Climatic influences on the oxygen isotopic composition of biogenic silica in prairie grass

Geochimica et Cosmochimica Acta, Vol. 66, No. 11, pp. 1891–1904, 2002 Copyright © 2002 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/02 $22.00 ⫹ .00

Pergamon

PII S0016-7037(02)00822-0

Climatic influences on the oxygen isotopic composition of biogenic silica in prairie grass ELIZABETH A. WEBB*,† and FRED J. LONGSTAFFE* Department of Earth Sciences, The University of Western Ontario, London, Ontario N6A 5B7, Canada (Received February 16, 2001; accepted in revised form November 27, 2001)

Abstract—Samples of Calamovilfa longifolia were collected from across the North American prairies to investigate the relationship between the oxygen-isotope composition of biogenic silica (phytoliths) deposited in this grass and relative humidity, temperature, and the oxygen-18 enrichment of soil water relative to local precipitation. The ␦18O values of silica in nontranspiring tissues were controlled by soil-water composition and temperature, whereas the oxygen-18 content of silica formed in leaf and inflorescence tissues was enriched further by transpiration. Accurate calculation of growing temperature was possible only when the oxygenisotope compositions of both stem silica and soil water were known. However, the oxygen-isotope values of stem phytoliths can be used to calculate the variation in the isotopic composition of soil water across a North American temperature gradient. As plant organic matter decays and phytoliths are transferred to the soil, the temperature and soil-water signals carried by the oxygen-isotope composition of silica from nontranspiring tissues can be masked by the oxygen-18 enrichment of phytoliths from transpiring tissues. However, the overall oxygen-isotope composition of a soil-phytolith assemblage can be related to temperature using an empirical relationship based on temperature and the difference between soil-phytolith and estimated soil-water oxygen-isotope compositions. Copyright © 2002 Elsevier Science Ltd al., 1996; Webb and Longstaffe, 2000) and in fossil soilphytolith assemblages (Fredlund, 1993). In this study we examine the variations in ␦18O values of silica phytoliths within one grass species over a range of climatic conditions. If the integrated effects of temperature and relative humidity on the ␦18O values of phytoliths are predictable for living grasses, potential then exists to use such data as a paleoclimatic indicator.

1. INTRODUCTION

Aqueous silicic acid (Si(OH)4) enters plants through their roots and is polymerized as amorphous opal-A (phytoliths) in cells and intercellular spaces along the transpiration stream (Raven, 1983). Silica phytoliths are believed to form in temperature-dependent, oxygen-isotope equilibrium with plant water (Shahack-Gross et al., 1996; Webb and Longstaffe, 2000). Hence, the oxygen isotopic composition of phytoliths has potential as a proxy for temperature and the ␦18O values of precipitation during plant growth. Use of this paleoclimate proxy is complicated by the possible variations in plant-water oxygen-isotope compositions. Not only does the oxygen-isotope composition of precipitation vary as a function of season, temperature, and latitude, but plant-water ␦18O values are also dependent on the susceptibility of water in soil and plant tissues to become enriched in 18O and D by kinetic and equilibrium fractionation associated with evaporation and transpiration (Allison et al., 1984; Flanagan and Ehleringer, 1991; Farquhar and Lloyd, 1993). Hence, the ␦18O values of phytoliths vary within individual plants as well as between plants grown in different climates. Phytoliths are produced in all plant species but are particularly abundant in grasses (Metcalfe, 1960). The production of phytoliths over wide geographic and climatic gradients, as well as their persistence in Recent to Quaternary soils, makes them available for paleoclimate studies in the terrestrial environment. The oxygen isotopic composition of phytoliths has been examined under controlled or closely monitored conditions for living grasses (Bombin and Muehlenbachs, 1980; Shahack-Gross et

2. METHODS 2.1. Sample Collection Calamovilfa longifolia, a C4 grass, was collected across the mid and northern prairies in Canada and the United States at the end of the 1995 and 1996 growing seasons (late August to early September) (Table 1, Fig. 1). This species produces abundant phytoliths, thrives in a variety of climates, and generally grows in sandy soils, where high infiltration rates minimize the evaporative loss of rainwater. We expect that the oxygen-isotope behaviour of phytoliths from C. longifolia will be representative of the majority of grass silica in the North American Great Plains. Comparable behaviour has been observed for the ␦18O values of silica extracted from the C3 grasses Triticum aestivum (Shahack-Gross et al., 1996) and Ammophila breviligulata (Webb and Longstaffe, 2000). Silica was extracted from dozens of individual plants at each location and represents an average for opal-A deposited in the grass over its lifetime. Each sample was divided into six parts: roots, rhizomes, stems, sheaths, leaves, and inflorescences. Silica was extracted from organic matter with sulfuric acid and hydrogen peroxide (Webb and Longstaffe, 2000). Purity of the extract was confirmed using X-ray diffraction. Gypsum and anhydrite, present in a few leaf-silica samples, are artifacts of the reaction between sulfuric acid and biogenic calcium oxalate. These sulphates were removed by dissolution in HCl. Meteoric water was collected from small local streams, ponds, lakes, or shallow ground water in close proximity to grass-sample locations. At Pinery Provincial Park, monthly precipitation samples were also available. For locations where local surface water, ground water, or precipitation were not accessible, appropriate data were gathered from the literature (Table 1). Summer precipitation is defined as the average

* Authors to whom correspondence should be addressed (ewebb@ mcmaster.ca, [email protected]). † Present address: School of Geography and Geology, McMaster University, 1280 Main Street West, Hamilton, ON, L8S 4M1, Canada. 1891

1892

E. A. Webb and F. J. Longstaffe Table 1. Sample locations and climatic informationa Temperatureb (°C) Location City, Province (State), Country

Latitude

Longitude

Year

1. Pinery, ON, CAN

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

43°19⬘N 81°45⬘W 1995 early summer (May–June) 1996 late summer (May–August) 1996 1997 Colorado Springs, CO, USA 38°51⬘N 104°45⬘W 1996 Havanna, IL, USA 40°18⬘N 90°03⬘W 1996 Kellogg, MN, USA 44°18⬘N 91°59⬘W 1995 Aroya, CO, USA 38°51⬘N 103°07⬘W 1996 Quinter, KS, USA 39°04⬘N 100°13⬘W 1996 Thedford, NE, USA 42°02⬘N 100°49⬘W 1995 Bonner, NE, USA 41°56⬘N 103°01⬘W 1995 Fertile, MN, USA 47°32⬘N 96°16⬘W 1995 Cheyenne, WY, USA 41°20⬘N 104°34⬘W 1995 Onefour, AB, CAN 49°07⬘N 110°28⬘W 1995 Dundurn, SA, CAN 51°49N 106°30⬘W 1995 Sprucewoods, MA, CAN 49°52⬘N 99°22⬘W 1995 Rawlins, WY, USA 42°03⬘N 106°56⬘W 1995 Broadus, MT, USA 46°24⬘N 106°16⬘W 1995 Kinsella, AB, CAN 53°00⬘N 111°32⬘W 1995 Kortes Dam, WY, USA 42°03⬘N 106°56⬘W 1995

May– Aug

July– Aug

20 15 18 17 19 21 20 21 22 20 19 18 16 16 15 17 15 20 14 15

22 20 19 20 23 22 22 24 25 24 20 21 19 16 19 19 24 14 19

Max/Minb RH (%)

Meteoric waterc ␦18O (‰)

Precip. May– July– (mm) Aug Aug annual 92/54 90/53 90/53 97/59 85/32 94/54 89/47 89/39 94/49 89/41 92/39 90/46 84/38 79/35 91/44 91/45 83/30 84/36 91/47 83/31

95/57 90/52 97/61 87/36 94/52 93/54 91/41 94/48 89/35 90/29 92/50 75/24 78/34 99/53 95/46 75/18 79/30 95/54 75/19

919 1229 1229 900 408 797 715 408 635 694 512 620 478 384 370 545 418 375 389 418

surface water

ground water

⫺7.2 ⫺7.2 ⫺7.2 ⫺7.2 ⫺5.8 ⫺6.2 ⫺8 (f) ⫺7.2 ⫺5.9 ⫺10.5

t

Soil water ␦18O (‰)

precip.

M-A temp t ⫽ 20⬚C

⫺10 (d) ⫺6.0 ⫺6.9 ⫺10 (d) ⫺5.8 ⫺9.3 ⫺10 (d) ⫺6.8 ⫺9.2 ⫺10 (d) ⫺6.1 ⫺8.5 ⫺13.5 ⫺7.8 ⫺6.4 ⫺4.1 (e) ⫺7.3 ⫺7 (g) ⫺6 (h) ⫺7.7 ⫺10.0 ⫺7.6 ⫺9.7 (i) ⫺7.9 ⫺12.1 ⫺5.8 (j) ⫺8.8 ⫺12.1 ⫺9.6 (s) ⫺9.0 ⫺12.7 ⫺8 (k) ⫺10.0 ⫺14.3 ⫺17 (l) ⫺11.5 ⫺14.5 (m) ⫺18 (d & n) ⫺14.2 (s) ⫺12.2 ⫺11.5 ⫺19 (d) ⫺13.6 (e) ⫺12.8 ⫺11.0 ⫺14.5 ⫺12.3 ⫺15.6 ⫺14.9 (s) ⫺13.2 ⫺14.4 ⫺19 (o) ⫺12.2 (s) ⫺11.7 ⫺18.5 (p) ⫺18.9 (q) ⫺15.6 (r) ⫺14.0 ⫺16.4 ⫺14.3

⫺6.9 ⫺7.5 ⫺8.5 ⫺7.5 ⫺7.5 ⫺7.6 ⫺7.7 ⫺7.9 ⫺8.6 ⫺8.8 ⫺9.3 ⫺10.1 ⫺10.8 ⫺11.0 ⫺11.2 ⫺11.4 ⫺11.7 ⫺11.9 ⫺12.5

a

Climatic information from Environment Canada, and the online Climate Visualization System, National Climatic Data Center. Average of daily values; RH ⫽ relative humidity. Average values from local rivers, lakes, summer precipitation, and shallow groundwater measured in this study and (d) Fritz et al. (1987); (e) IAEA (1992); (f) Yapp (1979); (g) median value from Siegal (1989); (h) Hunt et al. (1997); (i) this study and Clarke et al. (1998); (j) Fricke et al. (1998); (k) median value from LaBaugh et al. (1997); (l) Back et al. (1983); (m) this study and Flanagan et al. (1991); (n) Hendry and Schwartz (1988); (o) median value from USGS (1984); (p) this study and Maule´ et al. (1994); (q) this study and Fortin et al. (1991); (r) Maule´ et al. (1994); (s) Luz et al. (1990). t Soil-water values were calculated using the May to August temperatures or t ⫽ 20°C and Eq. 7 for temperature-dependent fractionation of phytoliths (see text). b c

␦18O value of rainfall over the growing season (April or May to August). 2.2. Isotopic Analysis The oxygen-isotope results are expressed in the standard ␦-notation, relative to VSMOW (Coplen, 1994) where R represents 18O/16O and:

␦ ⫽ 关共Rsample/Rstandard兲 ⫺ 1兴 ⫻ 1000 共‰兲.

(1)

An Optima, dual inlet, stable-isotope-ratio mass-spectrometer was used for all measurements. The difference in ␦-values between two phases (a and b) is expressed as: ⌬18Oa-b ⫽ ␦18Oa ⫺ ␦18Ob.

(2)

Two milliliters of each water sample were equilibrated with CO2 for oxygen-isotope analysis (Epstein and Mayeda, 1953). Reproducibility was better than ⫾0.1‰. Oxygen was liberated from silica by reaction with BrF5 (Clayton and Mayeda, 1963). During these experiments, the ␦18O value of the laboratory’s standard quartz was consistent with a value of ⫹9.7 ⫾ 0.2‰ for NBS-28. Before analysis, the oxygen isotopic exchange procedure for opal-A described by Juillet-Leclerc and Labeyrie (1987) was employed to fix the oxygen isotopic composition of hydroxyl groups, which exchange with unstable Si-O bonds during dehydration. Separate aliquots of the same silica sample, maintained at 200°C, were exchanged with the vapour produced at 0°C over two different isotopically labeled waters (␦18OH2O ⫽ ⫹41.6‰ and – 0.9‰; Webb and Longstaffe, 2000). The oxygen-isotope composition of the nonexchangeable oxygen (␦silica), and the percentage of exchangeable oxygen (X) in each phytolith sample, were calculated using the following equation:

␦silca ⫽ [␦measured 1 or 2 ⫺ X (␦exchanged 1 or 2)]/(1 ⫺X) ,

(3)

which also requires ␦exchanged 1 or 2 values, calculated following JuilletLeclerc and Labeyrie (1987). The standard deviation of phytolith ␦18O values averaged ⫾0.2‰, both when an exchanged silica sample was reanalyzed (26 pairs) and when the exchange process was duplicated using separate aliquots of the same sample (18 pairs). For samples in which only one exchange with either of the labeled waters was repeated, the measure of variance of the calculated ␦18Osilica values was ⫾0.1‰. 3. RESULTS

The ␦18O values of phytoliths and meteoric water from across the transect are illustrated in Figure 2. In addition, our best estimations for the oxygen isotopic composition of meteoric water, growing-season average daily temperatures, and minimum and maximum relative humidities are listed in Table 1 for each locality. Weather data for both the July to August and May to August intervals have been included, because many researchers believe that the majority of phytoliths in grasses are deposited later in the growing season after the plant cells have matured and begun to senesce (Johnston et al., 1967; Simkiss and Wilbur, 1989). Shahack-Gross et al. (1996) have suggested that the late growing-season climatic conditions can have a dominant influence on the ␦18O values of phytoliths. The amount of silica in dry grass tissues varied from 0 to 8.4% (Table 2a). The average silica content was ⬃2 to 35 times higher in the sheaths, leaves, or inflorescences, than in the stems, rhizomes, or roots (Table 2a). On average, C. longifolia

Climatic influences on the ␦18O values of grass phytoliths

1893

Fig. 1. Sample locations across North America. 1) Pinery Provincial Park, ON; 2) Colorado Springs, CO; 3) Longwoods Conservation Area, Havanna, IL; 4) Kellogg, MN; 5) Aroya, CO; 6) Quinter, KS; 7) Thedford, NE; 8) Bonner, NE; 9) Fertile, MN; 10) Cheyenne, WY; 11) Agriculture Canada Station, Onefour, AB; 12) Dundurn, SA; 13) Sprucewoods Provincial Park, MA; 14) Rawlins, WY; 15) Broadus, MT; 16) University of Alberta, Kinsella Ranch, AB; 17) Kortes Dam, WY.

contained 0.9% silica by weight (n ⫽ 20), but the average silica content was noticeably higher in arid than relatively humid locations (1.2%, n ⫽ 9 versus 0.7%, n ⫽ 11). Overall, more than 70% of the total silica in C. longifolia is contained in the leaves and sheaths (Table 2b). The ␦18O values of the silica phytoliths and their percentage of exchangeable oxygen are summarized in Table 3 for various tissues of C. longifolia. Variations in the morphology and surface area of phytoliths from different tissues have the potential to affect the amount of exchangeable oxygen in the opal-A. In general, we observed no significant difference in the amount of exchangeable oxygen among the silica fractions from different tissues (paired t-tests: p ⫽ 0.05). However, the percentage of exchangeable oxygen in sheath phytoliths is lower than in the inflorescence or rhizome (paired t-tests: p ⫽ 0.005 and p ⫽ 0.003, respectively), suggesting that the former comprise less surface area than the latter. This anomaly, however, is unlikely to affect the measured oxygen-isotope composition of the silica. For all samples analyzed, an average of 4.6 ⫾ 1.7% of phytolith oxygen was susceptible to exchange. We expect that this value is typical of most grass species and note that it is similar to that found for silica diatoms (Knauth and Epstein, 1982; Juillet-Leclerc and Labeyrie, 1987).

4. DISCUSSION

In an equilibrium system, the oxygen-isotope composition of phytoliths is ultimately dependent on plant-water ␦18O values and temperature throughout the period of silica formation. However, other parameters such as the distribution of silica within various tissues, the effect of relative humidity on transpiration and plant-water ␦18O values, and the manner in which the original isotopic composition of meteoric water and subsequent soil-water ␦18O values vary with climate must also be considered. How these factors collectively influence the ␦18O values of phytoliths will largely shape the quality of climatic information that potentially can be obtained from ancient phytolith assemblages. 4.1. Distribution and Concentration of Silica in Plant Tissues As water evaporates from the plant during transpiration, the remaining water becomes enriched in solutes, including silicic acid. Once the concentration of silicic acid exceeds saturation, or pH conditions become favourable, silica preferentially precipitates where the majority of water has been removed (Simkiss and Wilbur, 1989). As a result, the highest phytolith

1894

E. A. Webb and F. J. Longstaffe

Fig. 2. The ␦18O values of local meteoric water and phytoliths extracted from C. longifolia at different sites. The soil-phytolith ␦18O values have been calculated (see text). Data from Tables 1 and 3.

concentrations occur in the leaves and sheaths (Table 2a, Fig. 3). Increased transpiration rates under arid conditions should result in elevated rates of silica precipitation. Johnston et al. (1967) found no correlation between the abundance of silica deposited in grasses and climatic parameters. In this study, however, a weak negative correlation was obtained between the total silica content of the plant and relative humidity (Fig. 4). The average silica content of transpiring tissues (leaf, sheath, and inflorescence) is much higher at sites that are relatively arid than in localities with a higher relative humidity (wt.% silica ⫽ 2.7 to 4.8% versus 0.7 to 2.5%; Table 2a, Fig. 3). Even nontranspiring tissues (stem, rhizome) are richer in silica in arid regions (0.4 to 1.0%) than in more humid areas (0.2 to 0.6%; Table 2a, Fig. 3). In regions of greater aridity, an increased movement of silicic acid–rich root water through the plant is required to replenish water lost during transpiration, producing greater silica deposition even in the nontranspiring tissues. Overall, the average silica content of C. longifolia is twice as high in arid regions than humid regions (Table 2a). The relative contribution of silica from different plant tissues to the overall phytolith assemblage, however, is similar, regardless of growing conditions (Table 2b). The availability of silicic acid in the soil may also affect the rate of phytolith production. The silicic acid concentration in soil water is a function of the availability and solubility of soil minerals, including amorphous silica, as well as the residence time and pH of soil water (Bartoli and Wilding, 1980; Raven, 1983; Rosen and Weiner, 1994; So¨ jberg, 1996; Alexandre et al., 1997). The production of silicic acid in soil water from the

dissolution of silica phytoliths is an important consideration for future research on the silica content of plants and the stability of fossil soil-phytolith assemblages. The pattern of silica deposition within C. longifolia is indicative of passive uptake and transport of silicic acid along the transpiration stream (Jones and Handreck, 1965). This suggests that there is no significant metabolic oxygen-isotope fractionation associated with the formation of phytoliths. However, systematic variations in ␦18O values do exist for silica formed in different tissues of the same plant, which are differentially affected by climatic variables such as relative humidity and temperature. 4.2. Consequences of Transpiration for Phytolith ␦18O Values Tissues affected by transpiration (i.e., leaves, inflorescence) contain partially evaporated water that is enriched in 18O relative to soil water, the extent of which is largely controlled by relative humidity. Leaf-water 18O enrichment can be calculated from steady-state models, increasing with transpiration rate and aridity and fluctuating on a daily and seasonal basis (Flanagan and Ehleringer, 1991; Farquhar and Lloyd, 1993; Buhay et al., 1996). Previously, we have demonstrated that silica in the leaves of C. longifolia is not precipitated from the highly 18 O-enriched leaf-water predicted from steady-state models (Webb and Longstaffe, 2000). Instead, leaf silica forms from water with lower ␦18O values, similar to average daily, bulk- or lower-leaf water (Shahack-Gross et al., 1996; Webb and Longstaffe, 2000). By comparison, water in nontranspiring tissues

Climatic influences on the ␦18O values of grass phytoliths

1895

Table 2a. Silica content of C. longifolia. A. Humid sites (max/min relative humidity ⬎90/50). B. Arid sites (max/min relative humidity ⬍90/50). C. Average values for all sample locations. % weight silicaa Location City, Province (State), Country

Inflorescences

Leaves

Sheaths

Stems

Rhizomes

Roots

1.9 0.1 0.3 — 0.4 0.1 — 0.1 0.8 2.0 —

3.3 1.1 1.2 2.5 1.2 1.0 7.0 2.6 3.5 — 1.9

2.4 1.4 0.6 2.2 1.5 1.6 5.6 1.3 — 3.0 2.3

0.3 0.6 0.2 0.6 0.2 0.2 1.5 0.7 1.0 — 0.4

0.5 0.2 0.0 0.1 — — — 0.2 0.3 — 0.0

0.1 — — — — — 0.2 — 0.3 — 0.0

0.7

2.5

2.2

0.6

0.2

0.2

1.7 1.3 3.0 1.9 3.7 2.8 2.7 2.1 5.0

5.9 2.6 4.2 2.8 — 6.7 2.3 5.6 5.6

3.1 4.0 4.4 3.7 8.4 4.4 — 6.1 4.1

0.5 0.7 0.6 — 1.4 0.4 0.4 3.2 0.4

0.4 0.4 0.5 0.3 0.6 0.4 0.4 0.8 0.1

0.1 — 0.1 0.0 0.3 0.1 0.1 0.0 —

Arithmetic average

2.7

4.5

4.8

1.0

0.4

0.1

C. All locations Arithmetic average

1.8

3.4

3.3

0.7

0.3

0.1

A. Humid locations 1. Pinery, ON, CAN

3. 4. 6. 9. 12. 13. 16.

Year 1995 early 1996 late 1996 1997 1996 1995 1996 1995 1995 1995 1995

Havanna, IL, USA Kellogg, MN, USA Quinter, KS, USA Fertile, MN, USA Dundurn, SA, CAN Sprucewoods, MA, CAN Kinsella, AB, CAN

Arithmetic average B. Arid locations 2. Colorado Springs, CO, USA 5. Aroya, CO, USA 7. Thedford, NE, USA 8. Bonner, NE, USA 10. Cheyenne, WY, USA 11. Onefour, AB, CAN 14. Rawlins, WY, USA 15. Broadus, MT, USA 17. Kortes Dam, WY, USA

a

1996 1996 1995 1995 1995 1995 1995 1995 1995

Relative to dry weight of grass.

(stems and rhizomes) is not influenced directly by relative humidity. Because there is no isotopic fractionation of soilwater during uptake through roots, these tissues contain water with an isotopic composition similar to the soil water, regardless of daily or seasonal changes in relative humidity (White et al., 1985; Flanagan and Ehleringer, 1991). If all above-ground plant tissues at the same location grow at the same temperature, variations in the ␦18O values of their silica must be related to the water from which the phytoliths precipitated. There is no significant difference between the ␦18O values of inflorescence and leaf silica analyzed in this study (Table 3; paired t-test, p ⫽ 1.0); water in both of these tissues must have experienced similar amounts of transpirationrelated 18O-enrichment over the life of these plants.

Sheath tissues have stomata and contain high concentrations of silica, suggesting that the sheaths are also active in transpiration. However, there is no significant difference in ␦18O values between sheath phytoliths and those formed in entirely nontranspiring tissues (Table 3; paired t-tests for sheath-stem, p ⫽ 0.06; stem-rhizome, p ⫽ 0.5; sheath-rhizome, p ⫽ 0.6). Sheath water represents the beginning of a successive string of water pools, which are progressively enriched in 18O from the base to the tip of the leaf (Wang and Yakir, 1995). Therefore, sheath-water ␦18O values are most similar to stem water. In addition, when transpiration rates are high, the rate of water movement toward the leaf increases. This influx of nonfractionated stem water through the sheath counteracts back diffusion of 18O-enriched water from sites of evaporation

Table 2b. Relative contribution (%) of silica from plant tissues to the overall silica content of C. longifolia. % silicaa Location

Inflorescences

Leaves

Sheaths

Stems

Rhizomes

Roots

A. Humid locations B. Arid locations C. All locations

2.4 4.7 4.1

38.5 35.6 36.4

33.4 38.1 35.8

7.9 6.9 7.2

9.9 11.8 12.0

8.0 2.7 4.4

a Avg. silica content (Table 2a) ⫻ avg. percentage of total biomass (inflorescence ⫽ 0.022, leaves ⫽ 0.10, sheaths ⫽ 0.10, stem ⫽ 0.091, rhizomes ⫽ 0.344, and roots ⫽ 0.344), as estimated from personal observation and Maun (1985).

1896

E. A. Webb and F. J. Longstaffe Table 3. O-isotope results (‰, VSMOW) and % exchangeable oxygen of silica phytoliths from C. longifolia.

Location City, Province (State), Country 1. Pinery, ON, CAN

1995 early summer 1996 late summer 1996 1997 Colorado Springs, CO, USA 1996 Havanna, IL, USA 1996 Kellogg, MN, USA 1995 Aroya, CO, USA 1996 Quinter, KS, USA 1996 Thedford, NE, USA 1995 Bonner, NE, USA 1995 Fertile, MN, USA 1995 Cheyenne, WY, USA 1995 Onefour, AB, CAN 1995 Dundurn, SA, CAN 1995 Sprucewoods, MA, CAN 1995 Rawlins, WY, USA 1995 Broadus, MT, USA 1995 Kinsella, AB, CAN 1995 Kortes Dam, WY, USA 1995

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. a

Year

Inflorescence

Leaf

Sheath

Stem

Rhizome

␦18O %ex.a

␦18O %ex.a

␦18O %ex.a

␦18O %ex.a

␦18O %ex.a

31.8

6.5

29.8 32.8 30.6 30.0

7.0 2.4 3.9 8.6

28.1 27.4 26.4 27.5 27.5 27.3 27.2 27.0 26.3 26.1

6.8 5.3 4.4 3.2 4.9 2.3 4.6 5.6 4.9 5.6

5.0

3.7 6.2

26.6 27.4

5.7 5.2

37.2 32.5 29.0 30.6 37.1 34.7

4.9 7.6 8.7 6.9 2.6 5.9

5.1 2.6 5.0 3.7 3.9 3.1 1.0 3.5 5.5 4.4 3.2 3.6 5.4 5.1 4.9 4.2

25.5

32.5 36.0

27.7 27.5 26.7 27.2 26.8 29.0 26.6 27.4 27.5 26.8 25.3 25.4 24.8 24.5 25.1 25.6

6.1 5.5 5.3 4.8 5.7

4.7

3.8 4.4 4.7 5.7 2.4 2.8 2.2 4.0 3.7 9.5 2.6 1.9 5.9 4.4 3.4

27.5 28.5 29.1 28.1 24.2

35.3

32.4 28.8 28.7 31.7 33.2 32.3 29.7 35.1 29.3 33.3 33.4 28.6 36.3 32.7 29.3

4.6 6.9 4.1

4.0 1.6 5.1

1.5 4.3

7.1

22.6 24.3 26.6

24.7 22.8

38.1

4.8 5.9 4.1 4.8

1.6 3.2 7.0 4.9 5.2 2.3 3.0 5.5 3.5

25.3 24.8 23.9

36.9 33.1 28.2 38.9

25.7 24.8 24.2 23.9 23.7 23.6 23.3 23.0 22.5

25.0

7.1

Soilphytolithb

Root

␦18O

% ex.a

22.8

2.1

23.4

6.3

22.9

3.7

␦18O 29.5 28.1 27.9 29.2 28.9 29.8 28.0 30.2 27.9 29.3 29.0 26.7 29.5 27.8 26.5 27.1 29.1 27.0 25.5 30.9

Average amount of oxygen affected by exchange. Calculated from Eq. 10; see text.

b

(Farquhar and Lloyd, 1993). Hence, we regard the sheaths as weakly transpiring and recognize their affinity with nontranspiring rhizomes and stems in terms of their ␦18Osilica behaviour. We previously concluded that leaf silica is deposited in equilibrium with mildly 18O-enriched leaf water, with the enrichment between leaf and stem silica being equal to that between average-daily bulk leaf water and soil water (Webb and Longstaffe, 2000):

␦stem silica ⫺ ␦soil water ⫽ ␦leaf silica ⫺ ␦leaf water 18

(4)

Theoretically, leaf water will not become enriched in O relative to stem water at 100% relative humidity, when transpiration rates are minimal, and phytoliths will have the same isotopic composition in both leaves and stems. The 18O-enrichment of leaf silica relative to stem silica is plotted against the reciprocal of daily average relative humidity in Figure 5. The correlation using the July-August, daily average relative humidity (h):

Fig. 3. Silica content (dry weight %) of tissues of C. longifolia collected from humid and arid locations across North America. Data from Table 2a.

Climatic influences on the ␦18O values of grass phytoliths

1897

Fig. 4. Silica content of whole grass samples (dry weight %) versus average daily relative humidity (%) for May–August and July–August periods. The silica content of whole grass samples was determined by multiplying the dry wt.% silica content of each tissue by the contribution of that tissue to the total biomass (Table 2). In cases where the silica content was unavailable for specific tissues, the average silica concentration for that tissue was substituted.

⌬18Oleaf water-soil water ⫽ ⌬18Oleaf silica-stem silica ⫽ (12.5/h) ⫺13; (R2 ⫽ 0.9, p ⫽3.7 ⫻ 10⫺7)

(5)

is very similar to that reported by Yapp (1979) for the evaporation of body fluids in land snails, whose shells were used in paleoclimate studies. Bombin and Muehlenbachs (1980) used the relationship formulated by Yapp (1979) to calculate the ␦18O values of leaf water involved in leaf-silica formation. If this relationship is extrapolated to 100% relative humidity (e.g., h ⫽ 1), the value of ⌬18Oleaf silica-stem silica tends toward zero, as predicted above. The correlation obtained using average relative humidity values for the entire growing season (May–August) is poorer than for the July–August period, and produces an unrealistic value of ⌬18Oleaf silica-stem silica at 100% relative humidity (Fig. 5). It appears that relative humidity during the later part of the growing season is more accurately reflected in the ␦18O values of these phytoliths. Because the difference in ␦18O values between stem and leaf phytoliths can be related directly to relative humidity, it appears that neither metabolic processes nor the rate of phytolith precipitation in the leaves versus the stems played a significant role in determining their oxygen isotopic compositions.

4.3. Control of Temperature and Soil-Water on Phytolith ␦18O Values Differences in ⌬18Osilica-soil water values for various nontranspiring tissues (stems, rhizomes, and roots) within a single plant ought to be the result of differences in temperature. For example, the low thermal conductivity of sand insulates underground tissues from both high surface temperatures during the day and rapid cooling at night (Baldwin and Maun, 1983). Rhizome silica formed at lower average temperatures than coexisting stem silica would have a higher ␦18O value, as was observed for many C. longifolia samples (Table 3). Differences between rhizome- and stem-silica ␦18O values can also result from age differences of various tissues in this perennial species. Silica in the aboveground tissues, which regenerate annually, will record only the most recent growing-season temperatures, weighted towards late growing-season conditions. By comparison, rhizome-silica ␦18O values will integrate fluctuations in temperature and soil-water ␦18O values over several years of growth (up to 5 yr at the depth sampled for this study). The ␦18O values of silica within individual tissues should record a weighted average of diurnal and seasonal temperature variations. However, the weak positive correlation between ␦18O values of stem silica and May to August temperatures,

1898

E. A. Webb and F. J. Longstaffe

Fig. 5. Relationship between values of ⌬18Oleaf silica-stem silica and average daily relative humidity (h) for the May to August [⌬18Oleaf silica-stem silica ⫽ (21.3/h) ⫺ 26.0, R2 ⫽ 0.6, p ⫽ 6.4 ⫻ 10⫺5] and July to August [⌬18Oleaf silica-stem silica ⫽ (12.5/h) ⫺ 13.0, R2 ⫽ 0.9, p ⫽ 1.4 ⫻ 10⫺7] growing periods. h is given in fractional percent (e.g., 100% ⫽ 1).

␦ 18Ostem silica ⫽ 0.52 t ⫹ 16.1; (R2 ⫽ 0.5; p ⫽ 0.001), t in °C, (6) is opposite to that expected from temperature-dependent equilibrium isotope-fractionation between phytoliths and plant water (Fig. 6A). This relationship is instead dominated by the generally higher ␦18Osoil water values at warmer sites within the North American study area. The change in isotopic composition of precipitation across continental areas can be correlated with annual or seasonal surface temperatures, such that precipitation is enriched in 18O and D at locations with warmer climates (Dansgaard, 1964; Rozanski et al., 1993; Fricke and O’Neil, 1999). The isotopic composition of soil water can vary from local precipitation as a result of selective infiltration, mixing in the unsaturated zone, and evaporative enrichment. However, its ␦-values over 10 to 100 cm depth generally describe a reservoir accumulated over an entire season and hence correlate well with seasonal precipitation and temperature (Gage, 2001). Consequently, the correlation between silica ␦18O values and average temperatures for the May–August period is stronger than the same relationship for July–August temperatures, notwithstanding that most phytoliths are formed later in the growing season. Because transpiration and photosynthesis are constrained in part by temperature, the growth rate and phytolith productivity of C. longifolia may be optimized within a relatively narrow temperature range, regardless of climatic variation at a given site. Temperature dependency can produce a temporal separation in the primary production of different grass species at one

location. In addition, southern ecotypes of some species may mature later in the season than their northern counterparts (Madakadze et al., 1998). As the most productive period for this species may vary from site to site and from year to year, average growing-season temperatures may not reflect those at which most silica precipitated. Such behaviour further restricts the role of temperature in determining the isotopic composition of phytoliths. Recent studies have shown that the ␦18O values of tropical, fresh-water, silica diatoms are also dominated by the oxygen-isotope composition of precipitation rather than growth temperature (Barker et al., 2001). The ␦18O values of soil water from which stem phytoliths precipitated can be calculated using a constant “ideal” growingtemperature (e.g., 20°C) or measured growing-season (May– August) temperatures and the paleothermometer of ShahackGross et al. (1996): t ⫽ 5.8 ⫺ 2.8 (⌬18Ostem silica-soil water ⫺ 40), t in °C.

(7)

The soil-water ␦ O values calculated using a constant growing-temperature vary with measured growing-season temperatures as follows (Fig. 6B, Table 1): 18

␦ 18Osoil water ⫽ 0.52 t ⫺ 18.8; (R2 ⫽ 0.5, p ⫽ 0.001), t in °C (8) If the phytoliths formed at temperatures better represented by average values for the growing season (May to August) at each location, the relationship between ␦18Osoil water values and temperature becomes (Fig. 6B, Table 1):

Climatic influences on the ␦18O values of grass phytoliths

1899

Fig. 6. Relationship between ␦18O values of stem silica or soil water and average daily growing temperatures for: (A) C. longifolia stem silica and May to August and July to August temperatures; and (B) soil-water ␦18O values calculated using Eqn. 7 and May to August temperatures or a constant temperature (t ⫽ 20°C) versus the average May to August temperatures.

␦ 18Osoil water ⫽ 0.87 t ⫺ 25.9; ⫺6

(R ⫽ 0.7, p ⫽ 5.6 ⫻ 10 ), t in °C 2

(9)

The covariation between ␦18Osoil water values and temperature (0.52 or 0.87‰ per °C) is larger than reported by Fricke and O’Neil (1999) for summer precipitation (0.42‰ per °C), and brackets the values for annual precipitation (0.58 to 0.69‰ per °C) proposed by Dansgaard (1964) and Rozanski et al. (1993) for midlatitude continental regions. Such behaviour is consistent with increasing evaporative enrichment of soil-water under more arid conditions, which are commonly associated with higher temperatures. The actual variation in ␦18Osoil water values versus temperature across North America likely lies somewhere between the values obtained from these two interpretations. Regardless, either scenario shows how temperature-dependent variations of soil-water ␦18O values across the grasslands of North America can sufficiently account for the positive relationship observed between stem silica and temperature (Fig. 6A).

4.4. Surface Water, Ground Water, and Summer Precipitation as Proxies for Soil-Water ␦18O Values Because of temperature effects, summer rain is normally enriched in 18O relative to most winter precipitation at a given location. The shallow rooting systems of grasses preferentially utilize summer precipitation, which also prevents some of this water from reaching the water table (Gupta, 1979; Darling and Bath, 1988; Berndtsson et al., 1996). Some winter precipitation also remains in the soil, and is available for plant uptake (Maule´ et al., 1994). In addition, the ␦18O values of precipitation can be altered in the soil-water reservoir by daily cycles of evaporation and condensation (Allison et al., 1984; Walker and Brunel, 1990). Hence, accurate estimation of soil-water ␦18O values is a major challenge, particularly for ancient phytolith-bearing soils where direct measurement is not possible. In our search for suitable proxies, we examine next whether the ␦18O values of surface water, ground water, or summer precipitation can be used to approximate the oxygen-isotope composition of soil water across our study area. Figure 7 illustrates the variation in

1900

E. A. Webb and F. J. Longstaffe

Fig. 7. Correlations between growing-season temperatures (t in °C; May to August) and ⌬18Ostem silica-water for: (A) surface water, (B) ground water, and (C) summer precipitation. The dashed line represents the temperature-dependent equilibrium fractionation between phytoliths and water defined by Eqn. 7. Equations for the lines of best linear fit are: t ⫽ 15.3 ⫺ 0.68 (⌬18Ostem silica-surface water ⫺ 40), (R2 ⫽ 0.5, p ⫽ 0.0008); t ⫽ 17.7 ⫺ 0.53 (⌬18Ostem silica-ground water ⫺ 40), (R2 ⫽ 0.4, p ⫽ 0.004); t ⫽ 13.9 ⫺ 0.77 (⌬18Ostem silica-precipitation ⫺ 40), (R2 ⫽ 0.7, p ⫽ 0.0005).

our data between average growing-season temperature (May– August) and ⌬18Ostem silica-water for (a) surface water, (b) ground water, and (c) summer precipitation. Similar, although more poorly correlated, curves result if average temperatures for the July–August period are used instead. Wherever possible, the surface-water samples were collected from relatively small water bodies at the end of the growing season, which maximized the opportunity for local recharge by summer rains. The correlation between temperature and ⌬18Osilica-surface water has a much shallower slope than Eqn. 7 with the ⌬18Ostem silica-surface water values more or less bisected by the Shahack-Gross et al. (1996) geothermometer (Fig. 7A). The ⌬18Ostem silica-surface water values that plot to the right of this line must have been obtained using water compositions that

were depleted of 18O relative to soil water. Water transport by molecular diffusion through the stable boundary layer above the soil-water reservoir amplifies its evaporative 18O-enrichment via kinetic fractionation relative to an open body of water (Allison et al., 1983). As soil water is increasingly enriched in 18 O by evaporation, particularly in arid regions, estimates of ⌬18Ostem silica-soil water made using surface water become increasingly inaccurate. The ⌬18Ostem water-surface water values that plot to the left of Eqn. 7 in Figure 7A represent calculations in which soil-water ␦18O values were overestimated by summer surface-water compositions. This can occur if the soil-water reservoir contains a fraction of winter precipitation, which is generally depleted of 18O relative to most summer precipitation (Maule´ et al., 1994), or if small surface-water bodies experi-

Climatic influences on the ␦18O values of grass phytoliths

1901

Fig. 8. Correlation between the ␦18O values of silica from C. longifolia and the best estimates of local surface-water ␦18O values for stem silica, leaf silica, and predicted soil-phytolith assemblages. The solid line describes the linear correlation obtained between the ␦18O values of surface water and silica from stem tissues: ␦18Ostem silica ⫽ 0.41 ␦18Osurface water ⫹ 29.8; (R2 ⫽ 0.8, p ⫽ 1.7 ⫻ 10⫺7).

ence more evaporative 18O-enrichment than soil water. The natural variability in ␦18O values for different surface-water bodies at any one site also can limit the value of this proxy for growing-season soil-water compositions. Almost all ⌬18Ostem silica-water values calculated using ground-water compositions plot to the right of Eqn. 7 (Fig. 7B). Shallow ground water contains a large fraction of winter precipitation, and is largely unaffected by evaporation. Hence, ground water is generally depleted of 18O relative to soil water from which the phytoliths formed (Table 1), and is unsuitable as a soil-water proxy. The majority of summer-precipitation ␦18O values are higher than the average seasonal isotopic composition calculated for the soil water using Eqn. 7 (Fig. 7C). For example, the ␦18O values of soil water at Pinery Provincial Park in 1997 predicted using Eqn. 7 (␦18Osoil water ⫽ ⫺8.5‰; Table 1) and measured at this location (␦18O soil water ⫽ ⫺8.6‰; weighted average over a depth of 1 m in August, 1997; Gage, 2001) are both less enriched in 18O than summer precipitation (␦18Osummer precipitation ⫽ ⫺6.1‰; Table 1). This observation tends to confirm that a lower-18O fraction, such as winter precipitation or spring runoff, is available to mix with summer precipitation in the soil before moisture is taken up by grass roots. The shallow soil-water ␦18O values measured near Kinsella, Alberta (␦18O ⫽ –14‰; Maule´ et al., 1994) are also very similar to our calculated soil-water composition (␦18Osoil water ⫽ ⫺14.0‰, Table 1). In this case, however, the effects of evaporative 18O-enrichment have caused the predicted ␦18Osoil-water

values to be higher than summer-precipitation values (␦18Osummer ⫽ ⫺15.6‰). The oxygen-isotope composition of soil water entering the grass is a composite of evaporated water near the surface and a mixture of winter and summer precipitation stored in the soil-water reservoir at greater depths. The ultimate balance among these contributions depends not only on the distribution of soil water through the profile, but also on the (not unrelated) distribution of active root and rhizome biomass. precipitation

4.5. Potential of Soil-phytolith ␦18O Values for Paleoclimatic Reconstruction Extracting quantitative ␦18Osoil water, relative humidity, and temperature signals from soil-phytolith assemblages for paleoclimatic reconstruction is complicated by the combination of phytoliths from different plants and the mixing of phytoliths from transpiring and nontranspiring tissues. In addition, soil phytoliths accumulated over tens to thousands of years will have formed under variable diurnal, seasonal, and annual climate conditions. For example, climatic variations over a 4-yr period at Pinery Provincial Park, Ontario, caused the oxygenisotope compositions of C. longifolia stem silica to vary by 1.7‰ (Table 3). Physical disturbances arising from grazing, fire, aeolian transport, bioturbation, and downward soil drainage can also disrupt the initial deposition of soil phytoliths such that they no longer represent the overlying vegetation exclusively (Fredlund and Tieszen, 1994; Alexandre, 1996). Likewise, phytolith dissolution rates will vary with the size and

1902

E. A. Webb and F. J. Longstaffe

Fig. 9. Linear correlation between growing-season temperatures (t in °C; May to August) and the difference between the calculated ␦18O values of soil-phytolith assemblages and (A) surface water, and (B) soil water calculated using the May–August temperatures. The dashed line represents the temperature-dependent equilibrium fractionation between phytoliths and water defined by Eqn. 7. The lines of best linear fit through the data are: t ⫽ 17.6 ⫺ 0.42 (⌬18Osoil 2 18 2 phytolith-surface water ⫺ 40), (R ⫽ 0.5, p ⫽ 0.001); and t ⫽ 17.0 ⫺ 0.74 (⌬ Osoil phytolith-soil water ⫺ 40), (R ⫽ 0.5, p ⫽ 0.0006).

shape of the silica bodies, and soil-water chemistry and pH. Nevertheless, soil-phytolith assemblages are generally representative of native vegetation on an extra-local to regional scale in the North American prairies (Fredlund and Tieszen, 1994). Here, we shall assume that phytoliths from all tissues are transferred directly to the underlying soil on an annual basis and that they are equally well preserved. For such a scenario, Table 2b shows that the underground portions (roots and rhizomes) of C. longifolia contribute only ⬃16% of the total phytolith assemblage, which serves to minimize contamination of ancient phytoliths by modern root and rhizome phytoliths. Overall, 59% of the total phytolith assemblage will be produced by essentially nontranspiring tissues (sheath, stem, rhizome, and root). This amount, however, will be reduced if rhizome and root silica is not returned to the soil annually. Strongly transpiring tissues provide ⬃41% of the phytoliths. Table 3 contains the ␦18O values of this hypothetical soilphytolith assemblage at each site, calculated using the average silica concentrations of each tissue (Table 2b) and the measured ␦18O values of its phytoliths (Table 3):

␦ 18Osoil phytolith ⫽ 0.041 ␦18Oinflorescence ⫹ 0.364 ␦18Oleaves ⫹ 0.358 ␦18Osheath ⫹ 0.072 ␦18Ostem ⫹ 0.164 ␦18Orhizomes.

(10)

For plant parts whose ␦18Osilica values were not available, results for a comparable tissue were substituted (e.g., leaves for inflorescences). The ␦18Orhizome values were substituted for ␦18Oroot values in all cases. Stem-phytolith and surface-water ␦18O values show a systematic covariance, as both parameters increase with increasing temperature and decreasing latitude (Fig. 8; an identical relationship exists when summer-precipitation ␦18O values are used). The slope (m ⫽ 0.41) of this relationship is consistent with the temperature dependence of the ␦18O values of both annual precipitation (␦18Owater ⫽ 0.69 t ⫹ C; Dansgaard, 1964) and phytolith silica (Eqn. 7; Shahack-Gross et al., 1996), which combine to form a hypothetical ␦18Osilica versus ␦18Owater curve with a slope of 0.48. In contrast, the ␦18O values of leaf silica do not vary systematically with temperature, surface-water ␦18O values, or latitudinal gradients (Figs. 2 and 8). At several locations, the 18O-enrichment of leaf and inflorescence silica arising

Climatic influences on the ␦18O values of grass phytoliths

from increased transpiration rates at the arid-western boundary of the prairies (Fig. 1; sites 2, 5, 10, 14, 15, and 17) is counterbalanced by the lower ␦18O values of silica from nontranspiring tissues, whose compositions directly reflect the lower ␦18O values of surface water in these localities (Figs. 2 and 8). The result is that the calculated ␦18Osoil 18 phytoliths values fall within a relatively narrow range ( ␦ O ⫽ ⫹25.5 to ⫹30.9‰) regardless of the diverse climatic regimes in which the grasses have grown. A modest empirical relationship exists between temperature and both ⌬18Osoil phytolith-surface water (Fig. 9A) and ⌬18Osoil phytolith-soil water values (Fig. 9B). This may have some value in the estimation of growing-season temperatures for soil-phytolith assemblages, provided that ␦18O values can be obtained for concurrent soil or surface waters. The equilibrium relationship between the ␦18O values of plant silica and water implies that, for locations where soilwater ␦18O values remain relatively constant, a decrease in temperature will result in soil-phytolith assemblages that are more enriched in 18O. However, an increase in the ␦18O values of soil-phytolith assemblages over time can also be caused by greater evaporative enrichment in 18O of soil water or by increased transpiration. 5. CONCLUSIONS

The oxygen-isotope compositions of silica phytoliths from C. longifolia record valuable climatic data, including information about the ␦18O values of soil water. Non- or weakly transpiring parts of this grass (sheaths, stems, and rhizomes) precipitate silica in oxygen isotopic equilibrium with plant water, which in turn, is unfractionated in oxygen isotopes from soil water. Silica deposited in the leaves and inflorescence is formed from plant water that has been enriched in 18O during transpiration. The amount of this enrichment is dependent on the transpiration rate, which can be directly related to relative humidity. The variation of ␦18Osilica values among C. longifolia grown under a range of natural conditions is highly dependent on soil-water ␦18O values. Accurate calculation of the temperature of silica precipitation is only possible when plant-water ␦18O values for that tissue are known. For nontranspiring tissues, this ␦18O value corresponds to soil water at the average rooting depth. The ␦18O values of stem phytoliths can be related to the variation of ␦18Osoil water values over a continental temperature gradient. Soil-phytolith assemblages are composed of silica formed in both transpiring and nontranspiring tissues. Temperature and soil-water ␦18O signals carried by phytoliths from the stems, sheaths, and rhizomes may be masked by the ␦18O values of phytoliths from the leaves and inflorescence. These are variably enriched in 18O relative to phytoliths from nontranspiring tissues, depending on relative humidity. Accurate reconstruction of temperature, ␦18Osoil water values, and relative humidity using ancient phytolith assemblages will require the recognition (via morphology) and isolation (via handpicking) of phytoliths produced in transpiring versus nontranspiring tissues. Acknowledgments—We thank Paul Middlestead, Sharon Forbes, Dagmar Lacina, Shelly Morgan, Jamie Longstaffe, Raveenie Ratnasingam, and Mark Fazari for their assistance in the stable isotope laboratory.

1903

Grass samples were obtained with the help of many individuals and Onefour Agricultural Canada Research Station, University of Alberta’s Kinsella Ranch, Kaste Inc., Fertile, Minnesota, Illinois Nature Preserves Commission, Montana’s Natural Heritage Program, Manitoba Natural Resources, and Pinery Provincial Park, Ontario. Henry Schwarcz provided useful suggestions regarding an earlier version of this manuscript. We also thank Ronald K. Matheney, an anonymous reviewer and Associate Editor Simon Sheppard for their helpful reviews. This research was supported by the Natural Sciences and Engineering Research Council of Canada (Grant A7387 to F.J.L.). Associate editor: S. M. F. Sheppard REFERENCES Alexandre A., Meunier J-D., Colin F., and Koud J. M. (1997) Plant impact on the biogeochemical cycle of silicon and related weathering processes. Geochim. Cosmochim. Acta 61, 677– 682. Alexandre A. (1996) Phytolith interactions sol-plante et paleoenvironnements. Ph.D. thesis, Univ. d’Aix-Marseille III. Allison G. B., Barnes C. J., and Hughes M. W. (1983) The distribution of deuterium and 18O in dry soils 2. Experimental J. Hydrol. 64, 377–397. Allison G. B., Barnes C. J., Hughes M. W., and Leaney F. W. J. (1984) Effect of climate and vegetation on oxygen-18 and deuterium profiles in soils. In Isotope Hydrology (IAEA SM-270/20), pp. 105–124. International Atomic Energy Agency, Vienna. Back W., Hanshaw B., Plummer N., Rahn P., Rightmire C., and Rubin M. (1983) Process and rate of dedolomitization: Mass transfer and 14 C dating in a regional carbonate aquifer. Geol. Soc. Am. Bull. 94, 1415–1429. Baldwin K. A. and Maun M. A. (1983) Microenvironment of the Lake Huron sand dunes. Can. J. Bot. 61, 241–255. Barker P. A., Street-Perrott F. A., Leng M. J., Greenwood P. B., Swain D. L., Perrott R. A., Telford R. J., and Ficken K. J. (2001) A 14,000-year oxygen isotope record from diatom silica in two alpine lakes on Mt. Kenya. Science 292, 2307–2310. Bartoli F. and Wilding P. (1980) Dissolution of biogenic opal as a function of its physical and chemical properties. Soil Sci. Soc. Am. J. 44, 873– 878. Berndtsson R., Nodomi K., Yasuda H., Persson T., Chen H., and Jinno K. (1996) Soil water and temperature patterns in an arid desert dune sand. J. Hydrol. 185, 221–240. Bombin M. and Muehlenbachs K. (1980) Potential of 18O/16O ratios in opaline plant silica as a continental paleoclimatic tool. Am. Quat. Ass. Progr. Abstr. 6, 43– 44 (abstr.). Buhay W. M., Edwards T. W. D., and Aravena R. (1996) Evaluating kinetic fractionation factors used for ecological and paleoclimatic reconstructions from oxygen and hydrogen isotope ratios in plant water and cellulose. Geochim. Cosmochim. Acta 60, 2209 –2218. Clarke J. F., Davisson M. L., Hudson G. B., and Macfarlane P. A. (1998) Noble gases, stable isotopes and radiocarbon as tracers of flow in the Dakota Aquifer, Colorado and Kansas. J. Hydrol. 211, 151–167. Clayton R. N. and Mayeda T. K. (1963) The use of bromine pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis. Geochim. Cosmochim. Acta 27, 43–52. Coplen T. B. (1994) Reporting of stable hydrogen, carbon and oxygen isotopic abundances (technical report). Pure Appl. Chem. 66, 273– 276. Dansgaard W. (1964) Stable isotopes in precipitation. Tellus 16, 436 – 468. Darling W. G. and Bath A. H. (1988) A stable isotope study of recharge processes in the English Chalk. J. Hydrol. 101, 31– 46. Epstein S. and Mayeda T. K. (1953) Variation of 18O content of waters from natural sources. Geochim. Cosmochim. Acta 4, 213–224. Farquhar G. D. and Lloyd J. (1993) Carbon and oxygen isotope effects in the exchange of carbon dioxide between terrestrial plants and the atmosphere. In Stable Isotopes and Plant Carbon-Water Relations (eds. J. R. Ehleringer et al.), pp. 47–70. Academic Press. Flanagan L. B. and Ehleringer J. R. (1991) Stable isotope composition of stem and leaf water: Applications to the study of plant water use. Func. Ecol. 5, 270 –277.

1904

E. A. Webb and F. J. Longstaffe

Flanagan L. B., Bain J. F., and Ehleringer J. R. (1991) Stable oxygen and hydrogen isotope composition of leaf water in C3 and C4 plant species under field conditions. Oecologia 88, 394 – 400. Fortin G., vanderKamp G., and Cherry J. A. (1991) Hydrology and hydrochemistry of an aquifer-aquitard system within glacial deposits, Saskatchewan, Canada. J. Hydrol. 126, 265–292. Fredlund G. G. (1993) Paleoenvironmental interpretations of stable carbon, hydrogen and oxygen isotopes from opal phytoliths, Eutis Ash Pit, Nebraska. In Current Research in Phytolith Analysis: Applications in Archaeology and Paleoecology (eds. D. M. Pearsal and D. R. Piperno), MASCA Research Papers in Science and Archaeology 10, 37– 46. Fredlund G. G. and Tieszen L. T. (1994) Modern phytolith assemblages from the North American Great Plains. J. Biogeog. 21, 321–335. Fricke H. C., Clyde W. C., and O’Neil J. R. (1998) Intra-tooth variations in ␦18O (PO)4 of mammalian tooth enamel as a record of seasonal variations in continental climate variables. Geochim. Cosmochim. Acta 62, 1839 –1850. Fricke H. C. and O’Neil J. R. (1999) The correlation between 18O/16O ratios of meteoric water and surface temperature: Its use in investigating terrestrial climate change over geological time. Earth Planet. Sci. Lett. 170, 181–196. Fritz P., Drimmie R. J., Frape S. K., and O’Shea F. K. (1987) The isotopic composition of precipitation and ground water in Canada. In Isotope Techniques in Water Resources Development. IAEA Proceedings Series, 539 –550. Gage K. L. (2001) The stable isotope geochemistry of the unsaturated zone: Pinery Provincial Park. M.Sc. thesis. Univ. Western Ontario. Gupta J. P. (1979) Some observations of the periodic variations of moisture in stabilized and unstabilized dunes of the Indian desert. J. Hydrol. 41, 153–156. Hendry M. J. and Schwartz F. W. (1988) An alternative view on the origin of chemical and isotopic patterns in ground water from the Milk River Aquifer, Canada. Water Resour. Res. 24, 1747–1763. Hunt R. J., Bullen T. D., Krabbenhoft D. P., and Kendall C. (1997) Using stable isotopes of water and strontium to investigate the hydrology of a natural and constructed wetland. Groundwater 36, 434 – 443. IAEA. (1992) Statistical Treatment of Data of Environmental Isotopes in Precipitation. International Atomic Energy Agency Technical Report Series 331, Vienna. 448 –511. Johnston A., Bezeau L. M., and Smoliak S. (1967) Variation in silica content of range grasses. Can. J. Plant Sci. 47, 65–71. Jones L. H. P. and Handreck K. H. (1965) Studies of silica in the oat plant. III. Uptake of silica from soils by the plant. Plant and Soil 24, 79 –96. Juillet-Leclerc A. and Labeyrie L. (1987) Temperature dependence of the oxygen isotopic fractionation between diatom silica and water. Earth Planet. Sci. Lett. 84, 69 –74. Knauth L. P. and Epstein S. (1982) The nature of water in hydrous silica. Am. Min. 67, 510 –520. LaBaugh J. W., Winter T. C., Rosenberry D. O., Schuster P. F., Reddy M. M., and Aiken G. R. (1997) Hydrological and chemical estimates of the water balance of a closed basin in north central Minnesota. Water Resour. Res. 33, 2799 –2812.

Luz B., Cormie A. B., and Schwarcz H. P. (1990) Oxygen isotope variations in phosphate of deer bones. Geochim. Cosmochim. Acta 54, 1723–1728. Madakadze I. C., Coulman B. E., Mcelroy A. R., Stewart K. A., and Smith D. L. (1998) Evaluation of selected warm-season grasses for biomass production in areas with a short growing season. Biosource Tech. 65, 1–12. Maule´ C. P., Chanasyk D. S., and Muehlenbachs K. (1994) Isotopic determination of snow-water contribution to soil water and ground water. J. Hydrol. 155, 73–91. Maun M. A. (1985) Population biology of Ammophila breviligulata and Calamovilfa longifolia of Lake Huron sand dunes. I. Habitat, growth form, reproduction and establishment. Can. J Bot. 63, 113– 124. Metcalfe C. R. (1960) Anatomy of the Monocotyledons I Gramineae. Claredon Press. Raven J. A. (1983) The transport and function of silicon in plants. Biol. Rev. Cambridge Phil. Soc. 58, 179 –207. Rosen A. M. and Weiner S. (1994) Identifying ancient irrigation: A new method using opaline phytoliths from emmer wheat. J. Arch. Sci. 21, 125–132. Rozanski K., Araguas-Araguas L., and Gonfiantini R. (1993) Isotopic patterns in modern global precipitation. In Climate Change in Continental Isotopic Records. Geophysical Monograph. 78, 1–36. Shahack-Gross R., Shemesh A., Yakir D., and Weiner S. (1996) Oxygen isotopic composition of opaline phytoliths: Potential for terrestrial climatic reconstruction. Geochim. Cosmochim. Acta 60, 3949 – 3953. Siegal D. I. (1989) Geochemistry of the Cambrian-Ordovician aquifer system in the northern midwest, United States. U.S. Geol. Sur. Prof. Paper, 1405-D. Simkiss K. and Wilbur K. M. (1989) Biomineralization Cell Biology and Mineral Deposition. Academic Press. So¨ jberg S. (1996) Silica in aqueous environments. J. Non-Crystalline Solids 196, 51–57. USGS. (1984) Geochemistry of ground water in two sandstone aquifer systems in the Northern Great Plains in parts of Montana, Wyoming, North Dakota and South Dakota. U.S. Geological Survey Professional Paper, 1402-C. Walker C. D. and Brunel J. P. (1990) Examining evapotranspiration in a semi-arid region using stable isotopes of hydrogen and oxygen. J. Hydrol. 118, 55–75. Wang X. F. and Yakir D. (1995) Temporal and spatial variations in the oxygen-18 content of leaf water in different plant species. Plant Cell Environ. 18, 1377–1385. Webb E. A. and Longstaffe F. J. (2000). The oxygen isotopic compositions of silica phytoliths and plant water in grasses: Implications for the study of paleoclimate. Geochim. Cosmochim. Acta 64, 767– 780. White J. W. C., Cook E. R., Lawrence J. R., and Broecker W. S. (1985) The D/H ratios of sap in trees; implications for water sources and tree ring D/H ratios. Geochim. Cosmochim. Acta 49, 237–246. Yapp C. J. (1979) Oxygen and carbon isotope measurements of land snail shell carbonate. Geochim. Cosmochim. Acta 43, 629 – 635.