Tree Physiology Advance Access originally published online on December 5, 2008
Tree Physiology 2009 29(2):157-169; doi:10.1093/treephys/tpn019
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Seasonal changes in depth of water uptake for encroaching trees Juniperus virginiana and Pinus ponderosa and two dominant C4 grasses in a semiarid grassland
1 School of Natural Resources, University of Nebraska–Lincoln, 406 Hardin Hall, Lincoln, NE 68583-0974, USA
2 Corresponding author (tawada2{at}unl.edu)
3 Department of Soil, Water and Climate, University of Minnesota, Minneapolis, MN 55108, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We used the natural abundance of stable isotopic ratios of hydrogen and oxygen in soil (0.05–3 m depth), plant xylem and precipitation to determine the seasonal changes in sources of soil water uptake by two native encroaching woody species (Pinus ponderosa P. & C. Lawson, Juniperus virginiana L.), and two C4 grasses (Schizachyrium scoparium (Michx.) Nash, Panicum virgatum L.), in the semiarid Sandhills grasslands of Nebraska. Grass species extracted most of their water from the upper soil profile (0.05–0.5 m). Soil water uptake from below 0.5 m depth increased under drought, but appeared to be minimal in relation to the total water use of these species. The grasses senesced in late August in response to drought conditions. In contrast to grasses, P. ponderosa and J. virginiana trees exhibited significant plasticity in sources of water uptake. In winter, tree species extracted a large fraction of their soil water from below 0.9 m depth. In spring when shallow soil water was available, tree species used water from the upper soil profile (0.05–0.5 m) and relied little on water from below 0.5 m depth. During the growing season (May–August) significant differences between the patterns of tree species water uptake emerged. Pinus ponderosa acquired a large fraction of its water from the 0.05–0.5 and 0.5–0.9 m soil profiles. Compared with P. ponderosa, J. virginiana acquired water from the 0.05–0.5 m profile during the early growing season but the amount extracted from this profile progressively declined between May and August and was mirrored by a progressive increase in the fraction taken up from 0.5–0.9 m depth, showing plasticity in tracking the general increase in soil water content within the 0.5–0.9 m profile, and being less responsive to growing season precipitation events. In September, soil water content declined to its minimum, and both tree species shifted soil water uptake to below 0.9 m. Tree transpiration rates (E) and water potentials (
) indicated that deep water sources did not maintain E which sharply declined in September, but played an important role in the recovery of tree . Differences in sources of water uptake among these species and their ecological implications on tree–grass dynamics and soil water in semiarid environments are discussed.
Keywords: Great Plains, Nebraska Sandhills, Panicum virgatum, Schizachyrium scoparium, stable isotopes, water potential
Received February 8, 2008; Accepted September 25, 2008
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The distribution of soil water over time and with depth is a key determinate of vegetation structure, including the balance of grasses and woody species in semiarid and arid landscapes (Huxman et al. 2005, Darrouzet-Nardi et al. 2006). At 50,000 km2, the Nebraska Sandhills are the largest stabilized sand dune formation in the Western Hemisphere. Although the Sandhills are almost entirely stabilized by grasses, tree encroachment into these semiarid grasslands is increasing. Ponderosa pine (Pinus ponderosa P. & C. Lawson), a major forest type in the USA (Sala et al. 2005), has expanded into grasslands from historic grassland–woodland ecotones (Steinauer and Bragg 1987), and eastern red cedar (Juniperus virginiana L.) is spreading throughout the Sandhills and the Great Plains (Schmidt and Stubbendieck 1993, Willson et al. 2008). These land cover changes reflect shifts from open grasslands and savannas to more closed-canopy woodlands in semiarid landscapes worldwide (Scholes and Archer 1997, Bond et al. 2005). Although decreased fire frequency is certainly a driver of these changes (Bond et al. 2005), over-grazing, climate change, elevated atmospheric CO2 concentration, atmospheric nitrogen deposition and human-enhanced dispersal may also favor woody species over native C4 grasses such as those that dominate the Sandhills (reviewed by Scholes and Archer 1997). Regardless of the mechanisms behind woody plant expansion into semiarid grasslands, this transition has potentially large biogeochemical, hydrologic and ecologic impacts (Archer et al. 2001, Jackson et al. 2002, Huxman et al. 2005).
Root distributions and soil water dynamics play a key role in an ecosystem’s water balance, as evidenced by the emphasis placed in recent ecohydrology theory and modeling (e.g., Huxman et al. 2005, Rodríguez-Iturbe and Porporato 2005). Ecohydrology models make assumptions about the rooting zone, i.e., the depth to which roots effectively deplete soil water. For native grasslands, various modelers have used rooting zones ranging from 0.3 to 1.5 m depth. The classic studies of Weaver (1958 and references therein) remain the ‘benchmark for studies of root morphology and distribution’ in the Great Plains (Schenk and Jackson 2002). For the Sandhills, Weaver (1958) concluded that the ‘Sandhills grasses are characterized by moderately fine roots which spread widely in the surface, usually 1.5–3.0 ft (0.45–0.91 m), and penetrate deeply, mostly 4–7 ft (1.2–2.15 m)’, with a maximum rooting depth of 2.4 m. However, Weaver (1958) also clearly stated that most grass roots are shallow (> 75% in top 0.1 m). Measurements on our site indicated that 75–80% and 85–90% of cumulative grass root biomass and length, respectively, were found in the top 0.3 m (Wedin unpublished data, Wang et al. 2008). Trees and shrubs clearly have the potential for deeper roots than grasses (Schenk and Jackson 2002), but estimates of the depth to which soil water depletion occurs vary widely among species and ecosystems (Asbjornsen et al. 2008), making determination of source of water uptake for individual species important within functional groups and ecosystems (Goldstein et al. 2008).
Due to the general difference between tree and grass rooting depths, the partitioning of soil water as a resource into shallow and deep layers is often cited as a mechanism for the stable coexistence of trees and grasses in semiarid landscapes (Walter 1979, Hesla et al. 1985, Schulze et al. 1996, Darrouzet-Nardi et al. 2006). Several semiarid woodland findings support the niche partitioning hypotheses: woody species predominantly use deep soil water and grasses predominantly use soil water from the upper soil profile (Sala et al. 1989), although other studies have shown significant competition for soil water between woody species and grasses (Le Roux et al. 1995, Scholes and Archer 1997). Studies that compared the depth of water uptake by co-occurring woody species report both patterns: some species use only deep soil water, while others tap both shallow and deep layers (Dodd et al. 1998, Williams and Ehleringer 2000, West et al. 2007, Goldstein et al. 2008). When trees and grasses use overlapping regions of the soil profile, grasses are assumed to have a competitive advantage for water in the upper soil profile (Belsky 1994), whereas trees alone have access to deeper soil layers where water is available to them during drought periods (Walker and Noy-Meir 1982), contributing to their success in semiarid environments. Daly et al. (2000) modeled tree–grass interactions in water-limited landscapes and concluded that, when trees are more deeply rooted than grasses, trees inevitably become the dominant life form in the absence of fire. Conversely, when trees and grasses share the same rooting zone, grass dominance or co-dominance is predicted.
Determining species’ source of water uptake and role in an ecosystem’s water balance is critical for accurately modeling and predicting species dynamics and the effects of climate change on ecosystem processes (West et al. 2007). Our primary objective was to determine the seasonality and depth of water uptake by two native encroaching woody species (P. ponderosa and J. virginiana) and two C4 grasses (Schizachyrium scoparium (Michx.) Nash and Panicum virgatum L.). The selected grasses dominate the Sandhills of Nebraska and are a major vegetation component of the Great Plains. We used the natural abundance of stable isotopic ratios of hydrogen (
D) and oxygen (18O) to differentiate the source of soil water uptake (Ehleringer and Dawson 1992, Williams and Ehleringer 2000, Phillips et al. 2005, West et al. 2007). We hypothesized that: (1) during periods of high soil water content, grasses and trees use water from overlapping soil profiles, whereas when water is unavailable in the upper soil profile, because of drought or freezing temperatures, tree species acquire water from deeper soil profile, giving them a competitive advantage over grasses and (2) tree species differ in patterns of water uptake that might influence their future distributions.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Site description
The study was conducted at the Nebraska National Forest (NNF) at Halsey, NE, USA (825 m altitude, 41°51'45'' N and 100°22'06'' W). The NNF is an experimental forest established in the semiarid Sandhills grasslands of Nebraska in 1902 and planted with coniferous species, predominantly P. ponderosa. Most of the forest was planted in the 1930s and in the 1940s. In 2005, pine stands in the study area were 60–65 years old. The documented site history and relatively uniform stand age and soils at the NNF make it uniquely suited to investigate the ecosystem effects of tree encroachment in the Great Plains.
The climate is semiarid continental, with a mean annual precipitation of 573 mm of which 77% falls during the growing season (April–September). Mean annual temperature is 8.4 °C. The mean minimum temperature in January is –13.8 °C, and the mean maximum temperature in July is 31.3 °C. Soils are Valentine fine sand (mixed, mesic Typic Ustipsamments) with deep loose sandy well-drained soils (Sherfey et al. 1965). The vegetation of the study area consists of scattered P. ponderosa and J. virginiana and a perennial grassland dominated by little bluestem (S. scoparium), switchgrass (P. virgatum), sand dropseed (Sporobolus cryptandrus (Torr.) A. Gray), Kentucky bluegrass (Poa pratensis L.), and white sage (Artemesia ludoviciana Nutt.). Tree basal area of the study site is 5.6 m2 ha–1 with a density of 123 trees ha–1. Four native species were selected for study: two woody species (J. virginiana and P. ponderosa) and two dominant C4 grass species (S. scoparium and P. virgatum). The selected P. ponderosa trees averaged 0.44 m in diameter at breast height (DBH; range 0.35–0.52 m). Juniperus virginiana trees were 20–30 years old and averaged 0.18 m DBH (range 0.11–0.24 m).
Isotope measurements
Precipitation and groundwater
Precipitation in the study area was measured continuously with a tipping bucket rain gauge (TE525, Campbell Scientific, Logan, UT) connected to a Campbell Scientific data logger (CR10X). Precipitation samples for isotopic analyses were collected monthly from May 2003 to April 2004 using light mineral oil in the collector apparatus to prevent evaporation between sampling events. The collected samples were stored at –10 °C until isotopic analyses. Groundwater (~ 7 m deep, Harvey unpublished data) was sampled monthly from three wells within 2.5 km of the study site between May 2003 and April 2004. The groundwater samples were placed in 8-ml polypropylene vials and stored at –10 °C until isotopic analyses.
Plant and soil sampling
Stems were collected monthly between May 2003 and April 2004 from eight trees per species (P. ponderosa and J. virginiana). Stems were > 0.02 m in diameter and were 0.15 m long. Twenty thick fleshy root crowns were harvested on a monthly basis during the growing season (May–August 2003), from S. scoparium and P. virgatum individuals. In several herbaceous species, root crowns have been shown to be the least variable plant tissues to sample for the assessment of plant source water by isotope techniques (Barnard et al. 2006). Samples were put immediately in glass vials following the method of Dawson and Ehleringer (1998), and placed in plastic bags and stored at –10 °C until isotopic analyses.
Soil profiles were sampled with a truck-mounted hydraulic corer (Giddings Machine Co., Fort Collins, CO) with a barrel (0.0625 x 1.22 m) containing an inert PETG plastic liner. The liner was sealed with plastic caps to prevent drying or evaporation. Two soil cores (up to 3 m each) were collected monthly between May 2003 and April 2004 from two locations at the site. The soil cores were cut into 0.1 m increments from 0 to 0.2 m and 0.2 m increments from 0.2–3 m. Samples were placed in plastic bags and stored at –10 °C until isotopic analyses. Soil samples were not collected in December because winter conditions prevented site access. Similarly, winter conditions prevented access to Location 2 in February and so soil isotope values, although presented, were not included in the plant source of water analysis because the soil core was taken from the site edge and therefore might not be fully representative of Location 2.
Isotopic analyses
Isotopic analyses were performed at the Water Science Laboratory, University of Nebraska–Lincoln. Water was extracted from plant and soil samples by azeotropic distillation in dry toluene (Revesz and Woods 1990). This method has been shown to extract only hydrologically active water. The extracted water and meteoric water (precipitation) samples were analyzed for D (deuterium) by an online chromium reduction technique (Morrison et al. 2001) on a continuous flow isotope ratio mass spectrometer (EuroPyrOH-3110, Isoprime with Eurovector Pyrolysis Furnace, GV Instruments, Manchester, UK). Water samples were analyzed for 18O with an automated CO2 equilibration technique (Midwood et al. 1992) connected to an isotope ratio mass spectrometer (GV2003 with an IsoprepB Equilibration System, GV Instruments). The D and 18O values were expressed in parts per thousand (
) relative to the V-SMOW standard (Gonfiantini 1978)
|
| (1) |
for 18O and 2 for D).
Annual weighted means of D and 18O for precipitation (PWT) were calculated as
|
| (2) |
PWT is the weighted mean D and 18O for rainfall, x is isotopic composition of the ith month, 
x is the fraction contribution of the ith month to the yearly rainfall total, and n is the number of months (i = 1–12). To determine the fractional contribution of water sources to studied species, the IsoSource program was used (http://www.epa.gov/wed/pages/models/stableIsotopes/isosource/isosource.htm) (Phillips et al. 2005). IsoSource finds all feasible solutions of source contributions based on linear mixing models (Phillips and Gregg 2003, Phillips et al. 2005, West et al. 2007), and because the two studied isotopes were highly correlated (r = 0.94), it was not possible to use other traditional mixing models (West et al. 2007). At first, source endpoints were derived from means of the 0.05–0.3, 0.3–0.5, 0.5–0.7, 0.7–0.9 and 0.9–3 m soil depths. However, these sources were too many to allow for feasible solutions; therefore, we combined the data into three layers (0.05–0.5, 0.5–0.9 and 0.9–3 m) post hoc following Phillips et al. (2005).
Soil water content
Soil water content at depths of 0.2, 0.4, 0.6 and 1 m was measured at two locations by time domain reflectometry technology (ThetaProbe ML2 probes, Dynamax Inc., Houston, TX). Data were recorded every 30 min and stored on a Campbell Scientific data logger. Mean values for replicate sensors at the same depth are presented. Sensors were calibrated (High Plains Regional Climate Center, University of Nebraska–Lincoln) for sandy soils and values were compared with other sites equipped with similar sensors at the NNF. Additionally, gravimetric soil water content was determined in subsamples from the two cores collected for soil isotope measurements.
Plant water potential and transpiration
Predawn (pre MPa; 1–1.5 h before sunrise) and midday (mid MPa; 1300 h during winter and 1230 h during summer months) water potentials were measured to assess plant water relations. Measurements were followed once a month between May 2003 and April 2004 for the trees and over the growing season (June–August 2003) for the grasses. Water potential was determined with a pressure chamber (PMS Instrument Co., Albany, OR) on tree twigs and grass leaves collected from the same individuals that were used for stable isotope measurements. Foliar transpiration rates (E; mmol m–2 s–1) were determined at light saturation, concurrently with midday water potential measurements with an infrared gas analyzer (LI-6400, Li-Cor, Lincoln, NE). Detailed measurement conditions for gas exchange measurements have been described by Eggemeyer et al. (2006).
Statistical analyses
Data were analyzed with SAS statistical software (SAS 1996, SAS Institute, Cary, NC) using the mixed models procedure. Orthogonal contrasts were used to examine species differences in seasonal dynamics. Means were separated by the pairwise mean comparison in SAS.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Environmental parameters
Air temperature was consistent with the 30-year mean (High Plains Regional Climate Center, University of Nebraska–Lincoln) reaching a maximum in July and August and a minimum in January and February (Figure 1). Total precipitation for the study period was 370 mm, well below the long-term mean of 573 mm (Figures 1 and 2). The climate diagram (Figure 1) indicates that drought on site started in mid-July and ended in October. Volumetric soil water content (Figure 2) was highest in spring, and fell to a minimum in September. Soil water content was highly variable at 0.2 m depth, reflecting recent precipitation events. Soil water fluctuations were smaller with increasing depth (0.6 and 1 m), and exhibited a steady decline after May, before stabilizing by late August (Figure 2). For much of August and September, soil water content at 1 m was higher than in the upper soil profile. Although comparable continuous measurements of volumetric soil water below 1 m were not made, the gravimetric soil water content of the soil cores used for isotopic analysis averaged between 4% and 7.5% for the 0.9–3 m soil profile during the drought period.
|
|
Seasonal isotopic variations in precipitation, soil water and groundwater
The
18O and D values of precipitation were consistent with regional monthly means (Harvey and Welker 2000), varying seasonally as a function of temperature and ranging from –4.6 to –21 for 18O, and from –25 to 138 for D (Figures 3 and 4). Soil water 18O and D differed significantly between months (P < 0.0001) and depths (P < 0.0001) (Figures 3 and 4). Isotopic values in the 0.05–0.3 m soil profile reflected the seasonal variation in local precipitation and were generally less negative than precipitation during the hottest part of the summer (July–September), because of enrichment associated with evaporation (Figures 3 and 4). Values of 18O and D remained constant deeper in the soil profile, approaching values comparable with groundwater (–11 and –77 for 18O and D, respectively). Although we did not sample the groundwater (~ 7 m deep), minimal isotopic variation was expected below 3 m because a steady state value was likely reached (Zimmerman et al. 1967). Soil water 18O and D also indicated one or more pulses moving down the soil profile between January and April. Pulses were also evident in the May and June profiles, but not in late summer profiles. These pulses may indicate water percolating through the soil profile by piston flow movement of a precipitation event, or they may represent diffusion of a larger seasonal pulse. There was a significant interaction between month and depth (P < 0.003) indicating that the 18O and D of soil water were changing with depth throughout the year.
|
|
The isotopic composition of groundwater from nearby wells was relatively constant over the year averaging –11
± 0.063 for 18O and –77 ± 0.31 for D (Figures 3 and 4). The weighted mean isotopic composition of precipitation over the year was similar to the groundwater yearly mean from the three wells, implying direct local recharge (Gat and Matsui 1971).
Seasonal variations in plant xylem isotopic composition and water uptake
The 18O and D of xylem water did not differ among grass species (S. scoparium and P. virgatum, P > 0.05) (Figures 3–6
). Grass xylem water 18O and D ranged from –4 to –6.5, and –28 to –50, respectively. Values generally tracked recent precipitation and soil water in the upper soil profile. The 18O of grasses was slightly less negative than those of precipitation and soil water on a few sampling dates. The 18O and D of grass xylem water were significantly less negative than those of P. ponderosa and J. virginiana (P < 0.01; Figures 3–6), indicating that grasses and trees might be obtaining water from different depths within the soil profile. The 18O values differed significantly between tree species (P. ponderosa and J. virginiana, P = 0.04), and together with the difference in D between dates (P < 0.001) indicate that, in one or more months, P. ponderosa and J. virginiana were obtaining water from different depths within the soil profile.
|
|
Species were plotted on a
D–18O graph to evaluate fractionation processes (Figure 5). Results showed that grass species plotted along and slightly to the right of the meteoric water line (MWL), whereas tree species plotted on the MWL, indicating that grasses obtained shallow surface waters at or near the ground surface that had been enriched slightly by evaporation and that the trees used deeper water unaffected by evaporation.
A three-source mixing model was used to determine the fraction of water acquired by plants from three ranges of soil depths (0.05–0.5, 0.5–0.9 and 0.9–3 m in Figure 6). The model (Figure 7) predicted that grasses extracted most of their water (~ 100%) from the 0.05–0.5 m soil profile when water was available (May and June), and as the season progressed and soil water declined in this profile, S. scoparium and P. virgatum slightly increased their uptake of deeper water sources (~ 20% of water acquired from below 0.5 m). In August, S. scoparium reverted to obtaining its water almost exclusively from the upper soil profile (0.05–0.5 m). In contrast, P. ponderosa and J. virginiana showed plasticity in their patterns of water uptake. In November, January and February, when below zero air and soil temperatures were recorded, P. ponderosa and to a lesser extent J. virginiana extracted most soil water from below 0.9 m depth. In March, soil 18O and D values were irregular and isotopic values of the tree species overlapped with several soil depths, resulting in less accurate model estimations (large range of model solutions) for the 0.05–0.5 and 0.9–3 m soil profiles. However, with snow melt in March and increases in precipitation, temperature and soil water content in the upper soil profile, and the initiation of fine root production, we infer that water might have been predominately acquired from that source. The upper soil profile (0.05–0.5 m) was also the main source of water in April for P. ponderosa and in April and May for J. virginiana. From May to August, P. ponderosa extracted a sizeable fraction of its water from the 0.05–0.5 and 0.5–0.9 m soil profiles. This was not the case for J. virginiana, where water uptake from the 0.05–0.5 m profile progressively declined from May to August, mirrored by an increase in the fraction of soil water uptake from the 0.5–0.9 m profile, more or less tracking the increase in soil water content within this depth range (Figure 7). Juniperus virginiana had negligible water uptake from below 0.9 m between April and July. In September, soil water content was at its lowest throughout the measured profile and we observed a switch in soil water uptake to sources below 0.9 m depth. The plasticity in patterns of soil water uptake in P. ponderosa and J. virginiana was observed again in October, where both species switched back to acquiring a sizeable fraction of soil water from the upper soil profile (0.05–0.5 m), with a negligible contribution from the driest 0.5–0.9 m soil profile.
|
) and fraction of water uptake from 0.05–0.5, 0.5–0.9 and 0.9–3 m soil depths for trees (P. ponderosa and J. virginiana) and grasses (S. scoparium and P. virgatum) between May 2003 and April 2004. Solid and dashed lines represent the mean and the range of model solutions produced with the IsoSource program, respectively.Seasonal variations in water potential and transpiration
Water potentials differed among species (P < 0.01) and displayed a significant seasonal variation (P < 0.03; Figure 8). The
pre that is an indication of soil and plant water status corresponded well with the observed seasonal variations in volumetric soil water content. Pinus ponderosa had the least negative pre and mid, followed sequentially by J. virginiana and finally the grass species. The most negative pre and mid occurred in August for all species. The most negative pre (–1.0 MPa) and mid (~ –2.5 MPa) were observed in grasses, followed by J. virginiana (pre –0.95 and mid –2.0 MPa) and P. ponderosa (pre –0.8 and mid –1.8 MPa). Grasses senesced in late August in response to drought conditions, whereas water potentials of the tree species increased in September (Figure 8).
|
Transpiration rates (E, Figure 9) increased in spring with the improvement in weather conditions in P. ponderosa, but remained low in J. virginiana before sharply increasing between May and June. Transpiration rates remained high in both tree species for the remainder of the summer before sharply declining in September, because of severe drought conditions. Grasses did not differ between each other, but exhibited significantly higher E than trees in June, before declining in July to similar rates as observed in the tree species.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
To use stable isotopes to identify the depth of soil water uptake by plants, the isotopic values of soil water must show a clear vertical gradient (Ehleringer and Dawson 1992, Dawson 1993, Brunel et al 1995, Dawson and Ehleringer 1998). We observed such a vertical gradient in soil water
18O and D profiles taken over a year in an open savanna–grassland site in the Nebraska Sandhills. Generally, the 18O and D values below 1–2 m depth, depending on the season, were uniform and comparable with the signature of deep groundwater taken from nearby wells. The exception to this was a pulse of 18O- and D-enriched water observed at 1.5–2.5 m depth in May and June. In contrast, soil water between the surface and 1 m depth showed a clear 18O and D profile that changed dramatically during the year. Isotopic ratios of soil water in the upper profile varied and reflected both the isotopic ratio of recent precipitation and evaporative fractionation. Thus, the isotopic soil water profile at the study site was useful in identifying the depth of plant water uptake.
Water sources did not differ between grass species (S. scoparium and P. virgatum) except for August. Isotopic values generally tracked recent precipitation, and the mixing model showed that water was extracted from the upper soil profile (0.05–0.5 m). Soil water uptake from below 0.5 m depth slightly increased during drought in P. virgatum and to a lesser extent in S. scoparium, yet uptake of soil water by deep grass roots appeared to be minimal in relation to the total water use of these species. The reliance on water from the upper soil profile also contributed to the observed more negative in grasses compared with trees. Grasses of the shortgrass steppe of Colorado (Dodd et al. 1998), and the Corn Belt Region of Midwest (USA) (Asbjornsen et al. 2008), were also found to use shallow soil water predominantly. The use of shallow water in semiarid environments might reflect high tolerance to low resource environments and a reliance on rapid fine root turnover in response to water pulses and shortages (Asbjornsen et al. 2008). The fibrous roots of S. scoparium and P. virgatum are within 0.6–1.2 m of the surface, respectively (Stubbendieck et al. 1985), with the majority being in the top 0.3 m (Wang et al. 2008). Although the deeper roots of P. virgatum relative to S. scoparium contributed to a small fraction of total water uptake, they may be an important survival mechanism in the Sandhills and might explain in part the high susceptibility of S. scoparium to prolonged periods of drought compared with P. virgatum and other C4 grass species that dominate the Sandhills (Schacht et al. 2000, Awada et al. 2002). Values of grass 18O were less negative than both precipitation and uppermost soil profile values on a few sampling dates. This might reflect water uptake from very shallow soils that have been enriched by evaporation, or species-specific physiologic and morphologic traits, or both (Wang et al. 1998, Barnard et al. 2006, Asbjornsen et al. 2008). Barnard et al. (2006) reported a general progressive enrichment of plant water 18O from the root crown to leaves in several herbaceous species, concluding that root crowns showed the least variability and are the most suitable if not the best plant parts to sample for 18O analysis in herbaceous species.
In contrast to grasses, P. ponderosa and J. virginiana exhibited significant plasticity in sources of water uptake. During winter, snow accumulated on site and both tree species extracted most of their soil water from below 0.9 m depth, with P. ponderosa possibly reaching deeper water sources than J. virginiana. Deep water uptake during winter could be attributed to low and more extreme soil temperatures in the upper soil profiles, reduced shoot demands, or changing root system function, or a combination thereof (Retzlaff et al. 2001). Winter measurements of E (Figure 9) and photosynthesis (Eggemeyer et al. 2006) showed that P. ponderosa and J. virginiana were physiologically active on warm winter days resulting in possible carbon gain (Awada et al. 2003) in trees while grasses are dormant. In spring, when shallow soil water became available, tree species extracted water from the upper soil profile (0.05–0.5 m) and relied little on water from below 0.5 m depth. Water uptake from the upper soil profile in spring could be attributed to increases in fine root production, root system function as related to nutrient uptake (Retzlaff et al. 2001), and the ability of a species to switch its water uptake to soil profiles with highest water potentials (Plamboeck et al. 1999). During the growing season (May–August), P. ponderosa acquired a large fraction of its water from the 0.05–0.5 and 0.5–0.9 m soil profiles, whereas in J. virginiana, water uptake from the 0.05–0.5 m profile progressively declined between May and August, mirrored by an increase in the fraction of uptake from the 0.5–0.9 m profile, thus tracking the general increase in soil water within the 0.5–0.9 m profile, and being less responsive to growing season precipitation events. In September, soil water content declined to its minimum throughout the measured profile (0.05–3 m), and both tree species shifted their soil water uptake to below 0.9 m depth, and possibly groundwater (~ 7 m deep). We did not measure the rooting depth of tree species on site, but roots of J. virginiana and P. ponderosa are reported to penetrate up to 7.5 and 12 m deep, respectively (Burns and Honkala 1990). Root access to deep water sources and possibly groundwater did not seem to improve E, which significantly declined in September in response to severe drought, but might have played a role in the recovery of water potentials. This might be critical for plant survival during drought (West et al. 2007), especially for P. ponderosa, a species prone to xylem cavitation induced by water stress (Law et al. 2001, Martínez-Vilalta et al. 2004). However, Juniperus spp. are reported to be among the most resistant woody plants to drought-induced xylem cavitation (Willson et al. 2008). Juniperus virginiana has been found to maintain stomatal conductance, photosynthetic activity (Eggemeyer et al. 2006), and in this study E at significantly more negative water potentials compared with P. ponderosa. The latter species was reported to have strong stomatal control (Law et al. 2001), and high sapwood to leaf area ratios (Piñol and Sala 2000) to avoid xylem cavitation, rendering this species sensitive to prolonged periods of drought. These differences in drought resistance strategies between the study tree species might explain the widespread encroachment by J. virginiana in water-limited grasslands relative to P. ponderosa (Schmidt and Stubbendieck 1993). In October, and with the increase in soil water availability in the upper soil profile, J. virginiana switched to extracting water from the 0.05–0.5 m profile, and P. ponderosa used a mixture of water derived from both the 0.05–0.5 and below 0.9 m depths. Depth of water uptake by woody plants in other studies has been shown to be species and ecosystem specific (e.g., Phillips and Ehleringer 1995, Williams and Ehleringer 2000, Weltzin and McPherson 1997, Asbjornsen et al. 2008, Goldstein et al. 2008), which illustrates the importance of studying individual species within ecosystems (Goldstein et al. 2008). For example, although our results are in agreement with what has been reported for the piñon-juniper woodland of southern Utah (West et al. 2007), Pinus edulis Englmn acquired water from both shallow and deep water sources, whereas J. osteosperma Torr. used shallow water when water was available in the spring and shifted to deeper sources for the remainder of the growing season. Furthermore, Pinus taeda L. was reported to use water predominantly from the upper soil profile (Retzlaff et al. 2001). Darrouzet-Nardi et al. (2006) found that the invading shrub Artemisia rothrockii A. Gray used deeper water than native herbaceous species.
The two encroaching tree species acquired water outside of the growing season, competed for shallow water with grasses during spring and early summer, and depended on water drawn from deeper in the soil profile during drought. This will have implications for tree–grass dynamics in semiarid grasslands. A shallow grass rooting zone may facilitate the co-existence of woody species and grasses (Le Roux et al. 1995, Weltzin and McPherson 1997), or give woody species a competitive advantage in the absence of fire, and projected climate change (Daly et al. 2000, Bond et al. 2005, Huxman et al. 2005). Thus, although grasses may outcompete trees for shallow soil water when their rooting zones overlap (Belsky 1994, Daly et al. 2000), trees use deep soil water unavailable to grasses and avoid grass competition for water during the six months each year when the C4 grasses are dormant. Pinus ponderosa and J. virginiana appear to be suited for survival in a semiarid grassland, provided sufficient soil water exists in early spring, deep soil water persists during drought periods, and fire suppression continues to be a dominant policy (Eggemeyer et al. 2006). This has important implications for regional ecohydrology. The Nebraska Sandhills are a major recharge zone for the High Plains (Ogallala) Aquifer (36% and 65% of the aquifer area and water, respectively, are located under Nebraska; Bleed and Flowerday 1998). Szilagyi et al. (2005) estimated that between 6% and 14% of annual precipitation falling across the entire Sandhills region enters the groundwater system (i.e., it is unused by plant roots). This is a surprisingly high groundwater recharge rate for a well-vegetated region where potential evapotranspiration well exceeds actual evapotranspiration on an annual basis (Bleed and Flowerday 1998). However, if as our results suggest, grasses do not rely on water from deep in the soil profile, this high recharge rate is easier to explain. Our results indicate why high groundwater recharge rate may be affected with ongoing encroachment by native conifer tree species into grassland–savanna regions (Huxman et al. 2005). These results are also supported by soil water measurements (2005–2008, unpublished data) that have shown that soils (0.05–3 m depth) in dense sites of J. virginiana and of P. ponderosa are significantly drier than in open grasslands at NNF. The projected increase in the frequency and severity of droughts in semiarid ecosystems with climate change are predicted to further alter the competitive relationship between co-existing species and affect ecosystem processes. For the two studied tree species, prolonged summer drought would have greater effects on P. ponderosa than on J. virginiana, which might have further implications on the distributions of woody species in the semiarid grasslands of Nebraska.
|
Acknowledgements |
|---|
This work was supported by the McIntire Stennis Forest Research Funds-USDA, and an Interdisciplinary Agriculture Research grant, ARD-University of Nebraska–Lincoln. K. Eggemeyer would like to acknowledge a fellowship from the Center for Great Plain Studies. The authors appreciate the assistance of Dr. D. Snow (Water Sciences Laboratory, UNL), Kenneth Elgersma and J. Hellerick, and the cooperation of the US Forest Service. They would like thank the three anonymous reviewers for their comments on this manuscript.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
-
Archer S., Boutton T.W., Hibbard K.A. Trees in grasslands: biogeochemical consequences of woody plant expansion. In: Global Biogeochemical Cycles in the Climate System—Schulze E.D., Harrison S.P., Heimann M., Holland E.A., Lloyd J., Prentice I.C., Schimel D., eds. (2001) San Diego: Academic Press. 115–138.
Asbjornsen H., Shepherd G., Helmers M., Mora G. Seasonal patterns in depth of water uptake under contrasting annual and perennial systems in the Corn Belt Region of the Midwestern U.S. Plant Soil (2008) 308:69–92.
Awada T., Moser L.E., Schacht W.H., Reece P.E. Stomatal variability of native warm-season grasses from the Nebraska Sandhills. Can J. Plant Sci. (2002) 82:349–355.
Awada T., Radoglou K., Fotelli M., Constantinidou H.I.A. Ecophysiology of seedlings of three Mediterranean pine species in contrasting light regimes. Tree Physiol. (2003) 23:33–42.
Barnard R.L., De Bello F., Gilgen A.K., Buchmann N. The 18O of root crown water best reflects water 18O in different herbaceous species. Rapid Comm. Mass Spectrom. (2006) 20:3799–3802.[Medline]
Belsky A.J. Influences of trees on savanna productivity: tests of shade, nutrients, and tree-grass competition. Ecology (1994) 75:922–932.[CrossRef][Web of Science]
Bleed A.S., Flowerday C.A. An Atlas of the Sand Hills. In: Conservation and survey division, IANR (1998) Lincoln: University of Nebraska. 260.
Bond W.J., Woodward F.I., Midgley G.E. The global distribution of ecosystems in a world without fire. New Phytol. (2005) 165:525–538.[CrossRef][Web of Science][Medline]
Brunel J.P., Walker G.R., Kennett-Smith A.K. Field validation of isotopic procedures for determining sources of water used by plants in a semi-arid environment. J. Hydrol. (1995) 167:351–368.[CrossRef]
Burns R.M., Honkala B.H. Silvics of North America: vol. 1. In: Conifers. Agriculture Handbook 654 (1990) Washington D.C: USDA-Forest Service.
Daly C., Bachelet D., Lenihan J.M., Neilson R.P., Parton W., Ojima D. Dynamic simulation of tree-grass interactions for global change studies. Ecol. Appl. (2000) 10:449–469.
Darrouzet-Nardi A., D’Antonio C.M., Dawson T.E. Depth of water acquisition by invading shrubs and resident herbs in a Sierra Nevada meadow. Plant Soil (2006) 285:31–43.
Dawson T.E. Water sources of plants as determined from xylem-water isotopic composition: perspectives on plant competition, distribution, and water relations. In: Stable Isotopes and Plant Carbon–Water Relations—Ehleringer J.R., Hall A.E., Farquhar G.D., eds. (1993) San Diego: Academic Press. 465–496.
Dawson T.E., Ehleringer J.R. Plants, isotopes and water use: a catchment-scale perspective. In: Tracers in Catchment Hydrology—Kendall C., McDonnell J., eds. (1998) Amsterdam: Elsevier. 165–202.
Dodd M.B., Lauenroth W.K., Welker J.M. Differential water resource use by herbaceous and woody life-forms in a shortgrass steppe community. Oecologia (1998) 117:504–512.[CrossRef][Web of Science]
Eggemeyer K.D., Awada T., Wedin D., Harvey F.E., Zhou X. Physiology of two native invasive woody species and two dominant warm-season grasses in the semiarid grasslands of the Nebraska Sandhills. Int. J. Plant Sci. (2006) 167:991–999.[CrossRef][Web of Science]
Ehleringer J.R., Dawson T.E. Water uptake by plants: perspectives from stable isotope composition. Plant Cell Environ. (1992) 15:1073–1082.
Gat J.R., Matsui E. Atmospheric water balance in the Amazon basin: an isotopic evapotranspirational model. J. Geophys. Res. (1971) 96:13179–13188.
Goldstein G., Meinzer F.C., Bucci S.J., Scholz F.G., Franco A.C., Hoffmann W.A. Water economy of Neotropical savanna trees: six paradigms revisited. Tree Physiol. (2008) 28:395–404.
Gonfiantini R. Standards for stable isotope measurements in natural compounds. Nature (1978) 271:534–536.[CrossRef][Web of Science]
Harvey F.E., Welker J.M. Stable isotopic composition of precipitation in the semi-arid north-central portion of the U.S. Great Plains. J. Hydrol. (2000) 238:90–109.
Hesla B.I., Tieszen L.L., Boutton T.W. Seasonal water relations of savanna shrubs and grasses in Kenya, East Africa. J. Arid Environ. (1985) 8:15–31.
Huxman T.E., Wilcox B.P., Breshears D.D., Scott R.L., Snyder K.A., Small E.E., Hultine K., Pockman W.T., Jackson R.B. Ecohydrological implications of woody plant encroachment. Ecology (2005) 86:308–319.
Jackson R.B., Banner J.B., Jobbagy E.G., Pockman W.T., Diana H., Wall D.H. Ecosystem carbon loss with woody plant invasion of grasslands. Nature (2002) 418:623–626.[CrossRef][Medline]
Law B.E., Goldstein A.H., Anthoni P.M., Unsworth M.H., Panek J.A., Bauer M.R., Fracheboud J.M., Hultman N. Carbon dioxide and water vapor exchange by young and old ponderosa pine ecosystems during a dry summer. Tree Physiol. (2001) 21:299–308.
Le Roux X., Bariac T., Mariotti A. Spatial partitioning of the soil water resource between grass and shrub components in a West African humid savanna. Oecologia (1995) 104:147–155.[CrossRef][Web of Science]
Martínez-Vilalta J., Sala A., Pinol J. The hydraulic architecture of Pinaceae – a review. Plant Ecol. (2004) 171:3–13.[CrossRef]
Midwood A.J., Haggarty P., Milne E., McGraw B.A. Factors affecting the analysis of 18O enriched aqueous samples when using CO2 equilibration in vacutainers. Appl. Radiat. Isot. (1992) 43:1341–1347.
Morrison J., Brockwell T., Merren T., Fourel F., Phillips A.M. On-line high precision stable hydrogen isotopic analyses on nanoliter water samples. Anal. Chem. (2001) 73:3570–3575.[Medline]
Phillips S.L., Ehleringer J.R. Limited uptake of summer precipitation by Bigtooth maple (Acer grandidentatum Nutt) and Gambels oak (Quercus gambelii Nutt). Trees (1995) 9:214–219.
Phillips D.L., Gregg J.W. Source partitioning using stable isotopes: coping with many sources. Oecologia (2003) 136:261–269.[CrossRef][Web of Science][Medline]
Phillips D.L., Newsome S.D., Gregg J.W. Combining sources in stable isotope mixing models: alternative methods. Oecologia (2005) 144:520–527.[CrossRef][Web of Science][Medline]
Piñol J., Sala A. Ecological implications of xylem cavitation for several Pinaceae in the Pacific Northern USA. Funct. Ecol. (2000) 14:538–545.
Plamboeck A.H., Grip H., Nygren U. A hydrological tracer study of water uptake depth in a Scots pine forest under two different water regimes. Oecologia (1999) 119:452–460.[CrossRef][Web of Science]
Retzlaff W.A., Blaisdell G.K., Topa M.A. Seasonal changes in water source of four families of loblolly pine (Pinus taeda L.). Trees (2001) 15:154–162.
Revesz K., Woods P.H. A method to extract soil water for stable isotope analysis. J. Hydrol. (1990) 115:397–406.[CrossRef]
Rodríguez-Iturbe R., Porporato A. Ecohydrology of water-controlled ecosystems: soil moisture and plant dynamics (2005) Cambridge: Cambridge University Press. 442.
Sala O.E., Golluscio R.A., Lauenroth W.K., Soriano A. Resource partitioning between shrubs and grasses in the Patagonian steppe. Oecologia (1989) 81:501–505.[CrossRef][Web of Science]
Sala A., Peters G.D., McIntyre L.R., Harrington M.G. Physiological responses of ponderosa pine in western Montana to thinning, prescribed fire and burning season. Tree Physiol. (2005) 25:339–348.[Abstract]
Schacht W.H., Volesky J.D., Bauer D., Smart A.J., Mousel E. Plant community patterns on upland range in eastern Sandhills. Prairie Nat. (2000) 32:43–58.
Schenk H.J., Jackson R.B. Rooting depths, lateral root spreads and below-ground/above-ground allometries of plants in water-limited ecosystems. J. Ecol. (2002) 90:480–494.[CrossRef][Web of Science]
Schmidt T.L., Stubbendieck J. Factors influencing eastern redcedar seedling survival on rangeland. J. Range Manage. (1993) 46:448–451.
Scholes R., Archer S. Tree–grass interactions in savannas. Annu. Rev. Ecol. Syst. (1997) 28:517–544.[CrossRef][Web of Science]
Schulze E.D., Mooney H.A., Sala O.E., et al. Rooting depth, water availability and vegetation cover along an aridity gradient in Patagonia. Oecologia (1996) 108:503–511.[CrossRef][Web of Science]
Sherfey L.E., Fox C., Nishimura J. Soil Survey, Thomas County, USDA, Soil Conservation Service and Forest Service in cooperation with University of Nebraska Conservation and Survey Division. Series 1961, No 9 (1965).
Steinauer E.M., Bragg T.B. Ponderosa pine (Pinus ponderosa) invasion in the Nebraska Sandhills prairie. Am. Midl. Nat. (1987) 188:358–365.
Stubbendieck J., Nickols J.T., Roberts K.K. Nebraska Range and Pasture Grasses. In: Nebraska Cooperative Extension. E.C. 85-170F (1985) Lincoln, NE: University of Nebraska, Department of Agriculture, Cooperative Extension Service.
Szilagyi J., Harvey F.E., Ayers J.F. Regional estimation of total recharge to ground water in Nebraska. Ground Water (2005) 43:63–69.[Medline]
Walker B.H., Noy-Meir I. Aspects of the stability and resilience of savanna ecosystems. Ecol. Stud. (1982) 42:556–590.
Walter H. Vegetation of earth and ecological systems of the geo-biosphere (1979) New York: Springer-Verlag. 290.
Wang X.F., Yakir D., Avishai M. Non-climatic variations in the oxygen isotopic compositions of plants. Global Change Biol. (1998) 4:835–849.[CrossRef]
Wang T., Zlotnik V., Wedin D., Wally K.D. Spatial trends in saturated hydraulic conductivity of vegetated dunes in the Nebraska Sandhills: effects of depth and topography. J. Hydrol. (2008) 349:88–97.
Weaver J.E. Summary and interpretation of underground development in natural grassland communities. Ecol. Monogr. (1958) 28:55–78.
Weltzin J.F., McPherson G.R. Spatial and temporal soil moisture resource partitioning by trees and grasses in a temperate savanna, Arizona, USA. Oecologia (1997) 112:156–164.[CrossRef][Web of Science]
West A.G., Hultine K.R., Burtch K.G., Ehleringer J.R. Seasonal variations in moisture use in a piñon-juniper woodland. Oecologia (2007) 153:787–798.[Medline]
Williams D.G., Ehleringer J.R. Intra- and interspecific variation for summer precipitation use in pinyon-juniper woodlands. Ecol. Monogr. (2000) 70:517–537.
Willson C.J., Manos P.S., Jackson R.B. Hydraulic traits are influenced by phylogenetic history in the drought-resistant, invasive genus Juniperus (Cupressaceae). Am. J. Bot. (2008) 95:299–314.
Zimmerman U., Ehhalt D., Munnich K.O. Soil water movement and evapotranspiration: changes in the isotopic composition of the water. In: Isotopes in Hydrology (1967) Vienna: IAEA. 567–584.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||