Soil and Environmental Analysis: Modern Instrumental Techniques - Chapter 9 docx

9 Stable Isotope Analysis and Applications Charles M. Scrimgeour The Scottish Crop Research Institute, Dundee, Scotland David Robinson The University of Aberdeen, Aberdeen, Scotland 1. INTRODUCTION The biologically important light elements are hydrogen, carbon, nitrogen, oxygen, and sulfur. Each has at least two stable isotopes, and the most abundant isotope in a pair is the lighter: 2 H/ 1 H (i.e., D/H), 13 C/ 12 C, 15 N/ 14 N, 18 O/ 16 O, and 34 S/ 32 S. Variations in isotope abundances can reveal and quantify processes in which these elements are involved. Such processes include photosynthesis, respiration, evaporation, organic matter turnover, and C, N, and S metabolism. Stable isotopes can also be used in activities as diverse as monitoring pollution events, tracking animals’ food sources, reconstructing past climates, identifying plants’ water sources, and untangling biochemical pathways. Valuable general references include Fritz and Fontes (1980), Vose (1980), O’Leary (1981, 1988, 1993), Hoefs (1987), Raven (1987), Rundel et al. (1989), Coleman and Fry (1991), Griffiths (1991, 1998), Krouse and Grinenko (1991), Robinson and Smith (1991), Handley and Raven (1992), O’Leary et al. (1992), Ehleringer et al. (1993), Engel and Macko (1993), Knowles and Blackburn (1993), Lajtha and Michener (1994), Boutton and Yamasaki (1996), Handley and Scrimgeour (1997), Kendall and McDonnell (1998), Bingham et al. (2000), Mook (2001), Robinson (2001), and Dawson et al. (2002). The Internet is being used increasingly as a source TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. of the latest stable isotope information. The ISOGEOCHEM website at http://geology.uvm. edu/isogeochem.html is a good place to start; Kendall and Campbell (1998) list others. Many of the approaches described in these references rely on isotopes being used as tracers. An isotopically distinct, but chemically indistingui- shable, material, the tracer, is introduced into an experimental system, and isotope abundances are later measured in particular compartments of that system. Sometimes the tracer is a natural ly occurring substance that happens to be isotopically distinct from others in the system. Increasingly, however, use is being made of isotope fractionations. These can report on the operation of chemical and physical processes that change the natural isotopic abundances of substrates and products involved in those processes. The tracer and fractionation approaches are conceptually distinct and capable of addressing different research questions (Table 1). Each approach demands its own theory and protocols, but both require similar instrumentation. Most of this chapter (Secs. III and IV) describes current instrumentation and analytical techniques used for stable isotope analysis. It relies heavily on our experience of automated, continuous-flow mass spectrometers to analyze the isotopic contents of soil, plant, and animal samples. Examples of tracer and fractionation applications are discussed in Sec.V. Section VI is, finally, a brief preview of future developments. We begin, however, with an overview of terminology. II. TERMINOLOGY A. Isotope Ratio Mass spectrometers (see Sec. III) measure the ratio (R) of heavy to light isotopes: R ¼ n H n L ð1Þ where n H and n L are the numbers of atoms containing heavy and light isotopes, respectively. For example, if five out of every 100 N atoms in an N sample are 15 N and the rest 14 N, the sample’s 15 N/ 14 N ratio is 5/95 ¼ 0.0526. Isotope ratios are usually converted into more convenient quantities. For tracer work, atom percentages are suitable; for natural abundances,  values are more appropriate. 382 Scrimgeour and Robinson TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. Table 1 Tracer and Fractionation Approaches Compared Tracer approach Labeled tracers Natural tracers Fractionation approach Isotope abundance range Much greater than natural abun- dance range Within natural abundance range Within natural abun- dance range Extent of perturbation to system Large Small or zero Zero Cost of tracer Potentially huge Zero Zero Sensitivity of detection Excellent Poor to good Poor to excellent Appropriate duration of study < 1h to 1 yr Unsuitable for short-term (< 1 h) studies < 1h to > 3.8 10 9 yr (Schidlowski, 1988) Appropriate scale of investigation Usually pot or small plot, but for lightly enriched tracers, small catchment studies are feasible (Nadelhoffer and Fry, 1994) Pot to landscape Molecular to global Conditions required Isotopic composition of tracer greater than natural range. Steady-state labeling achieved within sinks (Dele ´ ens et al., 1994) Reliable and distinct differ- ences in isotopic composition of all potential source pools Reliable measurements of isotopic composi- tions of all potential source pools continued Stable Isotope Analysis and Applications 383 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. Table 1 Continued Tracer approach Labeled tracers Natural tracers Fractionation approach Information required Isotopic composition of tracer before addition to system; system components before addition; system components after addi- tion. Amount of tracer added Isotopic composition of all potential tracer sources; of system containing no tracer; of components of system containing tracer. Fluxes among pools if more than two sources are involved Isotopic composition of all important pools. Amounts of element in each pool. Fractionation factors for candidate processes Interpretive model Mixing Mixing Fractionation and mixing Information obtained Amounts and rates of mixing of tracer in nontracer pools Amounts and (possibly) rates of mixing of tracer in nontracer pools Quantitative identification of likely processes causing fractionations 384 Scrimgeour and Robinson TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. B. Atom Percentage The abundance of the heavier isotope in an isotope pair as the fraction of the total amount of the element is the atom fraction or, more usually, atom percentage (A): A ¼ 100 n H n H þ n L ¼ 100 R R þ 1 ð2Þ In the example for 15 N given above, A is [5/(5 þ95)] 100 ¼5 atom %. The mass percentage (A m )of 15 N in this sample is not, however, 5%. It is calculated by multiplying n for each isotope by its atomic mass, m, so that A m ¼ 100 m H n H m H n H þ m L n L ð3Þ where m L and m H are the masses of the light and heavy isotopes, respectively. For our N sample, m L and m H are 14 and 15, respectively, and A m is 5.338%. A and A m are related by A m ¼ 100 m H A 100m L þ Aðm H  m L Þ ð4Þ Using Eq. (2) rather than Eq. (3), as often happens in practice, slightly underestimates the true mass fraction. The amount (X  ) of an element in a sample that is derived form a tracer is given by X  ¼ XðA sample  A background Þ½m L ð100  A tracer Þþm H A tracer  ðA tracer  A background Þ½m L ð100  A sample Þþm H A sample  ð5Þ where X is the total amount of the element in the sample, A sample is the sample’s atom % (Eq. 2), A tracer is the atom % of the trace r originally added, and A background is the background atom % in the system before the tracer was added. For a given isotope pair, Eq. 5 is simplified considerably by substituting the appropriate values for m L and m H . Let us suppose that our N sample for which A sample ¼5 atom % is from an experiment to which a tracer containing 7.5 atom % 15 N had been added (A tracer ), and assume that the background 15 N abundance in the system (A background ) is 0.3663 atom % (cf. Table 2; see Sec. V.A.1). If the sample contains a total of Stable Isotope Analysis and Applications 385 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. 100 g N (X), then the amount of N in the sample that is derived from the tracer (X*) is 65.1 mg. The term atom percent enrichment or atom percent excess (APE) is used frequently. This is simply the difference in atom % between a sample and background. In Eq. (5), the terms (A sample A background ) and (A tracer  A background ) are APE values. C. d Notation The natural abundances of D, 13 C, 15 N, 18 O, and 34 S range from only 0.01 to 4 atom % (Fig. 1). Close to natural abundance, it is more convenient to express the isotope ratio of a sample as the relative difference from that of a standard; this is the  notation.  values are expressed in parts per thousand or ‘per mille’ (ø). The  value of a sample,  sample , is given by  sample ¼ 1000 R sample  R standard R standard ð6Þ where R sample and R standard are the isotope ratios of sample and standard, respectively [Eq. (1)]. Values of R standard are listed in Table 2, with the corresponding atom % values. In practice, working standards are calibrated against these primary standards using materials supplied by the interna- tional Atomic Energy Agency (Vienna), the Los Alamos National Laboratory (U.S.A.), and other agencies. By definition [Eq. (6)], each standard in Table 2 has a  value of 0ø. Samples with negative  values are ‘‘depleted’’ in the heavier isotope relative to the standard; those that are positive are ‘‘enriched’’ (see Kendall and Caldwell, 1998). For example, if a sample has a 13 C/ 12 C ratio of 0.0111372, this differs by only 0.0001 from the standard (Table 2). This is Table 2 Heavy : Light Isotope Ratios (R standard ) and Atom % Values (A) in the International Standards Used for the Analysis of D/H, 13 C/ 12 C, 15 N/ 14 N, 18 O/ 16 O, and 34 S/ 32 S. By Definition, the  Value of Each is 0ø Isotope pair Standard material R standard A D/H Vienna Standard mean Ocean Water (V-SMOW) 0.00015576 0.01557 18 O/ 16 O Vienna Standard mean Ocean Water (V-SMOW) 0.00200520 0.20012 18 O/ 16 O Vienna PeeDee Belemnite (V-PDB) 0.0020671 0.20628 13 C/ 12 C Vienna PeeDee Belemnite (V-PDB) 0.0112372 1.11123 15 N/ 14 N Atmospheric N 2 0.0036765 0.3663 34 S/ 32 S Canyon Diablo Troilite 0.0450045 4.30663 386 Scrimgeour and Robinson TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. equivalent to  13 C ¼8.9ø in the sample, i.e., it is 13 C-depleted compared with the standard. For a particular isotope pair,  and A are related by  ¼ 1000 A R standard ð100  AÞ  1  ð7Þ where R standard is the value in Table 2 appropriate for the isotope pair. D. Fractionation and Discrimination D, 13 C, 15 N, 18 O, and 34 S occur naturally in varying amounts in different materials. These variations reflect isotopic fractionations of the heavier and lighter isotopes in a pair. Fractionations occur because more energy is needed to break or form chemical bonds involving the heavier isotope of a pair (Atkins, 1998, p. 833). For a reaction occurring over an infinitesimal time interval, a fractionation factor, a, can be defined. This is the isotope ratio of the substrate divided by that of the product for that time interval:  ¼ R substrate R product ð8Þ Figure 1 Natural abundances of 2 H/ 1 H, 13 C/ 12 C, 15 N/ 14 N, 18 O/ 16 O, and 34 S/ 32 S. The insets show the range of natural variation in isotope ratio and the  values (ø)to which these correspond. Stable Isotope Analysis and Applications 387 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. When a > 1, the heavier isotope accumulates in the substrate as a reaction proceeds; when a< 1, it accumulates in the product. If a¼1, R substrate ¼ R product and there is no fractionation. (Note that some authors define a as R product /R substrate : e.g., Mariotti et al., 1981). An expression that translates a values onto a ø scale for direct comparison with a values is " ¼ 1000ð 1Þð9Þ where " is the instantaneous isotopic enrichment factor; see Mariotti et al. (1981). a and " values are not strictly constants, but depend on temperature, the identities of the reactants (including any enzymes that mediate the reaction), and bond energies (Atkins, 1998, p. 833). This is true whether the fractionation occurs during a unidirectional kinetic reaction (e.g., NH þ 4 þ OH  ! NH 3 þ H 2 O) or in a system at equilibrium (e.g., NH þ 4 þ OH  $ NH 3 þ H 2 O). In each of these examples, the NH þ 4 becomes more 15 N- enriched than the NH 3 . For a given reaction under defined, closed conditions, however, a is effectively constant irrespective of substrate availability and may be characteristic of the reaction. a values for some biologically important reactions are tabulated in Friedman and O’Neill (1977), Leary et al. (1992), Handley and Raven (1992), O’Leary (1993), Nordt et al. (1996), Wada and Ueda (1996), and Handley et al. (1999). Most of these indicate the magnitudes of fractionations when substrate availability is not limiting and other conditions are favorable. They do not necessarily indicate the fractionations that occur in vivo and which are often smaller than those in vitro. As a reaction proceeds, the  values of substrates and products change in a predictable way, as described by Rayleigh equations (Mariotti et al. 1981; Hoefs, 1987). The  value of the substrate ( S ) depends on its initial  value ( 0 ), on " (Eq. 9), and the fraction ðÞ of the substrate that has been consumed in the reaction:  S ¼  0  " lnð1 Þð10Þ The  value of the instantaneous product ( Pi ) is approximated as  Pi   S  " ð11Þ The product created in any particular time-step mixes with that from earlier time-steps. The resulting  value of the accumulated product ð "  P Þ is given by "  P ¼  0 þ "ð1  Þ½lnð1  Þ  ð12Þ 388 Scrimgeour and Robinson TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. Figure 2 illustrates how  S ,  Pi and "  P vary with . Equations 10–12 apply strictly to a unidirectional single-step reaction in a closed system. In such a reaction, the final product has the same isotopic composition as the initial substrate, i.e., as  tends to 1, "  P tends to  0 . Equations 10–12 can, however, also be applied to natural systems comprising multiple open reactions, especially if one fractionating process dominates (O’Leary, 1988; Sec. V.C.3). Open reactions involve the addition of new substrate and/or the removal of accumulated product, and are never ‘‘completed’’ Isotopic differences between substrates and products persist (although the differences are not necessarily constant). Such differences are termed isotope Figure 2 Changes in  values of substrate and product in a closed system as a function of the fraction () of substrate consumed in a reaction. The initial  of the substrate in this example is 0ø. The  values of substrate, instantaneous product, and accumulated product are calculated using Eqs. (10–12). " is the instantaneous isotope fractionation factior [Eq. (9)], which is constant and, in this example, ¼10ø. Discrimination (Á; Eq. (14)] is not constant but approximates to the difference between the  values of substrate and accumulated product. Only when  0 does Á ¼". Stable Isotope Analysis and Applications 389 TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. discriminations (symbolized as Á). If substrate availability is effectively unlimited (i.e., if   0: see Fig. 2), then Á  " ð13Þ Combining Eqs. 6, 8, 9, and 13 gives (O’Leary, 1981) Á ¼  substrate   product 1 þð product =1000Þ   substrate   product ð14Þ A positive discrimination indicates that the heavier isotope accumu- lates in the substrate ( substrate > product ); a negative discrimination indicates the opposite. The most extensive biological use of Á has been to compare dis- criminations against 13 C during photosynthesis by C 3 plants (Sec. V.C.2). Unfortunately, other systems are not so amenable. For example, it is not yet possible to calculate Á for N assimilation by plants growing in soil. This is because the availability and  15 N values of putative substrates [e.g., NO  3 , NH þ 4 , dissolved organic-N (DON)] at the assimilatory site(s) or metabolic branch points (O’Leary, 1981) where 15 N/ 14 N fractionations may occur cannot be assumed or measured reliably by current methods (Sec. V.C.3). E. Isotope Mass Balances and Mixing Models One of the most useful and frequently encountered isotope equations is the isotope mass balance. The relates the  value of a composite sample to those of its components, each weighted by its mass. If a sample has two components of mass X and Y with  values  X and  Y , respectively, then  value of the composite sample ( " )is "  ¼ X X þ Y Y X þ Y ð15Þ If it has more than two components, Eq. (15) is modified accordingly. Depending on the available information, Eq. (15) can be solved to estimate an unknown  value or mass. For example, suppose one wished to enrich 1 kg of C 3 plant material with 13 C so that its  13 C value was about 500ø. How much 13 C-enriched CO 2 containing 5 atom % 13 C would be needed? 1 kg (fresh weight) of plant material would contain about 40 g C (X)witha 13 C value ( Y ) of about 27ø. 5 atom % 13 C is equivalent to a  13 C value ( Y ) of 3684ø [Eq. (7)]. 390 Scrimgeour and Robinson TM Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved. [...]... hydraulic lift by deep-rooted vegetation (Caldwell and Richards, 198 9) Tracing recently assimilated CO2 in ´ whole plants (Schnyder, 199 2; Deleens et al., 199 4) Identifying water sources (rainvs groundwater) for trees (White et al., 198 5; Dawson, 199 3) Estimating C transfers among various organic matter pools in soil (Balesdent and Mariotti, 199 6), trees and fungi (Hogberg et al., ¨ 199 9) Recovery of N... geological specimens and biological processes (Canfield and Teske, 199 6; Habicht and Canfield, 199 6) Tracing recently assimilated S in whole plants (Monaghan et al 199 9) Copyright n 2004 by Marcel Dekker, Inc All Rights Reserved Identifying origins of soil NOÀ 3 (Wassenaar, 199 5)(15N also used) Simulating effects of atmospheric S deposition on soils (Preitzel et al., 199 4; Krouse et al., 199 6) Identifying... et al., 198 6; Powlson and Barraclough, 199 3) Quantifying water and C turnover in animal metabolism (Nagy, 198 3) (D also used) Identifying NH3 use by vegetation (Erskine et al., 199 8) Analyzing water fluxes and energy balance of field crops (Bariac et al., 198 9) (18O also used) Analyzing gas exchange of C3 plants in terms of physical and biochemical processes (Farquhar et al., 198 2; Vogel, 199 3) Reconstructing... Reconstructing prehistoric climate and vegetation distributions (Street-Perrott et al., 199 7) and carbonate pedogenesis (Nordt et al., 199 6) Detecting denitrification using 15N values of residual NOÀ (Mariotti 3 et al., 198 8) Analyzing processes (e.g., photosynthesis, respiration) that influence gas exchange between vegetation and the atmosphere (Farquhar et al., 199 3; Yakir and Wang, 199 6) Reconstructing the evolution... atmospheric N inputs to, and net N losses from, the system to be estimated TM Copyright n 2004 by Marcel Dekker, Inc All Rights Reserved 416 Scrimgeour and Robinson More sophisticated 15N techniques allow N fluxes through various soil and microbial pools to be derived (Powlson and Barraclough, 199 3; Hart and Myrold, 199 6; Mary et al., 199 8) These rely on initially labeling either soil NHþ or NOÀ pools... Pichlmayer and Blochberger ( 198 8) and Haystead ( 199 1) However, it was somewhat later that this technique was reassessed and widely used (Giesemann et al., 199 4) Even then, it has been more commonly applied to mineral S prepared from biological samples than to intact plant or animal samples We have successfully used CF-IRMS for S in plant samples with as little as 0.1% S (Monaghan et al., 199 9) and in a... be on- or off-line DI-IRMS are still widely used, despite now being replaced in many laboratories by more convenient continuous-flow IRMS (Sec III.D) DIIRMS remain the most usual instruments for measuring D/H, but recent developments in continuous-flow approaches (Prosser and Scrimgeour, 199 5; Begley and Scrimgeour, 199 7) will change this in the future D Continuous-Flow IRMS The development of solid-state... of terrestrial and marine animals with $ 0.5% S (Neilson, 199 9) Most soil samples we have examined contained too little S for direct analysis by CF-IRMS and would still require the extraction of relatively large samples before analysis (e.g., Krouse et al., 199 6) 34 1 Instrumentation The general principles and methods described above for C and N isotopes apply to S isotope analysis by CF-IRMS The conditions... isotope technique has been the use of 15N-enriched fertilizer to estimate how much fertilizer N is captured by crops Powlson and Barraclough ( 199 3) described the practicalities of introducing 15N-labeled fertilizers into agricultural soil, as did Watkins and Barraclough ( 199 3) for 15N-labeled crop residues Powlson et al ( 198 6) measured the N balance in a long-term winter wheat experiment In spring they... or three working standards used as dummies, followed by a working standard, and then ten samples, and then a pair of working standards The first working standard (sometimes referred to as a check standard) is used for quality control and the second as a standard The check standard can also be substituted for the standard if there is a problem such as an electrical spike while the standard is being measured . Coleman and Fry ( 199 1), Griffiths ( 199 1, 199 8), Krouse and Grinenko ( 199 1), Robinson and Smith ( 199 1), Handley and Raven ( 199 2), O’Leary et al. ( 199 2), Ehleringer et al. ( 199 3), Engel and Macko ( 199 3), Knowles. tabulated in Friedman and O’Neill ( 197 7), Leary et al. ( 199 2), Handley and Raven ( 199 2), O’Leary ( 199 3), Nordt et al. ( 199 6), Wada and Ueda ( 199 6), and Handley et al. ( 199 9). Most of these indicate. Engel and Macko ( 199 3), Knowles and Blackburn ( 199 3), Lajtha and Michener ( 199 4), Boutton and Yamasaki ( 199 6), Handley and Scrimgeour ( 199 7), Kendall and McDonnell ( 199 8), Bingham et al. (2000),