INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Xylem is a defining feature of the higher plant. Its primary function is water transport, but it also plays major roles in mechanical support, in nutrient transport, and in signaling and integration of plant responses (Malone, 1996). Some mineral nutrients, particularly Ca²⁺, move almost exclusively in the xylem (Raven, 1983), whereas others, such as K⁺ and Mg²⁺, can move via both xylem and phloem.

Despite its importance for crop growth, xylem transport has rarely, if ever, been analyzed directly. This is because of the difficulty of extracting xylem sap without contamination from neighboring tissues and from cells at cut surfaces. It is especially difficult to obtain xylem sap from plants that are transpiring, because in these, the sap is held under strong negative pressure. Penetration by a sampling device is liable to cause rapid cavitation of the xylem. Probes can sometimes be located within the vessels of transpiring plants (Tyree, 1997), but extraction of sap with such probes is not possible when the sap is at strong negative pressure.

Studies of xylem transport have therefore almost invariably used indirect methods to obtain xylem sap: Root exudates can be collected from the cut stumps of detopped plants; xylem sap can be sucked from pieces of excised stem using modest vacuum (Bollard, 1960); sap can be expressed from the cut ends of excised organs using the pressure bomb. These methods have provided a wealth of comparative information on mineral transport, but they are destructive and cannot be used for analysis of nutrient dynamics within the shoot. In addition, the solute composition of fluids extracted by these methods may not match that of the transpiration stream. This is because of contamination from cut cells and other factors. In addition, methods such as root exudation generate fluid by forces and pathways that may differ from those in operation during transpiration, and they may therefore generate sap with a different solute profile.

A promising and mostly noninvasive method for collecting xylem sap was introduced by Gollan et al. (1992). This uses compressed air to drive water up the plant from root to shoot and to force sap expression at a cut petiole. This method requires a pressurized root volume, and it is not applicable to plants in the field. In addition, it cannot be used on plants in hydroponic culture because the roots of such plants become flooded when pressurized. Thus, there remains a need for a method for routine extraction of xylem fluid from intact, transpiring plants. Any such method should ideally be nondestructive and applicable to a wide range of plant species, tissues, and nutrient ions.

Certain insects are believed to feed on xylem sap. A number of workers have suggested that such insects offer insights into the chemistry, physiology, and transport properties of xylem (Mittler, 1967; Newby, 1980; Horsfield, 1978; Andersen et al., 1989, 1992; Crews et al., 1998). The xylem feeder Homalodisca coagulata has been used to examine C and N dynamics in several plant-insect systems (Brodbeck et al., 1999). Andersen et al. (1989) used a combination of five analytical approaches to determine the mineral nutrient content of H. coagulata excreta.

The meadow spittlebug (Philaenus spumarius) is a common UK insect that feeds on xylem sap. Its prolonged and copious production of watery excreta is compatible only with feeding from the main transpiration stream (Malone et al., 1999). This insect has been used to study Na⁺ exclusion from the shoot xylem of certain wheat (Triticum aestivum) varieties (Watson et al., 2001) and to investigate insect amino acid preferences (Horsfield, 1978). Here, the meadow spittlebug is used in combination with ion chromatography (IC) for rapid nondestructive analysis of all the macronutrient ions in the xylem of intact, transpiring plants. Example results from plants in laboratory and greenhouse are presented.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

IC Assay Performance

IC was used to determine all of the macronutrient cations (Na⁺, K⁺, NH₄⁺, Mg²⁺, and Ca²⁺) and anions (Cl, NO₃, PO₄³, SO₄²) in meadow spittlebug excreta. Performance of the assay was tested by dilution analysis. Dilutions of samples gave results that were linear and parallel to the standard curves for all the ions tested (Fig. 1). When samples were mixed with standards, dilution curves remained linear indicating that there was no interference from other constituents of the sample (Fig. 1). Sensitivity of the analysis was high; it was possible to measure accurately all the sampled ions in triplicate in only 2 µL of excreta. For convenience of handling, a larger amount (5-10 µL) was normally used for each injection.

View larger version (25K): [in this window] [in a new window]   Figure 1.   IC assay performance. A, The chromatogram shows nitrate and phosphate peaks for two sets of dilutions. The peaks have been offset for clarity. One set (thin lines) depicts injections of 1, 2, 3, and 4 mM standard solution. The second set (thick lines) depict injections of the same 1 mM standard, but spiked with 1, 2, 3, and 4 volumes of an excreta sample. B, The graph shows regression lines fitted to peak areas from this chromatogram, and from additional ion peaks in regions of the chromatogram which are not shown. Regression lines for the standard dilutions (white symbols, solid lines) should ideally be straight and pass through the origin. Regression lines for a standard spiked with increasing amounts of sample (black symbols, dashed lines) should similarly be straight and should pass through the points for level 1 of the standard.

IC has been used successfully with a range of plant materials (Masson and Andrieu, 1996), and it offers several advantages over physicochemical methods, particularly for anions (Weiss, 1995). With bulk samples, it is usually necessary to ash the material to remove organic contaminants before IC. This is time-consuming, and it drives off some minerals of interest. With meadow spittlebug excreta, neither ashing nor any other sample preparation was required: The diluted excreta was injected directly onto the IC columns. The method is thus rapid and simple. It was also robust: A single pair of IC columns received over 3,000 excreta sample injections with no detectable deterioration in performance. Excreta samples were routinely stored frozen at 20°C for extended periods. Samples stored in this way for over 2 years showed no change in their apparent macronutrient content.

Colorimetric Assay of Ca²⁺

Where IC is not available or where assay of only a single mineral ion is required, various alternative procedures can be used. An inexpensive colorimetric method was tested here for rapid monitoring of Ca²⁺ in hundreds of excreta samples. Ca²⁺ distribution in the xylem is an important factor in many horticultural disorders (Malone and Andrews, 2001). Purification of excreta before the colorimetric assay was not necessary, and Ca²⁺ concentrations measured in this way correlated well with those obtained by IC (Fig. 2). The meadow spittlebug approach should therefore be accessible to nonspecialist laboratories.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Colorimetric assay. The Ca2+ content of a set of 140 excreta samples was determined using both colorimetric assay and IC. The line is a fitted regression but forced through the origin. Without such forcing, the correlation coefficient was 0.9.

Insect Feeding Patterns

The meadow spittlebug fed throughout the day and night (Fig. 3). Feeding rates were variable, and individuals sometimes showed extended breaks in feeding. However, when plotted against time of day for a population of insects, neither consistent breaks nor any clear diurnal rhythm was apparent (Fig. 3). An increase in feeding activity during the predawn period might have been expected, because at this time, xylem tensions and the energy costs of extraction are lowest. No such increase was apparent.

View larger version (30K): [in this window] [in a new window]   Figure 3.   Feeding rate with time of day. Accumulation of excreta was measured periodically for 3 weeks, in six cages on petioles of greenhouse peppers. The hourly excretion rate is plotted against time of day. A different symbol is used for each cage. Each collection interval spanned several hours, and the point is plotted at the mid-time of the interval. Two mean accumulation rates are shown (solid lines): The lower one is the mean for all data. The upper is similar but it excludes periods of zero feeding.

The meadow spittlebug is highly polyphagous (Thompson, 1994), and it fed to some extent on all the host species that were offered in our experiments. Besides tomato (Lycopersicon esculentum), pepper (Capsicum annuum L. cv Bellboy), and broad bean (Vicia faba), these included sunflower (Helianthus annuus), bean (Phaseolus vulgaris), coleus (Coleus blumei), wheat, and barley (Hordeum vulgare).

In the greenhouse, feeding rates were about double on mature pepper plants than on mature tomato plants. Some insects continued to feed for periods of 8 weeks or more in the greenhouse, but many survived only for a few weeks. For a cohort of insects placed simultaneously on mature pepper plants in the greenhouse, the survival half-time was 9 d.

Evaporation

Evaporation of excreta from the cages was monitored carefully because it could alter the apparent ion concentration. To reduce evaporation, cages were wrapped in polythene film (cling film), and the sides and upper surface of the cages were shaded with aluminum foil. Evaporation rates were measured by injecting known amounts of test solutions into blank cages placed on foliage in the greenhouse, as for the test cages, and collecting after various times to ascertain how much liquid remained. These tests showed that the rate of evaporation averaged 1.7 µL h¹ cage¹ (Fig. 4). This was low compared with the rate of feeding (12-20 µL h¹ cage¹; Fig. 3). When plotted against time of day, evaporation rates showed some suggestion of an elevation during the afternoon, but no clear diurnal rhythm (Fig. 4). The influence of evaporation is reduced by frequent collection of excreta and by discarding of the smaller samples. The latter will be more prone to evaporation because of their greater surface area to volume ratio and because they come from cages with slow or intermittent feeders. In practice, we collected every 3 to 4 h during the day and discarded samples under 40 µL.

View larger version (15K): [in this window] [in a new window]   Figure 4.   Evaporation rate with time of day. Evaporative loss was measured from blank cages (without insects) on pepper plants in the greenhouse. The measurement period was the same as that shown for feeding rates in Figure 4. Each collection interval spanned several hours, and the point is plotted at the mid-time of the interval. The two points showing apparently negative rates of evaporation have probably arisen from pipetting errors.

Ion Content of Meadow Spittlebug

The content of the major mineral cations in meadow spittlebug tissues was measured after ashing at 550°C and was found to be very low (Table I) as in many insects (Clark, 1958; Sutcliffe, 1963). Mineral cation content of tomato leaf tissue is also shown in Table I for comparison. The insect's contents were much lower than the host shoot except in the case of Na⁺, which was present at similar levels in both (Table I).

View this table: [in this window] [in a new window]   Table 1.   Content of the major mineral cations in P. spumarius L. (Homoptera: Cercopidae) and in tomato leaf Measurement was carried out on a pooled sample of 20 individual insects; hence, variance is not available. Values are mg g1 dry wt.

The daily requirement of these insects for minerals and water will be extremely small. The adults do not grow in volume at all. When compared with the quantities flowing through the insect during feeding, their bodily requirements will be negligible. This means that the macronutrient composition of meadow spittlebug excreta will be practically identical to that of the food source, xylem sap.

In the case of K⁺, for example, insects feeding on peppers in the greenhouse were found to contain about 5.5 mg g¹ dry weight (Table I). This will overestimate the amount within the insect's body tissue, because some of this K⁺ will be in transit in the gut. Thus, an individual insect of 2.25 mg dry weight will contain up to 12 µg of K⁺ in its body tissues. The concentration of K⁺ in excreta from these insects is typically about 10 mM (Fig. 5). Thus, when feeding at a normal rate of 20 nL s¹ (Malone et al., 1999), an individual insect will excrete K⁺ at about 0.4 ng s¹. Every 30 min, therefore, the insect will void an amount of K⁺ equal to that accumulated within its body throughout its entire life (1-4 months). Only a vanishingly small proportion of the K⁺ taken in by these insects is retained within the body tissues; virtually all appears in the excreta. Even in the case of Na⁺, which is present at lower levels in xylem, an amount equivalent to the insect's entire body content will be voided every 10 h or so. We can therefore estimate that <1% of the Na in xylem sap is retained by the insect.

View larger version (23K): [in this window] [in a new window]   Figure 5.   Macronutrient dynamics in a mature pepper growing in rockwool in the greenhouse. The raw data for ion levels in xylem sap (excreta) are shown by white symbols. To highlight trends, a moving average is also plotted for each ion (gray and black symbols). The horizontal markers toward the bottom of each graph indicate periods of night.

Toward the end of the season (October/November), the adult females produce about 20 eggs each (Halkka et al., 1967). The eggs are up to 1 mm long and are produced at a rate of three or four per day in the wild. In the vast majority of cages used here, no eggs appeared, but occasionally two or three were found. Nutrient levels within the eggs were not measured, but unless they contain macronutrients at several orders of magnitude more concentrated than the adults, egg laying will also have negligible impact on levels of ions in excreta.

Mineral anion levels within the insect body were not measured, but the arguments above almost certainly apply to these as well as to the cations. One of the major mineral anions in xylem, nitrate, is not used at all by insects.

The meadow spittlebug will not metabolize any significant proportion of the ammonium in xylem sap, but it could generate a significant amount of ammonium by deamination of dietary amino acids (Andersen et al., 1989). It is difficult to estimate how much amino acid there might be in xylem sap. The arguments of Raven (1983) suggest that, in the extreme case in which all the plant's organic N is assimilated in the root, the level of amino acid in xylem sap could average some 2.5 mM. If the insect deaminated all of this, 2.5 mM of ammonium could be delivered into the excreta. This would account for most or all of the ammonium observed in excreta.

To test the importance of insect metabolism on the levels of ammonium in excreta in these experiments, xylem sap was extracted from greenhouse peppers by the method of Bollard (1960). This uses mild vacuum to draw fluid from the base of an excised piece of stem, while successively cutting increments from its distal end. The method is entirely destructive but it should produce transpiration-stream fluid with only limited contamination from live cells. There was no difference between the levels of ammonium in this sap and that in excreta from insects feeding on the same plants immediately before excision (data not shown). This indicates that insect metabolism had very little impact on ammonium levels in excreta. Thus, for ammonium as for other minerals, the levels in excreta are very similar to those in xylem sap.

Attempts have been made to analyze more directly the effect of insect metabolism on levels of macronutrient ions in excreta (Ponder et al., 2002). The approaches used by these authors are not without problems, but they support the conclusion that insect metabolism has little or no effect on the level of mineral ions in excreta, even in the case of ammonium.

Example Data

Figures 5 and 6 illustrate dynamics of macronutrient concentrations in the transpiration stream of intact plants, measured using meadow spittlebug in combination with IC. Results from individual plants are shown, but similar results were obtained simultaneously from neighboring plants. These two data sets come from plants in very different situations: Figure 6 illustrates dynamics of xylem-borne nutrients in a young tomato plant growing in compost in the laboratory, whereas Figure 5 shows the same for a mature pepper plant growing in hydroponic (rockwool) culture in the greenhouse.

View larger version (25K): [in this window] [in a new window]   Figure 6.   Macronutrient dynamics in a young tomato plant growing in compost in the laboratory. Ion levels in xylem sap (excreta) are plotted over a period of several days. The horizontal markers toward the bottom of each graph indicate periods of night.

Note that samples collected at 8 AM represent material that has accumulated during the night, whereas those collected at 12 PM have accumulated during the morning. The rates of delivery of these various ions to the shoot can be calculated by multiplying the concentration data by the prevailing transpiration rate.

The compost-grown plants (Fig. 6) showed pronounced diurnal rhythms in mineral cations: Xylem K⁺ was low during the night, but it rose sharply during the morning. Mg²⁺ and Ca²⁺ varied together, also with a pronounced diurnal rhythm, and out of-of-phase with that for K⁺. Na⁺ concentration remained low throughout. Among the anions, Cl and sulfate varied with a marked diurnal rhythm, which tended to match that for Ca²⁺. Dilution effects arising from the daily rhythm in transpiration cannot account for these patterns. This is because the concentrations of the various ions did not remain constant relative to each other and because there were many intervals during which the concentrations of some ions increased while those of others decreased.

Plants in rockwool were supplied with ample nutrient and had much more K⁺ and nitrate in their xylem sap (excreta) and much less of other cations and anions (Fig. 5) than did the compost-grown plants. Concentrations of the individual cations (particularly K⁺, Ca²⁺, and Mg²⁺) tended to vary together in the rockwool plant, and they never moved in opposite directions. The rockwool-grown plants showed no evidence of diurnal rhythms in any of the ions, although there were differences with time. Both sets of plants exhibited a strong diurnal rhythm in transpiration rate (not shown).

The plants compared in Figures 5 and 6 differ in several respects, including species, size, age, and environment. They were chosen to illustrate the versatility of the technique rather than to make comparisons between the plants. However, from this and similar data, a pattern is suggested in which compost-grown plants show strong diurnal rhythms whereas hydroponic (including rockwool) plants do not. The rhythms in xylem nutrients in our compost-grown plants (Fig. 6) have similarities to those reported from soil-grown castor bean (Ricinus communis) by Schurr and Schulze (1995): Both show marked diurnal rhythms in individual ions and marked changes in the ratios of the various ions through the cycle. Both data sets show a diurnal cycle of xylem K⁺ (with maxima during the day), which is out of phase with the rhythm in Ca²⁺ and Mg²⁺ (maxima at night). Dilution effects do not explain these phenomena, but some possible mechanisms are discussed in Schurr and Schulze (1995). It may also be that the increase in xylem Ca²⁺ and Mg²⁺ during the night reflects an increased contribution from root pressure at this time. Preliminary findings in our laboratory indicate that root exudate, which is also derived from root pressure, has similarly elevated profiles of Ca²⁺ and Mg²⁺. Further work is currently in progress with meadow spittlebug to assess directly whether there is a fundamental difference between nutrient dynamics in plants in hydroponic culture and those in soil.

In summary, data presented here show that meadow spittlebug in combination with IC, provides a versatile and sensitive method for nondestructive analysis of the dynamics of macronutrient transport in the transpiration stream.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS AND DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

For experiments in the laboratory, seedlings of tomato (Lycopersicon esculentum L. Mill cv Ailsa Craig) and broad bean (Vicia faba) were grown to about 25 cm height in 10-cm pots of Levington's compost. The plants were maintained on a bench by a window, with supplementary lighting of 100 µmol m² s¹ (at plant height) for 14 h d¹ from overhead fluorescent lamps. The midday light level at the leaf surface was typically 200 to 350 µmol m² s¹. Plants were watered every 2nd d but were not given supplementary nutrients.

Experiments were also conducted in the greenhouse on mature pepper plants (Capsicum annuum L. cv Bellboy). These were grown in rockwool under semicommercial conditions with a comprehensive nutrient supply (van de Vooren et al., 1986) and natural lighting. Their stems were 2 to 3 m in length at the time of these experiments (July-September 1999). For the example data, excreta samples were collected from cages situated midway along the petiole of leaves midway up the plants (about 1.5 m above the root).

Insects

Meadow spittlebug (Philaenus spumarius L. [Homoptera: Cercopidae]) adults were collected using a sweep net and aspirator. At our location (52° 12.4' N 1° 36.1' W) the nymphs of this insect, conspicuous by their "spittle" masses, appear on a wide variety of host plants from mid-May. The adults appear some 3 weeks later and remain evident until late October. They were locally very common on uncultivated land and, later in the year, on pasture land. Meadow spittlebug adults are about 6 mm long. They exist in several color-pattern morphs and are extremely variable in appearance (Stewart and Lees, 1996). Meadow spittlebug drags the rear pair of legs when walking; this feature is useful in distinguishing it from several generally similar species in the field. Both males and females were collected.

The insects were kept in the laboratory, on 4- to 8-week-old broad bean seedlings, growing in 4 cm of Levington's compost in a glass aquarium tank of dimensions 30 × 30 × 60 cm. The tank was laid on its side, and the open front of the tank was covered by plastic gauze of mesh size 1 mm. The tank was placed under supplementary lighting as above.

Cage Design

The insects were caged onto petioles at any desired position on the plant. Each cage was made from a 30-mm diameter petri dish (Fig. 7). A pair of diametrically opposed U-shaped notches was cut into the side walls of the base of the petri dish to receive the petiole. A disc of plastic gauze was placed in the bottom of the cage; capillary action along the fibers of this gauze helped to draw all the excreta fluid to a common collection point. The lid of a smaller petri dish (25 mm in diameter) was inverted in the bottom of the cage to hold the gauze flat, and to prevent meadow spittlebug from falling into any accumulated excreta. The cage was secured onto the petiole by a paper clip and elastic band (Fig. 7). When assembled, the single large central hole (approximately 4 mm diameter) in the cage lid was obscured by the elastic band. Smaller holes (approximately 1 mm in diameter) were placed around the periphery of the lid to permit collection of the clear excreta fluid with a Hamilton syringe. Excreta was collected into 0.5-mL Eppendorf tubes, weighed, and stored at 20°C before analysis.

View larger version (19K): [in this window] [in a new window]   Figure 7.   Diagram of cage assembly. For work in the greenhouse, cages were also covered by cling film and shaded with aluminum foil.

An aspirator was used to add insects through the hole in the cage lid, usually two insects per cage. To reduce evaporation, the entire cage was then enclosed in clear plastic film (cling film), and its top and sides were shaded by aluminum foil.

IC

Excreta samples were thawed, shaken, and centrifuged briefly to pool the tube contents. Ten microliters of excreta (or standard) was typically transferred to a 0.5-mL vial (part no. 038142, Dionex, Camberley, Surrey, UK) and diluted with 300 µL of water or eluent. The vials were then capped, mixed by shaking, and loaded onto an autosampler (AS40, Dionex) attached to a Dionex DX120 IC system. Cation content was determined using an "Ionpac" analytical column (CS12A 4 mm) with matching guard column. The eluent was 20 mN H₂SO₄ made up in ultra pure water (>18 M cm¹). Eluent cations were suppressed with a self-regenerating suppressor (CSRS-ULTRA 4 mm), and sample cations were detected by conductivity. Anions were determined in the same way, except using the corresponding anion columns and suppressor. The eluent for anions was a mixture of 0.4 mM NaHCO₃ and 5 mM Na₂CO₃.

Colorimetric Assay of Ca²⁺

This was conducted in ELISA plates. Three microliters of excreta or standard was added to each well, plus 190 µL of Tris-buffered calcein reagent (586C, Sigma-Aldrich, St. Louis; Mann and Green, 1988). The plate was covered and incubated at room temperature for 5 min. Absorbance was read at 595 nm.