thermal ionisation mass spectrometry (TIMS) [13]orinduc-
tively coupled plasma mass spectrometry (ICP-MS) [14–16].
Anal Bioanal Chem (2008) 390:487–494
DOI 10.1007/s00216-007-1582-7
S. Swoboda
:
M. Brunner
:
S. F. Boulyga
:
P. Galler
:
T. Prohaska (*)
Department of Chemistry-VIRIS Project,
University of Natural Resources and Applied Life Sciences,
Muthgasse 18,
1190 Vienna, Austria
e-mail: thomas.prohaska@boku.ac.at
M. Horacek
Austrian Research Centers GmbH,
2444 Seibersdorf, Austria
Mainly the analysis of isotope ratios of the ‘bio elements’
like carbon, nitrogen, oxygen, hydrogen or sulphur using
GS-IRMS is currently applied to provide evidence of
authentic food. Seasonal variations such as humidity, dryness
or onset of winter, fertilisers and environmental or human
impacts influence the isotope signature of food. Furthermore,
C
3
and C
4
plants used for animal feed have an influence on
the carbon isotope ratios in meat or diary products. These
effects allow a discrimination of even small regions but can
also undergo seasonal variation. Further systems are current-
ly under investigation to provide additional information of
regional origin. Isotope ratios of Pb or Sr are applied for
determining the provenance because they reflect the geo-
genic and anthropogenic background and annual variances
are expected to be less pronounced [17–26]. Moreover, new
isotope systems are under investigation to improve the
analytical reliability. For example, Coetzee and Vanhaecke
[27] used boron isotopes as a fingerprint.
The aim of this study is to generate a reliable and rapid
method to determine
87
Sr/
86
Sr ratios in agriculturally
produced Asparagus officinalis in order to distinguish
asparagus from the Marchfeld from products originating
from neighbouring countries. The term ‘ Marchfelder
Spargel’ arose as a kind of trademark in Austria in 1980.
According to legislation 1263/96 from 2 July 1996, the
farmers who belong to the association of Marchfeld
asparagus farmers are allowed to name their asparagus
Marchfelder Spargel for marketing purposes. Food authen-
tication plays an important role in ident ifying the authentic
Marchfelder Spargel and in distinguishing it from asparagus
from other major producing and importing areas since a
significant amount of money is lost every year owing to
undeclared imports sold as Marchfelder Spargel.
Experimental
Reagents
Pro analysi (p.a) grade 65% HNO
3
(Merck, Darmstadt,
Germany) and deionised water (18 MΩ;REWAHQ5
Austria Wasseraufbereitung, Guntramsdorf, Austria) were
used through out this work. Water and HNO
3
were
additionally cleaned one and two times, respectively, by
subboiling in an ultrapure quartz apparatus (MLS DuoPur,
MLS, Leutkirch im Allgäu, Germany). Subboiled 37% HCl
(p.a., Merck), 70% HClO
4
(suprapure, Merck), 48% HF
(ultrapure, Merck) and 31% H
2
O
2
(p.a., Merck) were used
for digestion. NH
4
NO
3
(recapture , VWR International,
Vienna, Austria) was prepared for soil extraction in
subboiled water with a concentration of 1 mol L
−1
.
Polyethylene flasks and tubes were cleaned subsequently
with HNO
3
(10% w/w) and HNO
3
(1% w/w) and rinsed
with deionised water before use. Dilution steps involved in
the standard, sample and reagent preparation were per-
formed with HNO
3
(1% w/w), prepared from subboiled
deionised water and double-subboiled HNO
3
.
Quality control was performed by analysing a 5 ng g
−1
solution of SRM 987 SrCO
3
(NIST, Gaithersburg, MD,
USA). The certified value of SRM 9 87 is a
87
Sr/
86
Sr ratio
of 0.71034±0.00026. A generally ‘accepted value’ of the
87
Sr/
86
Sr ratio for this reference material is reported in the
literature as 0.710263±0.000016 (the error represents a stan-
dard deviation of 2σ from the external reproducibility) [28].
Asparagus samples and sample preparation
In total 155 white asparagus samples were collected over a
period of 2 years (2005–2006) (Fig. 1, Table 1). Seventy-
five samples were harvested from the fields of 12 farmers in
Marchfeld (Austria) by labor atory staff, 31 samples were
taken from four regions in Slovakia close to the Austrian
border and six samples were taken from Germany (highway
L419, near Mainz). Twenty-five white asparagus samples
came from three different regions in Hungary. Furthermore,
14 samples we re c ollected from four regions in the
Netherlands. Six white asparagus samples originating from
Peru were bought in a local supermarket in Vienna.
The fresh asparagus samples were cut with a ceramic
knife into small slices inside a class 100,000 clean room
and freeze-dried for approximately 24 h (Alpha 1–2 LD,
Christ GefriertrocknungsGmbH, Osterode am Harz, Germany).
A 0.2-g amount of each freeze-dried sample was directly
weighed into Teflon bombs for further microwave-assisted
digestion (MLS 1200mega, MLS, Leutkirch im Allgäu,
Germany). Concentrated double-subboiled HNO
3
(3 mL)
and H
2
O
2
(0.5 mL) were used as digestion reagents. The
time and temperature programme used was as follows:
2
1
3
4
5
6
NL
A
HU
D
SK
Fig. 1 Geographical origin of the European asparagus samples: 1
Marchfeld, Austria; 2 Bratislava, Nové Zamky, Velke Levare and
Kuty, Slovakia; 3 Kiskörös and Kecel, Hungary; 4 Mainz, Germany; 5
Groesbeek and Venlo, The Netherlands; 6 Escharen and Langenboom,
The Netherlands
488 Anal Bioanal Chem (2008) 390:487–494
5 min with 250 W, 5 min using 400 W, further digestion for
10 min with 600 W and finally 5 min with 250 W. After this
digestion procedure the microwave oven was vented for
10 min. The samples were finally transferred into 50-mL
flasks, filled with HNO
3
(1% w/w) to 25 g and stored at 4 °C
for further investigation.
Soil samples and sample preparation
Approximately 100 g soil samples were collected randomly
from 13 different asparagus fields in Marchfeld at a soil
depth of 20 cm. The samples consisted of sandy soils (Ise,
Wei, Maz and Hab) and loamy soils (Muh, Mag and Ung)
(Table 1). Residual plant material was removed by hand.
The soil samples were air-dried and 1.5 g of each sample
was finely ground with a ball mill (MM 2000, Retsch,
Haan, Germany).
Several studies have already been established to decom-
pose specific compartments of soil or sediment via
microwave digestion. HF is used for digestion of mineral
aluminosilicate matrices and the oxidising HClO
4
decom-
poses the organic matrices. [29, 30] Thus, 0.2–0.3 g of
the milled soil was weighed into Teflon bombs. HF (4 mL)
and HClO
4
(1 mL) were added and digested via micro-
wave-assisted digestion (MLS 1200mega) using the same
temperature program as described before. The samples were
finally transferred into perfluoroalkoxy (PFA) tubes and
evaporated to approximately 0.5 mL at 180 °C. Subsequently,
2.5 mL aqua regia (3:1, HCl/HNO
3
) was added and the
samples were evaporated to 0.5 mL at 100 °C. Finally, the
solution was made up to 10 g with 6 mol l
−1
HNO
3
.
The 13 soil samples were passed through a 2-mm sieve
prior to NH
4
NO
3
extraction. The extraction followed DIN
V 19730 [31], which descri bes an extraction method of
soils with 1 mol l
−1
NH
4
NO
3
solution for the determination
of the mobile metal fraction in soils. A 50-mL aliquot of the
NH
4
NO3 solution was added to 20 g of sieved soil. The
samples were shaken with an overhead shaker for 2 h at
20 rpm at room temperature and subsequently the solutions
were filtered with filter paper (Whatman no. 42) into 50-mL
polyethylene vials. A 0.5-mL aliquot of HNO
3
solution
(1% w/w) was added to stabilise the samples for further
storage.
Strontium and rubidium separation
Liquid soil and asparagus samples were finally separated
according to an optimised batch process described below
using Eichrom Sr resin (Eichrom Industries, Darien, IL,
USA).
Resin (5–10 g) is filled into a 100-mL bottle and
approximately 50 mL HNO
3
(1% w/w) is added for
conditioning of the resin. The resin is soaked for at least
30 min but better is overnight. The supernatant is removed
owing to colloid formation at the surface and the bottle is
refilled with fresh HNO
3
(1% w/w). The resin is ready for
use and is stored dispersed in the acid in a refrigerator
(8 °C). Recycled resin is washed three times before reuse
with 50 mL HNO
3
(1% w/w) or subboiled water. The flask
is subsequently shaken for 1 to 2 min. The procedure is re-
peated twice. The soaked resin can be reused up to four times.
The method applied for Sr/Rb separation in our
laboratory so far [32] turned out not to be applicable for a
large number of samples since it is not very reproducible
with respect to effective separation and Sr recovery and
was therefore further optimised. Moreover, the method is
very time-consuming, as the glass wool has to be packed
manually; thus, the flow rate is not constant and resin
particles can be released into the sample solution , leading to
subsequent blockage of the nebuliser. A combination of 3-
mL frit tubes (Separtis, Grenzach-Wyhlen, Germany) and
appropriate filters (Separtis, Grenzach- Wyhlen, Germany,
Table 1 Asparagus samples
Country Location Sample
code
Number of
samples
Soil
type
Austria Raasdorf Maz 5 Sandy
Aderklaa Muh 6 Loamy
Franzensdorf Ung 1 4 Loamy
Ung 2 4 Loamy
Lassee Wei 1 5 Sandy
Wei 2 4 Sandy
Mannsdorf Mag 1 5 Loamy
Mag 2 3 Loamy
Mag 3 5 Loamy
Aderklaa Hab 1 5 Sandy
Hab 2 6 Sandy
Aderklaa Ise 1 5 Sandy
Marchegg Mak 5 Sandy
Gerasdorf Hora 1 2 -
Raasdorf Hora 2 3 -
Raasdorf Hora 3 3 -
Raasdorf Hora 4 3 -
Markgrafneusiedl Hora 5 2 -
Germany Mainz-Finthen Mai 6 -
The
Netherlands
Groesbeek Gb 2 -
Venlo Ven 2 -
Escharen Teu 7 -
Langenboom Dal 3 -
Slovakia Velke Levare Vel 9 -
Bratislava Bra 5 -
Kuty Kut 9 -
Nove Zamky Zam 8 -
Hungary Kecel Kec 7 -
Kiskörös Bac 8 -
Unknown Hun 9 -
Peru Unknown Peru 5 -
Anal Bioanal Chem (2008) 390:487–494 489
pore size of 10 μm) resulted in a constant flow rate of 1.8 mL
min
−1
for 10-μm filters. Filters with 20-μmporediameter
resulted in a flow rate of about 3.0 mL min
−1
. A recovery rate
of Sr below 1% led to the assumption that the latter flow rate
was too hig h. The volume of 3 mol L
−1
HNO
3
for
conditioning was increased from 1 to 2 mL to ensure that
the resin was sufficiently activated and the resin volume was
increased from 0.3 to 0.5 mL. The capacity of the resin is
approximately 10 mg mL
−1
, resulting in a theoretical binding
capacity of 5 mg in 0.5 mL resin. Eichrom recommends that
the resin should only be loaded with 10–20% of its maxi-
mum capacity [35]. The sample volume was kept at 2 mL,
but the concentration of the HNO
3
was increased from 3 to
6 mol L
−1
for the further washout of Rb. The wash volume
was increased from 1.5 to 4 mL for improved washout.
It was proven that practically no Sr remained in the resin
after the separation of asparagus samples. Moreover, the
results show that the applied washing steps are sufficient
for removing the Rb quantitatively. Additionally, no
memory effect was observed even when the resin was used
up to four times. The procedu ral Sr blanks (including
digestion, separation etc.) were lower than 0.4 ng g
−1
and
the Rb blank concentration was below the detection limit of
0.005 ng g
−1
.
This optimisation leads to the following separation
procedure. The separation starts with filling 0.4 mL of the
resin into frit tubes. After the resin has been washed with
2 mL subboiled water, it is conditioned four times with
0.5 mL HNO
3
(3 mol L
−1
). A 2–mL aliquot of the sample is
acidified to reach a concentration of 3 mol L
−1
HNO
3
solution and is pipetted into the tubes. The samples are
eluted with 2 mL subboiled H
2
O after washing eight times
with 0.5 mL HNO
3
(6 mol L
−1
).
The separated asparagus samples needed no further
dilution step for isotope ratio analysis of Sr since the
concentration of Sr after separation was between 5 and
40 ng g
−1
. The digested and extracted soil samples were
diluted 1:50 after separation to obtain concentrations in the
range from 5–30 ng g
−1
. Sr concentrations in analysed
samples below 5 ng g
−1
led to wor se precision and thus
increased uncertainty of isotope measurements therefore,
18 asparagus samples from those collected could not be
analysed.
Instrumentation
Screening of the digestion solutions for Rb and Sr was
performed using a quadrupole inductively coupled plasma
mass spectrometer (DRC II, PerkinElmer, ON, Canada)
operated in standard mode.
Isotope data were acquired using a double-focusing
high-resolution multicollector inductivel y coupled plasma
mass spectrometer (Nu Plasma, Nu Instruments, Wrexham,
UK) coupled to a membrane desolvating system (DSN-100,
Nu Instruments, Wrexham, UK). The DSN-100 instrument
was equipped with a PFA nebuliser (MicroFlow nebuliser,
Elemental Scientific, Omaha, NE, USA) and additionally
had a spray chamber with hot gas flow which eliminated
droplet and condensation formation. The multicollector
inductively coupled plasma mass spectrometer is provided
Table 2 Operating parameters
and scheme of the monitored
isotopes for the Sr isotope
measurements
Nu Plasma settings
Rf power 1,300 W
Auxiliary gas flow rate/cooling gas flow rate 0.75 mL min
−1
/ 13.0 mL min
−1
Sample uptake rate 100 μL min
−1
Sample/skimmer cone Ni
Nebuliser Perfluoroalkoxy nebuliser
Sampling mode 6 blocks of 10 measurements
Measurement time 10 min per sample
Mass analyser pressure < 10
−8
mbar
Background/baseline determination HNO
3
(1% w/w)
Washout time 3 min
Axial mass/mass separation 86.05/0.5
Detection system 12 Faraday collectors
Cups L5 L4 L3 L2 L1 Ax H1 H2 H3 H4 H5 H6
Isotope
82
Kr
83
Kr
84
Sr
85
Rb
86
Sr
87
Sr
88
Sr
DSN-100 nebuliser settings
Nebuliser pressure 2 bar (30 psi)
Hot gas flow 0.7–0.9 L min
−1
Membrane gas flow 4 L min
−1
Spray chamber temperature 112 °C
Membrane temperature 122 °C
490 Anal Bioanal Chem (2008) 390:487–494
with a collector configuration consisting of 12 Faraday cups
and three ion counters. All isotopes in this work (
82
Kr,
83
Kr,
84
Sr,
85
Rb,
86
Sr,
87
Sr,
88
Sr) were measured using the
Faraday cups. The operation parameters and the details of
the Faraday cups used for Sr isotope ratio measurement are
given in Table 2.
Experimental parameters of the multicollector inductively
coupled plasma mass spectrometer, including nebuliser gas,
rf power and ion transfer lens potentials, were optimised to
achieve the maximum ion intensity for
88
Sr, using the NIST
SRM 987 solution with a concentration of 5 ng g
−1
.
In order to test the reproducibility of the Sr isotope ratios
obtained, all measurements carried out during the 8-month
study were evaluated. The normality test indicated that the
measured Sr isotope ratios in NIST SRM 987 corresponded
well to a normal distribution (Fig. 2). The mean
87
Sr/
86
Sr
ratio of the 103 measurements of SRM 987 was 0.71039±
0.00013. It is evident that the value is higher than the
generally ‘accepted value’ of SRM 987 and is slightly
higher that the certified isotope ratio. Slightly increased
87
Sr/
86
Sr isotope ratios measured by multicollector ICP-MS
(MC-ICP-MS) in comparison, for instance, with TIMS
seem to be a general observation and they underline the
necessity for a more detailed study of mass bias correction
procedures in MC-ICP-MS.
Analytical protocol
Blank correction was performed by the ‘measure zero’
method, provided by the Nu Plasma software. The magnet
was set on the axial mass (mass 86) and the background
signal (HNO
3
, 1% w/w, for SRM 987 and an average
digestion blank for the asparagus samples) of all masses (88,
87, 86, 85, 84 and 83) was measured for 100 s. An average
value was subtracted from the measurements of the following
samples. Mass bias for both
87
Rb/
85
Rb and
87
Sr/
86
Sr isotope
ratios was corrected for by using the exponential mass
fractionation law [33] via the
88
Sr/
86
Sr isotope ratio.
87
Rb correction was performed via
85
Rb even after Rb/Sr
separation as a routine analysis procedure. Several mixtures
of SRM 987 (1, 2, 5, 10 and 20 ng g
−1
) spiked with Rb (0,
1, 5 and 10% of the Sr concentration) were prepared to
check the math ematical Rb correction. The results showed
clearly that Rb correction in samples with less than 10% Rb
concentration lead to proper results. Table 3 shows
deviations of
87
Sr/
86
Sr isotope ratios measured in SRM
987 solutions containing 20 ng g
−1
Sr spiked with different
amounts of Rb together with the uncertainty contribution of
the natural abundances of Rb isotopes solely. An uncer-
tainty of 0.003% was obtained in unspiked SRM 987
solution. When the samples with Rb are spiked with 1–10%
Sr concentration the uncertainty increases by up to 0.03%
owing to the imprecisely known isotope composition of Rb
used for correction of interference by
87
Rb. In practice, the
deviation of the corrected
87
Sr/
86
Sr isotope ratio from the
certified value exceeds the contribution from the uncertain-
ty of the Rb isotope compo sition (type B uncertainty) by
about twofold, which cannot only be explained by the
uncertainty of the Rb isotope composition. An additional
error (type A uncertainty) might be introduced by different
mass bias factors for Sr and Rb isotope ratios.
Results and discussion
Asparagus from Marchfeld (range/annual variation/
strontium in soil )
The Sr concent ration in the investigated asparagus samples
ranges from 24 to 810 μgg
−1
(dry weight). Screening prior
to separation indicated a Rb range of 24–990 μgg
−1
(dry
weight) in asparagus samples from all origins.
Normal Distribution SRM 987
0
5
10
15
20
25
30
35
87
Sr/
86
Sr ratio
Frequency (absolute %)
0.7101 0.7102 0.7103 0.7104 0.7105 0.7106
Accepted value ± uncertainty
Certified value ± uncertainty
Measured value ± 1
(in this study)
Normal Distribution SRM 987
0
5
10
15
20
25
30
35
87
Sr/
86
Sr ratio
Frequency (absolute %)
0.7101 0.7102 0.7103 0.7104 0.7105 0.7106
Accepted value ± uncertainty
Certified value ± uncertainty
Measured value ± 1
(in this study)
Normal Distribution SRM 987
0
5
10
15
20
25
30
35
87
Sr/
86
Sr ratio
Frequency (absolute %)
0.7101 0.7102 0.7103 0.7104 0.7105 0.7106
Normal Distribution SRM 987
0
5
10
15
20
25
30
35
87
Sr/
86
Sr ratio
Frequency (absolute %)
0.7101 0.7102 0.7103 0.7104 0.7105 0.7106
Accepted value ± uncertainty
Certified value ± uncertainty
Measured value ± 1
(in this study)
Accepted value ± uncertainty
Certified value ± uncertainty
Measured value ± 1σ
(in this study)
Fig. 2 Distribution of the measured SRM 987
87
Sr/
86
Sr isotope ratios
(n=103, D=0.0647, P=0.7824) including the mean value of the
measured SRM 987 Sr isotope ratios ± 1σ (0.71039±0.00013, light
grey arrow), the certified range of SRM 987 (0.71034±0.00026, grey
arrow) and the accepted value reported in [28] (0.71026±0.00002,
black arrow)
Table 3 Uncertainty contribution from the natural abundances of Rb isotopes (Sr concentration 20 ng g
−1
)
0% Rb 1% Rb 5% Rb 10% Rb
Contribution from the uncertainty of the Rb isotope composition - 0.0053 0.0166 0.0267
Observed deviation from the certified values - 0.016 0.066 0.062
Anal Bioanal Chem (2008) 390:487–494 491
The Sr isotope composition of Marchfeld asparagus from
13 different farmers was used to identify a range of Sr
isotope ratios which corresponds to the Marchfeld aspara-
gus. The resul t is given in Table 4. The range of Sr isotope
ratios determined as the median ± 2 times the standard
deviation can be defined as a probability of 95% that a
sample can be identified as an asparagus sample from
Marchfeld within this range.
One advantage of the application of Sr isotope ratios is
the expectation that the ratio does not undergo seasonal or
annual variations. Since the harvesting period is limited to
about 2 months, seasonal variation is not identified as a
major problem. Therefore, the year-to-year variation was
investigated in two subsequent years. The results in Table 5
show clearly that the range of Sr isotope ratios is almost
identical and that no significant variation within these
2 years can be observed.
In order to identify the possible source of Sr in the
asparagus sample in the soil, we investigated both the
isotope composition of total Sr and of Sr which was
extracted by a solution of NH
4
NO
3
. This extract represents
the mobile fraction or ‘bioavailable’ Sr [35]. The principle
is based on the consideration that met als in soil, which
correspond to this fraction, form soluble metal amine
complexes during the extraction. NH
4
NO
3
mobilises more
cationic elements in the soil owing to specific sorption
processes compared with, e.g., deionised water [36].
The plant uptake of mobile Sr takes place via water and
represents therefore the source of isotope information. The
rhizomes of aspara gus lie 35 cm below the earth ’s surface,
but the plant can root down to 5-m depth; therefore, its
water reservoir basically comes from the ground water.
The water source is in a soil depth beyond 1 m (the so-
called saturated zone) for deep-rooted plants. Hydraulic
redistribution studies show that water moves passively
through plant roots between deep and plane soil layers via
water gradients. The water can therefore move up and
down within the soil, depending on the environmental
conditions [37].
Minerals release monovalent and divalent alkali and
alkaline-earth cations during the formation of soil, and
these are leached. A soil profile displays a weathering
gradient from deep soil layers towards the surface [38] and
the availability of Sr and therefore the implemented isotope
composition for plant s can vary wi th the soil depth [37]. A
varying
87
Sr/
86
Sr ratio within a hydro system can therefore
provide information about the sources of Sr and the
different mixing processes involved. [39] However, we did
not make these observations within the areas investigated.
The Sr isotope ratios of the soil samples are given in
Table 6 in comparison with the Sr isotope range in
asparagus from Marchfeld. The Sr isotope ratios of total
Sr in soil are significantly different compared with the Sr
isotope composition in asparagus, which corresponds to
previous findings [38]. In addition, the range of the Sr
isotope composition (Table 6) of total soil is remarkably
wider than t hat of the solubl e fr acti on. Th is can b e
explained by the fact the Marchfeld region represents a
large lowland water reservoir filled by floods of the Danube
and Morava rivers, resulting in a variation of sediment. The
composition of the ground water in contrast c an be
expected to be homogenised over the whole area. Thus,
even if soil components could originate from geological
formations with different elemental composition (i.e. Rb/Sr
ratios) and geologic history resulting in different
87
Sr/
86
Sr
isotope ratios in bulk soil, the fraction of Sr available for
plants has a comparably narrow range of isotope compo-
sition. The Sr isotope compo sitions both of the NH
4
NO
3
extract and of the asparagus samples of locations with the
highest Sr isotope ratios correspond well to the Sr isotope
composition of soil extracts and asparagus of the other
locations in Marchfeld. Therefore, it was evident that the Sr
Table 4 Sr isotope composition in asparagus originating from different harvesting sites (2006)
Sampling sites Number of samples Range Median 2σ
Austria (Marchfeld) 75 (68 analysed) 0.7083–0.7102 0.7095 0.0008
Germany (1 site) 6 (6 analysed) 0.7081–0.7097 0.7086 0.0013
Hungary (3 site) 24 (19 analysed) 0.7056–0.7080 0.7069 0.0011
Peru (1 site) 5 (3 analysed) 0.7078–0.7081 0.7079 0.0002
Slovakia (4 site) 31 (28 analysed) 0.7062–0.7093 0.7079 0.0014
The Netherlands (4 sites) 14 (13 analysed) 0.7096–0.7105 0.7098 0.0006
Table 5 Sr isotope composition of asparagus from the Marchfeld region (2005 and 2006)
Sample Number of samples Range Median 2 σ
Asparagus harvest 2005 14 0.7081–0.7105 0.7094 0.0014
Asparagus harvest 2006 75 (68 analysed) 0.7083–0.7102 0.7095 0.0008
492 Anal Bioanal Chem (2008) 390:487–494
in the mobile phase corresponds to the Sr source which is
taken up by asparagus.
Asparagus samples from different origins
The range, median and standard deviation of the Sr isotope
ratios of the asparagus samples from other sources are
given in Table 4. It is evident that the samples from
Hungary and Marchfeld do not overlap within the total
range and they can be clearly distinguished. The range
shown in Table 4 reflects the lowest and the highest result,
whereas the median ±2 times the standard deviation reflects
the probability that a sample from the area can be identified
with a probability of 95%. Fig. 3 shows the plot of the
concentration of the asparagus samples versus the
87
Sr/
86
Sr
isotope ratios of the asparagus samples produced in Austria,
Slovakia and Hungary. Samples from other regions were
excluded because of fewer data.
The picture indicates that most asparagus samp les from
Marchfeld can be distinguished from asparagus from the
neighbouring countries of Hungary (100%) an d Slovakia
(80%). Peruvian asparagus Sr ratios lie between the data
from Hungarian and Slovakian asparagus but are different
from the values for samples from Germany, Austria and the
Netherlands.
The asparagus samples from the Netherlands also differ
from the asparagus samples from the two eastern European
countries.
It has to be mentioned that more asparagus samples
originating from the Netherlands, Germany and Peru are
necessary to represent statistically relevant data.
Conclusion
Isotope ratios of particular heavy elements are specific
functions of the local environment and therefore precise
isotope analysis of Sr by using MC-ICP-MS demonstrated
in this work is a promising tool for food authentication
studies and routine traceability applications. This study
produced a representative isotope database for authentic
Marchfeld asparagus for 2006 and determined natural
variations in isotope compositions ty pical for selected
European countries by measuring Sr isotope composition
in authentic asparagus samples originating from nearby
(Austrian, Slovakia and Hungarian regions) and distant (e.g.
German, Dutch) production sites. A particularly important
result is the demonstrated possibility to distinguish agricul-
tural products originating from neighbouring regions (e.g.
Marchfeld region, Slovakia and Hungary) by means of Sr
isotope analysis. Furthermore, it has been shown that the
annual variation of the investigated
87
Sr/
86
Sr isotope ratio is
not significant, and thus this isoto pe system shows the
potential for the establishment of a stable isotope dataset
which can be valid over a longer time period.
On the other hand, Sr isotope analysis solely does not
provide a complete and all-sufficient tool for reliable au-
thentication of Marchfeld asparagus. Thus, Dutch, German
and Marchfeld asparagus samples have similar Sr isotope
compositions and for their separation other isotope systems,
e.g. light isotopes like hydrogen (δ
2
H), carbon, ( δ
13
C),
nitrogen (δ
15
N), oxygen (δ
18
O), and sulfur (δ
34
S), must be
employed. In summary, light and heavy isotopes provide
complementary information, increasing the reliability of
food authentication and traceability. Future studies will
focus on exploring isotope systems, not yet being taken
advantage of, and on identification of the most reliable
isotope marker combination for authentication of agricul-
tural plants produced in Lower Austria.
Table 6 Sr isotope composition of soil and asparagus samples in the Marchfeld region
Sample Range Median 2σ
Asparagus: total Sr 0.7083–0.7102 0.7095 0.0008
Soil: NH
4
NO
3
-extracted Sr 0.7083–0.7098 0.7092 0.0007
Soil: total Sr 0.7137–0.7194 0.7146 0.0041
Sr concentration vs. Sr isotope ratio
0
100
200
300
400
500
600
700
0.705 0.706 0.707 0.708 0.709 0.71 0.711
87
Sr/
86
Sr Isotope ratio
Concentration µg g
-1
Austria
Slovakia
Hungary
Fig. 3 Sr concentration versus Sr isotope ratio of the samples
originating from Austria (Marchfeld), Slovakia (all sites) and Hungary
(all sites). Relative standard deviation less than 0.03%. Error bars
cannot be seen because of their small range
Anal Bioanal Chem (2008) 390:487–494 493
Acknowledgements This work was supported by the Austrian
Science Fund FWF (START Project 267 N11). This study was
partially funded by the independent research programme of Austrian
Research Center GmbH. Some of the asparagus samples were
provided by the Verein der Marchfelder Spargelbauern, which is
gratefully acknowledged.
References
1. John MD (1998) Analyst 123:151R–156R
2. Cosio MS, Ballabio D, Benedetti S, Gigliotti C (2007) Food
Chem 101:485–491
3. Consolandi C, Palmieri L, Doveri S, Maestri E, Marmiroli N,
Reale S, Lee D, Baldoni L, Tosti N, Severgnini M, De Bellis G,
Castiglione B (2007) J Biotechnol 129:565–574
4. Bonizzi I, Feligini M, Aleandri R, Enne G (2006) J Appl
Microbiol 102:667–673
5. Ruoff K, Luginbuhl W, Kunzli R, Bogdanov S, Bosset JO, von
der Ohe K, von der Ohe W, Amado R (2006) J Agric Food Chem
54:6858–6866
6. Tay A, Singh RK, Krishnan SS, Gore JP (2002) Lebensm Wiss
Technol 35:99–103
7. Brescia MA, Alviti G, Liuzzi V, Sacco A (2003) J Am Oil Chem
Soc 80:645–650
8. Rossmann A, Haberhauer G, Hölzl S, Horn P, Pichlmayer F,
Voerkelius S (2000) Eur Food Res Technol 211:32–40
9. Boner M, Förstel H (2004) Anal Bioanal Chem 378:301–310
10. Camin F, Wietzerbin K, Cortes AB, Haberhauer G, Lees M,
Versini G (2004) J Agric Food Chem 52:6592–6601
11. Camin F, Bontempo L, Heinrich K, Horacek M, Kelly SD,
Schlicht C, Thomas F, Monahan FJ, Hoogewerff J, Rossmann A
(2007) Anal Bioanal Chem 389:309–320
12. Schmidt O, Quilter JM, Bahar B, Moloney AP, Scrimgeour CM,
Begley IS, Monahan FJ (2005) Food Chem 91:545–549
13. Anderson KA, Smith BW (2006) J Agric Food Chem 54:1747–1752
14. Castineira Gómez MdM., Feldmann I, Jakubowski N, Andersson
JT (2004) J Agric Food Chem 52:2962–2974
15. Prohaska T, Köllensperger G, Krachler M, De Winne K, Stingeder
G, Moens L (2000) J Anal At Spectrom 15:335–340
16. Barbaste M, Halicz L, Galy A, Medina B, Emteborg H, Adams
FC, Lobinski R (2001) Talanta 54:307–317
17. Kawasaki A, Oda H, Hirata T (2002) Soil Sci Plant Nutr 48:635–640
18. Almeida CM, Vasconcelos MT (1999) Anal Chim Acta 396:45–53
19. Almeida CM, Vasconcelos MT (2001) J Anal At Spectrom
16:607–611
20. Almeida CM, Vasconcelos MT (2003) J Agric Food Chem
51:3012–3023
21. Almeida CM, Vasconcelos MT (2004) Food Chem 85:7–12
22. Milhajevic M, Ettler V, Šebek O, Ladislav Strnad L, Chrastný V
(2006) J Geochem Explor 88:130–133
23. Larcher R, Nicolini G, Pangrazzi P (2003) J Agric Food Chem
51:5956–5961
24. Barbaste M, Medina B, Sarabia L, Ortiz MC, Pérez-Trujillo JP
(2002) Anal Chim Acta 472:161–174
25. Barbaste M, Halicz L, Galy A, Medina B, Emteborg H, Adams
FC, Lobinski R (2001) Talanta 54:307–317
26. Augagneur S, Medina B, Grousset F (1997) Fresenius J Anal
Chem 357:1149– 1152
27. Coetzee PP, Vanhaecke F (2005) Anal Bioanal Chem 383:977–984
28. Stein M, St arinsky A, Katz A, Goldstein SL , Machlus M,
Schramm A (1997) Geochim Cosmochim Acta 61:3975–3992
29. Sandroni V, Smith CMM, Donovan A (2003) Talanta 60:715–723
30. Sastre J, Sahuquillo A, Vidal M, Rauret G (2002) Anal Chim Acta
462:59–72
31. DIN V 19730 (1997): Bodenbeschaffenheit-Extraktion von
Spurenelementen mit Ammoniumnitratlösung. Beuth, Berlin
32. Prohaska T, Latkoczy C, Schultheis G, Teschler-Nicola M,
Stingeder G (2000) J Anal At Spectrom 17:887–891
33. Albarède F, Télouk P, Blichert-Toft J, Boyet M, Agranier A,
Nelson B (2004) Geochim Cosmochim Acta 68:2725–2744
34. EIChrom (2007) http://www.eichrom.com/products/info/sr_resin.
cfm. Accessed 16 Mar 2007
35. Zeien H, Brümmer GW (1989) Mitt. Dtsch. Bodenkdl. Ges.
59:505–510
36. Ettler V, Mihaljevič M, Šebek O, Nechutný Z (2007) Chemosphere
68:455–463
37. Mc Culley R, Jobbágy E, Pockman W, Jackso n, R (2 004)
Oecologia 141:620–628
38. Prohaska T, Wenzel WW, Stingeder G (2005) Int J Mass Spectrom
242:243–250
39. Negrel P, Grosbois C (1999) Chem Geol 156:231–249
494 Anal Bioanal Chem (2008) 390:487–494