Uptake and translocation of manganese in seedlings of two varieties of Douglas fir (Pseudotsuga menziesii var. viridis and glauca)

Authors


Author for correspondence: Andrea Polle Tel: +49 551 393480 Email: apolle@gwdg.de

Summary

  • • Douglas fir (Pseudotsuga menziesii) variety glauca (DFG) but not the variety viridis (DFV) showed symptoms of manganese (Mn) toxicity in some field sites. We hypothesized that these two varieties differed in Mn metabolism.
  • • To test this hypothesis, biomass partitioning, Mn concentrations, subcellular localization and 54Mn-transport were investigated.
  • • Total Mn uptake was three-times higher in DFG than in DFV. DFV retained > 90% of 54Mn in roots, whereas > 60% was transported to the shoot in DFG. The epidermis was probably the most efficient Mn barrier since DFV contained lower Mn concentrations in cortical cells and vacuoles of roots than DFG. In both varieties, xylem loading was restricted and phloem transport was low. However, sieve cells still contained high Mn concentrations.
  • • DFV displayed higher biomass production and higher shoot : root ratios than DFG. Our results clearly show that both varieties of Douglas fir differ significantly in Mn-uptake and allocation patterns rendering DFG more vulnerable to Mn toxicity.

Introduction

Douglas fir (Pseudotsuga menziesii) occurs naturally in North America. Because of its fast growth and good wood properties it has been cultivated outside of its native range, in Europe for almost 200 yr. World-wide, silviculture of Douglas fir is currently expanding (Herman & Smidt, 1999). Estimations predict that the proportion of Douglas fir stands will increase up to 10% of the total forest area in Germany (Kleinschmit, 1991; Knoerzer, 1999). To test the performance of exotic trees, various subspecies of Douglas fir have been planted in Germany, among them the two main varieties: P. menziesii var. viridis (DFV), the coastal type originating from the coastal regions of the mountains from British Columbia to California, and P. menziesii var. glauca (DFG), the interior type spreading along the Rocky mountains to the south-west of the USA to Mexico (Kleinschmit et al., 1974). On the basis of quantitative traits such as growth rate, frost, and disease resistance the coastal variety (DFV) was found to be more appropriate for cultivation in Germany than the interior type (DFG) (Kleinschmit et al., 1974; Larsen, 1978, 1981; Schober et al., 1983, 1984; Liesebach & Stephan, 1995).

Although the performance of this introduced tree species was generally good, in some regions e.g. in Rhineland-Palatinate (Rheinland-Pfalz, Germany) severe disease symptoms such as needle necrosis and defoliation, as well as formation of dark slimes and bark necrosis, were observed in young stands of Douglas fir (20–40 yr) (Schöne, 1992). Needles of injured trees contained excessive concentrations of manganese. Therefore, the decline of Douglas fir in some sites in Germany has been related to manganese (Mn) toxicity (Schöne, 1992). By analysis of the genetic structures, the damaged stands in Rhineland-Palatinate were identified as belonging to the interior variety (DFG), whereas healthy stands were more closely related to the coastal variety (DFV) (Leinemann, 1996). Since healthy and damaged trees grew in close vicinity, it has been suspected that the two varieties differed in Mn-uptake rates (Schöne, 1992). It is unknown whether this might be a plant-inherent feature or modulated by tree–environmental interactions. Experiments to investigate the physiology of Mn metabolism in the two subspecies of Douglas fir have not yet been conducted.

Manganese is an essential nutrient element necessary for activation of a wide range of enzymes and an indispensable constituent of the catalytic centre of enzymes such as Mn-superoxide dismutase and the water splitting complex. To fulfil its metabolic functions, Mn is only necessary at low concentrations (20 µg g−1 dry mass, Marschner, 1995). However, tissue concentrations of Mn may vary considerably. In their natural habitats, the coastal variety, DFV contained 100–800 µg Mn g−1 dry mass and the interior type, DFG 200–2000 µg Mn g−1 dry mass, respectively (Zasoski et al., 1990; Baronius & Fiedler, 1996).

Plant availability of Mn depends on soil properties and on root exudates for Mn chelation or reduction. The availability of Mn increases with decreasing soil pH (Marschner et al., 1986). In general, excess Mn causes chlorosis and necrosis, the appearance of brown, necrotic spots or small reddish purple spots and sometimes, dark root tips (Horst, 1988). High concentrations of Mn interfere with the absorption, translocation, and utilization of other mineral elements such as calcium (Ca), magnesium (Mg), iron (Fe) and phosphorus (P) (Clark, 1982), stimulate phenolic metabolism (Brown et al., 1984; Wissemeier & Horst, 1992), affect energy metabolism, respiration rates (Nable et al., 1988) and cause oxidative stress (Del Rio et al., 1985; Panda et al., 1986; Horst et al., 1999; Fecht et al., 2001). Therefore, Mn uptake by plants is strictly regulated (Duc̆ić & Polle, 2005).

To date, most studies on Mn uptake and translocation have been conducted with agronomically important plants or with unicellular model organisms such as bacteria and yeast. In crops such as spinach, pea, bean and various cereals race-specific differences in Mn toxicity have been reported (Horst, 1983, 1988; Foy et al., 1988; Graham, 1988; Rout et al., 2001). Much less is known about uptake, translocation and toxicity of Mn in forest trees and virtually nothing about varietal differences in Mn metabolism. Since Douglas fir is of increasing silvicultural interest and field data point to differences in Mn toxicity in different subspecies, the goal of this study was to find out whether DFG and DFV differ in Mn metabolism under nontoxic conditions. For this purpose, Douglas fir seedlings were exposed to radioactively labelled Mn. Uptake, transport and allocation of Mn were investigated in the whole plant. Furthermore, transmission electron microscopy coupled with X-ray microanalysis was used to characterize the subcellular Mn localization.

Materials and Methods

The varietal origin of seed lots

Seeds of P. menziesii L. DFV and DFG were purchased from Niedersachsen Forstamt (Oerrel, Munster-Oerrel, Germany) and Sheffield's Seed Company (Locke, NY, USA), respectively. To evaluate the varietal origin of the two seed lots, genetic structures at the isozyme gene loci for 6-phosphogluconate-dehydrogenase (6-PGDH-A) were analysed according to Leinemann (1996, 1998). Endosperm and embryo from each seed were extracted and separated by horizontal starch gel electrophoresis using a Tris-citrate buffer, pH 7.4 (Leinemann, 1998). Activity staining was conducted according to Rothe (1994).

The zymograms of the two seed lots showed strong differences in their banding patterns (Fig. 1). As expected the allele 6-PGDH-A3 is dominant in P. viridis samples, whereas the P. glauca samples shows higher frequencies of homozygote varieties containing the allele 6-PGDH-A6 (Leinemann, 1996). This result confirms the varietal origin of the two seed samples.

Figure 1.

Zymograms of 6-phosphogluconate-dehydrogenase from seed endosperm and embryos of Pseudotsuga menziesii viridis (a) and glauca (b). Alleles from A1 to A7 are marked. Each pair of tracks shows the isozyme pattern from one endosperm and its corresponding embryo, respectively.

Plant material

Seeds of both varieties of Douglas fir were soaked in tap water for 7 d at 2°C and surface-sterilized in 96% ethanol for 30 s, in 0.2% HgCl2 for 30 s and in 30% H2O2 for 45 min. Subsequently, the seeds were placed on sterile 1.5% (w : v) water-agar, pH 4.5 in Petri-dishes (140 mm diameter), maintained for 7 d in darkness at 21°C and subsequently for 3 wk with a day/night regime of 16 h/ 8 h (white light of 150 µmol m−2 s−1 photosynthetic photon flux; OSRAM L 18-W/21-840) at 23°C/ 21°C air temperature. After germination for 7 d, the plants were transferred to hydroponic solutions. Aerated nutrient solution contained the following nutrient elements: 300 µm NH4NO3, 100 µm Na2SO4, 200 µm K2SO4, 60 µm MgSO4, 130 µm CaSO4, 30 µm KH2PO4, 10 µm MnSO4, 92 µm FeCl3, and 5 ml of a stock solution of micronutrients (0.1545 g l−1 H3BO3, 0.012 g l−1 NaMoO4, 0.0144 g l−1 ZnSO4 and 0.0125 g l−1 CuSO4 per litre of nutrient solution). The pH was adjusted to 4. The solution was changed every third day.

Experimental set-up to determine Mn transport

Young DFG and DFV plants were transferred into specially constructed exposure boxes (Fig. 2). The exposure boxes consisted of four separated small chambers (Fig. 2a). Three seedlings were inserted in chamber B. The root system of each seedling was divided and spread into chamber A as well into chambers C and D. Aliquots of 54MnCl2 (0.86 MBq or 0.75 MBq of 54Mn in 0.5 m HCl in two experiments; Perkin Elmer, Boston, MA, USA) was added to the nutrient solution in chamber A once at the beginning. The pH was adjusted to 4 by 0.5 m NaOH. The intermediate chambers B and C served to avoid spill over and contamination of the root parts in chamber D. The boxes were covered with a lid to avoid evaporation. The nutrient solution was aerated with syringes introduced into the solution through perforations in the lid. The level of the nutrient solution in each chamber was checked through a transparent side of the apparatus. Nutrient solution was supplied with a syringe as necessary and the added volume was recorded. The plants were maintained for 21 d in the exposure box under the following conditions: temperature 20°C, relative humidity 40% and white light 24 h (200 µmol m−2 s−1 photosynthetically active photon flux; OSRAM Powerstar HQI-T/D).

Figure 2.

Experimental set-up for manganese-54 (54Mn) feeding of Douglas fir (Pseudotsuga menziesii) seedlings. (a) Exposure box with four chambers to expose split roots to 54Mn in chamber A. (b) Exposure boxes containing plants. (c) Tissues taken for biomass, Mn, and 54Mn analyses after harvest: needles from the top, middle and bottom, roots from the labelled side (tip-ra, middle-rb and upper part-rc) and roots from the non-labelled side (tip-a, middle-b and upper part-c).

Needles from the bottom and upper part of the crown were harvested every third day to determine radioactivity. After 21 d plants were separated into root, stem and needles, and these parts were further divided in three pieces: upper part, middle and bottom part (as indicated in Fig. 2c). Each experiment was conducted with six to nine individual seedlings in three replicate exposure systems. The experiments were repeated with DFG and DFV seedlings of the same age and of similar biomass.

Radioactivity was quantified in a γ-counter (Automatic gamma counter 1480 Wizard 3″; Wallac, Turku, Finland) using 50 mg of dry mass of each tissue.

Distribution of 54Mn in whole plants was detected by autoradiography using a phosphor imaging plate scanner (Fuji BAS 1500) on imaging plates (BAS-III, Fuji) with 85 min to 21 h exposure times.

Element analyses

Samples for element analyses (root, stem and needles) were dried to a constant weight and ashed at 170°C in 65% HNO3 for 12 h (Feldmann, 1974). Elements were determined by inductively coupled plasma–atomic emission spectroscopy (Spectro Analytical Instruments, Kleve, Germany).

Energy dispersive X-ray microanalyses (EDX)

Needle and root pieces taken 15 mm from the root tip from 1.5-month-old seedlings were cut in several 2-mm and 5-mm long pieces, respectively, and were rapidly frozen in a mixture of propane–isopentane (2 : 1) cooled with liquid nitrogen to −196°C in a steel mesh. Samples were freeze-dried at −45°C for 3 d and stored at −20°C in a desiccator over silica gel. For transmission electron microscopy, freeze-dried samples were infiltrated with ether in a vacuum-pressure chamber and embedded in styrene-methacrylate using a technique specifically developed for analysis of diffusible elements (Fritz, 1989). Sections 1-µm thick were cut dry using glass knives, mounted on adhesive-coated 100-mesh hexagonal grits, coated with carbon and stored over silica gel. Details and testing of the method have been reported previously (Fritz, 1989; Fritz & Jentschke, 1994). The samples were analysed with a Philips EM 420 with the energy dispersive system EDAX DX-4 (EDAX Inc., Mahwah, NJ, USA). The accelerating voltage was 120 kV, the take-off angle 25° and counting time 60 live seconds. The Mn concentrations in cross-sections of roots and needles were analysed in cell walls and vacuoles of the following tissues: epidermis, needle mesophyll, cortex of roots, endodermis, xylem and phloem. Nine replicates were analysed in each compartments in three different plants.

Statistical analyses

Data are means (± SD) of five to nine seedlings. Statistical analysis of the data was performed using student's t-test or analysis of variance (anova) followed by a multiple range test (LSD, Statgraphics 2.1; StatPoint, Inc., St Louis, MO, USA). Means were considered to be significantly different from each other, if the level of significance was P  0.05.

Results

Growth performance, biomass, Mn partitioning and subcellular distribution in Douglas fir var. viridis and glauca

Growth of DFV and DFG seedlings differed significantly resulting in 1335 ± 118 and 250 ± 25 mg dry mass per plant, respectively (P = 0.023), after 5 months’ culture in hydroponic solutions. The DFV seedlings reached shoots lengths and biomass similar to those of 5-month-old DFG seedlings after only 3.5 months (Table 1). Biomass partitioning to the major plant compartments showed pronounced differences between the two varieties with DFV favouring above-ground growth and DFG below-ground growth (Table 1). Although DFG developed almost three-times longer and, thus, more root biomass than DFV (Table 1), the root system of DFV was more branched displaying many small lateral roots. The subspecies-related differences in above- and below-ground partitioning reported here for hydroponically grown plants were also found in Douglas fir seedlings cultivated in soil (data not shown).

Table 1.  Biomass, growth parameters, and manganese (Mn) concentrations of Pseudotsuga menziesii var. viridis (DFV) and var. glauca (DFG)
 DFV Mean ± SEDFG Mean ± SEP
  1. DFV was harvested after 3.5 months and DFG after 5 months. Data are means of n = 5–9 (± SE).

Biomass (per plant)
Needles (mg)129 ± 18 93 ± 120.319
Stem (mg) 41 ± 3 29 ± 40.014
Root (mg) 95 ± 14128 ± 130.137
whole plant (mg)265 ± 33250 ± 250.874
Growth morphology
Ratio root: shoot 0.6 ± 0.3 1.0 ± 0.30.002
Root length (mm)134 ± 48352 ± 970.002
Shoot height (mm)103 ± 18 91 ± 140.348
Mn concentrations
Mn in needles (mg kg−1)130 ± 28216 ± 280.068
Mn in stem (mg kg−1)131 ± 39135 ± 300.963
Mn in root (mg kg−1)238 ± 25178 ± 240.405
Mn content per whole plant (µg)646 ± 106616 ± 840.855

The needle concentrations of Mn tended to be slightly elevated in DFG compared with DFV (Table 1) and were in the same range as those of field-grown mature Douglas firs (Baronius & Fiedler, 1996). Stem and root Mn concentrations did not differ between the two varieties (Table 1). The Mn content of whole seedlings of DFG was similar to that of DFV (Table 1). Because of the differences in biomass partitioning, both DFV and DFG had slightly different Mn allocation patterns with relative portions of 37%, 12% and 50% Mn and of 43%, 8% and 49% Mn in needles, stem and roots of DFG and DFV, respectively.

To find out whether the varieties differed in the subcellular distribution of Mn, we investigated cross-sections of root tips employing energy dispersive X-ray microanalyses (Fig. 3a). Epidermal and cortex cell walls of DFG contained significantly higher concentrations of Mn than those of DFV (Fig. 3a). Furthermore the vacuolar Mn concentration in DFG epidermis cells was almost eight times higher than in DFV. The relative enrichment of Mn in DFG compared with DFV decreased towards the endodermal barrier and was absent in the vascular system (Fig. 3a). In DFV maximum Mn concentrations were found in epidermal cell walls. In all other locations analysed in DFV root cross sections, Mn was present at low concentrations displaying no obvious differences between cells types and subcellular compartments (Fig. 3a).

Figure 3.

Spatial resolution of manganese (Mn) concentrations in different cell types and subcellular compartments of roots (a) and needles (b) of Pseudotsuga menziesii var. viridis (DFV, closed bars) and var. glauca (DFG, open bars). Cross-sections were analysed by transmission electron microscopy (TEM)–energy dispersive X-ray microanalyses (EDX). Bars indicate means of n = 6–9 (± SE).

In needles, no significant subspecies-related differences in the localization of Mn were observed (Fig. 3b). However, unlike in roots, both varieties showed differences with respect to the subcellular location of Mn with significantly higher concentrations in cell walls than in vacuoles (P = 0.025 and P = 0.010 for DFV and DFG, respectively, Fig. 3b). It is also remarkable that Mn concentrations in needle cell walls were generally higher than in root cell walls (Fig. 3), which might be due to the fact that Mn is considered a relatively phloem-immobile element (Loneragan, 1988) not easily re-translocated to other tissues. Bearing this in mind it was even more surprising to find very high Mn concentrations in sieve cells of the needle phloem of both varieties (Fig. 3b). This observation suggests that Mn must be either immobilized by unknown mechanisms in sieve cells or that Mn can be transported and circulated in both Douglas fir varieties.

Uptake, transport, and allocation of Mn in Douglas fir var. viridis and var. glauca

To find out whether the two varieties of Douglas fir show differences in Mn-uptake and transport, young seedlings were exposed to 54Mn in the nutrient solution. A split-root system was used to investigate the possibility of re-translocation of Mn (Fig. 2). The kinetics of 54Mn accumulation was determined in top (i.e. younger needles) and bottom (i.e. older needles) of DFG and DFV, respectively (Fig. 4). The accumulation rate of Mn in old needles of DFV was similar to that of DFG during the first 2 wk of the experiment (Fig. 4b). Thereafter, no further accumulation of 54Mn in older needles of DFV was found and after 3 wk the concentration of 54Mn tended to decrease (Fig. 4b). Although there was a restriction in Mn translocation to older needles in DFV, an accelerated increase in 54Mn was observed in DFG after the initial lag-phase (Fig. 4b).

Figure 4.

Time-dependent accumulation of 54Mn in needles of Pseudotsuga menziesii var. viridis (DFV, closed circles) and var. glauca (DFG, open circles). Radioactivity was applied to chamber A of the exposure box (see Fig. 2). Radioactivity was determined in top (a) and bottom needles (b), as indicated in Fig. 2. n = 5–9 (± SE). When no error bars are apparent, they were smaller than the symbols.

Varietal differences in Mn transport were even more pronounced in young needles (Fig. 4a). After an initial lag-phase of c. 10 d, young needles of DFG showed about two-times higher 54Mn accumulation rates than old needles (Fig. 4). Unlike DFG, transport of Mn to young needles was strongly suppressed in DFV (Fig. 4a) and even lower than to older needles (Fig. 4b). Since we observed that DFG consumed higher volumes of nutrient solution per plant than DFV (Table 2), despite less needle biomass (Table 1), we suspect that higher transpiration rates in DFG may have led to higher transport of Mn to the above-ground compartment.

Table 2.  Consumption of nutrient solution during the whole experimental period (21 d) expressed per plant and expressed as daily consumption per needle biomass in Pseudotsuga menziesii var. viridis (DFV) and var. glauca (DFG)
Consumption of nutrient solutionDFVDFG
  • *

    , P  0.05.

Total volume in 21 d (ml)150 ± 10211 ± 41*
Rate (ml g−1 dry mass × d−1) 55108

After 3 wk, the seedlings were harvested to image Mn distribution at the whole-plant level by autoradiography. Figure 5 displays typical examples of whole plants and their corresponding autoradiograms. We found that DFV was characterized by more needle and less root biomass than DFG (Fig. 5a,e, and Table 1) and showed strong 54Mn accumulation in those root parts exposed to the labelled solution (Fig. 5b). The DFV autoradiograms showed little 54Mn in stem and needles (Fig. 5b). Unlike DFV, stem and needles of DFG contained strong 54Mn activity (Fig. 5d). It appeared that Mn was retained at the bottom of needles and not transported to the needle tip (Fig. 5d). Roots of both DFG and DFV exposed to the non-labelled solution also displayed radioactivity (Fig. 5f,h). This suggests that in respect to important differences in Mn allocation between both varieties, DFG and DFV might be able to re-translocate and probably exude Mn.

Figure 5.

Photographs and autoradiograms of whole seedlings of Pseudotsuga menziesii viridis (DFV) and glauca (DFG) after exposure to 54Mn in a split-root system. Photographs of whole, dry seedlings and their corresponding roots from the nonlabelled chamber (a,e = DFV, c,g = DFG). Autoradiograms of whole, dry seedlings and their corresponding roots from the nonlabelled chamber (b,f = DFV, d,h = DFG). Colours from blue to red represent increasing intensities of 54Mn labelling.

To quantify varietal differences in Mn-uptake and allocation, radioactivity was determined in different plant fractions as indicated in Fig. 2 and expressed on the basis of dry mass (Fig. 6a) or as specific activity on the basis of the tissue concentration of Mn (Fig. 6b). The analysis of root segments from different parts of the exposure chamber and analysis of stem and young, middle and older needles confirmed the differences of the Mn distribution in DFG and DFV found by autoradiography: DFV root segments from the labelled solution contained 25 times more 54Mn concentrations than those of DFG (Fig. 6a). In DFV, 54Mn decreased strongly towards the stem and was maintained at low, relatively even levels in stem and needles. By contrast, the concentration of 54Mn was relatively stable throughout the part of the root system of DFG that had access to the labelled solution. In stem and needles of DFG a sharp increase in 54Mn concentrations occurred resulting in 22-times higher activities and 14-times higher specific activities than in DFV. A decrease in activity towards the root tip was found in roots spread across the nonlabelled solutions (Fig. 6a). The DFV roots from the nonlabelled chambers also contained 54Mn, although at lower concentrations than those of DFG.

Figure 6.

Manganese-54 (54Mn) in different parts of seedlings of Pseudotsuga menziesii var. viridis (DFV, closed bars) and var. glauca (DFG, open bars) after exposure to 54Mn applied in a split-root system. Bars indicate means on the basis of dry mass (a) and as specific activity on the basis of the Mn concentration of the same tissue (b). n = 5–9 (± SE). Denomination of tissues, root from the labelled solution: tip-ra, middle-rb and upper part-rc; and non-labelled solution: tip-a, middle-b and upper part-c.

Analysis of whole-plant uptake of 54Mn revealed that DFV acquired 14.7% and DFG 41% of the total radioactivity supplied with the nutrient solution. In DFV c. 90% of the activity taken up by the plants was retained in the root system. Only a small (less than 7%) fraction of radioactivity was allocated to stem and needles (Fig. 7). Under the same conditions, DFG retained c. 43% of newly taken up Mn in roots and allocated c. 54% to stems and needles. In both varieties fractions of c. 3% of the total activity taken up by the plants were present in nonlabelled root parts (Fig. 7). At harvest, the nutrient solution in the nonlabelled chamber D (Fig. 2) contained 0.043% and 0.15% of the total radioactivity administered to DFV and DFG, respectively.

Figure 7.

Allocation of radioactivity to different tissues of seedlings of Pseudotsuga menziesii var. viridis (DFV) and var. glauca (DFG) after exposure to 54Mn applied in split-root system; n = 5–9. Pies indicate relative distribution of radioactivity in the following plant parts: roots from nonlabelled solution (black), roots from labelled solution (grey), stem (white) and needles (hatched).

Discussion

The most important result of this study is that uptake and translocation of Mn differed fundamentally in seedlings of DFG and DFV, the two main varieties of Douglas fir (Figs 5 and 6). First, uptake rates were c. three-times higher in DFG than in DFV, despite similar plant size (Fig. 7, Table 1) and second, Mn was more readily transported to the shoot, in particular to the youngest needles in DFG than in DFV resulting in pronounced differences in Mn allocation patterns (Figs 6 and 7). To date, few studies have addressed Mn uptake and translocation in tree species. Lin et al. (1995) applied 54Mn and 65Zn to the soil surface of balsam fir (Abies balsamea) seedlings and found the following 54Mn distribution: 31% in roots, 31% in twigs, 26% in stems and 12% in needles, with a preferential allocation to younger needles. Our study on the two varieties of Douglas fir showed that whole-plant allocation of the micronutrient Mn is a genetically determined trait, which differs not only between different species but also between closely related subspecies. Lin et al. (1995) also studied Zn translocation and observed that roots retained 86% of the total 65Zn taken up, indicating significant differences in the translocation of two the micronutrients Zn and Mn, respectively. Such differences in element mobility of different nutrients are known (Marschner, 1995). Our study shows that pronounced differences also exist for same element in different varieties of the same species.

Race-specific differences in Mn metabolism have mainly been investigated in agricultural crops in relation to Mn toxicity or tolerance (Horst, 1988; Rengel, 2000; Duc̆ić & Polle, 2005). In the present study the subject of Mn toxicity has not been addressed. However, the study was based on the observation that in some plantations DFG showed excessive Mn accumulation in needles as well as bark necrosis and slime flow, leading to a decline of these stands (Schöne, 1992; Kaus & Wild, 1998). Similar symptoms have previously been observed in apple tree plantations and could by related to Mn toxicity (Zeiger & Shelton, 1972). Since Douglas fir is an introduced species, planted because of its high productivity, the selection of suitable provenances is important. We have shown that the capacity for Mn uptake is higher in DFG than in DFV, probably also because of its more extended root system (Table 1). This may render DFG more prone to Mn toxicity than DFV in acidic soils with high Mn availability. However, it should be taken into account that under field conditions additional factors such as biotic interactions are important in controlling nutrient supply to the plant. Douglas fir forms symbioses with both arbuscular and ectomycorrhizal fungi (Cazares & Trappe, 1993; Parlade et al., 1996), which modulate nutrient supply to the host. At least for agricultural plants (e.g. wheat and lettuce) it has been shown that arbuscular mycorrhizal symbioses diminished Mn uptake (Azcon et al., 2003; Ryan & Angus, 2003). To fully understand the consequences of the differences in Mn metabolism in the two Douglas fir varieties in different forest ecosystems, it will also be necessary to study the influence of symbiotic interactions on micronutrient supply.

Apart from additional effects of symbiotic fungi on Mn supply to the plant, uptake and transport of nutrient elements occur in the following steps: first from the rhizosphere into the root, then inside the root cortex towards the endodermis using apoplastic or symplastic pathways, loading into xylem, transport with the transpiration stream to the leaves and recirculation via the phloem. An intriguing question is therefore which of these steps was regulated in different ways in DFG and DFV. The steep gradient of radioactivity persisting in roots of DFV (Fig. 6) even after an extended feeding period, together with lower total plant 54Mn accumulation, indicate that uptake into the roots is more restricted in DFV than in DFG and that Mn once taken up is less mobile in DFV than in DFG. Analysis of Mn at the subcellular level of DFV roots supported this view since only epidermal walls but not cortical, endodermal or xylem walls contained elevated Mn concentrations (Fig. 3). Unlike DFV, high Mn concentrations in cortical walls as well as in the vacuoles of root cells indicated higher accessibility of Mn into roots of DFG.

In both varieties the endodermis appears to have important barrier functions since Mn transport to needles displayed a pronounced delay of almost 2 wk. Assuming that Mn is freely mobile in the xylem as in other species (Rengel, 2000) and considering the substantial consumption of nutrient solution during the experiment (Table 2), the delayed appearance of Mn in needles can only be explained by restricted xylem loading. Anatomical microanalysis indicated higher Mn concentrations in endodermal and parenchyma cell walls of DFV needles than in those of roots (P = 0.04, Fig. 3). This could be due to the fact that Mn once transported into the needles can hardly be remobilized and, thus, accumulated mainly in cell walls but also in vacuoles as shown here (Fig. 3). In mature conifers grown on calcareous soils with low Mn availability, Mn translocation also appears to be a slow process. On these sites, young needles displayed severe symptoms of Mn deficiency only in the first year after emergence. These symptoms disappeared after gradual accumulation of sufficient Mn over a time course of 2–3 yr (Kreutzer, 1972; Polle et al., 1992).

A puzzling question is how DFG avoided overaccumulation of Mn and maintained needle concentrations that were not excessively higher than those of DFV (Table 1), despite much higher uptake rates. Since DFG produced less above-ground biomass than DFV, ‘dilution’ by growth was not the reason. Another possibility is recirculation via the phloem. It is still a matter of discussion to which extent Mn can be recirculated in plants. In an earlier publication (Epstein, 1971) suggested that Mn is mobile in the phloem but that the degree of mobility varies with plant species. For example, Mn added by spraying directly to oat shoots resulted in translocation to roots over an extended period (Boken, 1960). In other studies also employing agricultural crops, it was concluded that the movement of Mn in the phloem occurs only in physiologically insignificant amounts (Mengel & Kirkby, 1982; Nable & Loneragan, 1984a,b; Loneragan, 1988). Pearson & Rengel (1995) analysed Mn and Zn distribution during grain development of wheat and observed a pattern suggesting that Mn was supplied via the xylem and Zn via the phloem. More recently, Rengel (2000) concluded that manganese mobility in the phloem sap depends on the type and age of the plant part. Our result that roots of Douglas fir seedling in the nonlabelled nutrient compartments contained low, but significant radioactivity indicates that phloem transport must have occurred but at slow rates (Figs 5, 6). Whether this transport is circulation in the strict sense of leaf-to-root transport (as defined by Marschner et al., 1997), is not clear since radial exchange of Mn between xylem and phloem in the stem cannot be excluded. The presence of radioactivity in compartment D, to which only nonlabelled nutrient solution was added and which contained the root tips of the split root system was small. Unspecific spill-over cannot be completely excluded. However, this would have been expected to result in similar degrees of contamination for both varieties. Higher 54Mn concentrations in DFG root tips in the nonlabelled compartment than in those of DFV suggest that the radioactivity was at least to some extent due to transport towards the root tips and exudation or root leaching. If circulation took place at similar rates as root-to-shoot transport, the time-course of the present experiment would have been too short to give conclusive data. Recently, Krüger et al. (2002) identified a protein in the phloem of castor bean, which bound preferentially Fe but also Mn and was involved in Fe transport in vivo. Since our data reveal high concentrations of Mn in the sieve cells (Fig. 3), it is tempting to speculate that Mn can be transported from the shoot back to the root by employing specific proteins.

Although the major topics of this study were Mn uptake and transport, it is also notable that both varieties of Douglas fir showed distinct differences in growth and biomass partitioning (Table 1). A yet unresolved question is whether the traits ‘root : shoot ratio’, respective ‘biomass’ and ‘Mn uptake’ are independent or linked. The latter option is not unlikely since root elongation and side-root formation are regulated via auxin metabolism involving Mn-stimulated auxin oxidases (Morgan et al., 1976). Recently, Arabidopsis mutants of the metal transporter ilr2-1 have been identified, which constitute a link between auxin metabolism and Mn homeostasis (Magidin et al., 2003). Whether a modification of a single micronutrient transporter gene can have such profound effects on biomass production and partitioning as found in the two Douglas fir varieties is currently speculative but deserves further attention as a potential functional marker for breeding purposes. Genetic structures at the marker gene locus 6-PGDH-A are unlikely to be functionally related to the physiological differences discussed above, but prove strong differentiation between the two varieties at the level of single genes.

Acknowledgements

We thank T. Riemekasten for maintenance of the plants and B. Kopka and G. Lehmann (Labor für Radioisotope, Universität Göttingen) for excellent technical assistance during the radioactive exposure experiments. We are grateful to Dr R. Heyser and Dr E. Fritz for advice and introduction to anatomy and X-ray electron dispersive transmission electron microscopy. Our special thanks go to the German Science Foundation for financial support to A. P.

Ancillary