Experimental site and plant material.

Seed from 12 Carya species native to the United States were collected from each of three wild mother trees growing in native habitats within each species’ natural range. Mother trees were generally separated by ≈100 m or greater. Seed were germinated and grown in pots for 2 years, after which two seedlings were randomly selected from each of three maternal parents of each species and planted in a common field plot established on a well-drained Norfolk loamy fine sand [Ng (fine-loamy, Kaolintic, thermic Typic Kandiudult soil; a common soil in the Southern Coastal Plain, Atlantic Coast Flatwoods, and Tidewater Area of the United States)] to give six replications of any one species. The planting site was at Byron, GA, a humid climatic zone (lat. 32°39′54″ N, long. 83°44′31″ W; 155 m elevation) with a frost-free growing period of ≈280 d and an annual precipitation of ≈130 cm. Test trees were grown at 10 × 10-m spacing without irrigation, fertilization, or pesticides (except for contact herbicides to control weeds). The site had not received fertilizers for ≈20 years before tree establishment. A single 2-year-old seedling of each species was randomly established within a block containing 12 Carya species with the pattern replicated to give six blocks (i.e., 72 trees). Trees were 22 years old at sampling.

Sampled Carya species represent both Apocarya and Carya sections of subtribe Caryinae of the tribe Juglandeae (Juglandaceae) (Manos and Stone, 2001) (Table 1). Species differ in not only morphology and habitat, but also in ploidy (i.e., 2n = 32 vs. 64) with Apocarya section species being diploid (100% of species) and Carya section species being either diploid (40% of taxa) or tetraploid (60% of taxa). Species are grouped by section based on morphology with those within the Carya section being shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], nutmeg hickory [C. myristiciformis (2n = 32)], and scrub hickory [C. floridana (2n = 64)]. Apocarya section species are pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)]. Species not included were mexican hickory [C. palmeri (found only in Mexico)] of the Apocarya section and red hickory (C. ovalis) of the Carya section. Other excluded species are the Asian section Sinocarya (C. cathayensis, C. dabieshanensis, C. hunanensis, C. kweichowensis, C. poilanei, and C. tonkinensis).

Table 1.

Seed source location of North American Carya species evaluated for their rare-earth metallome.

View Table

The hyperaccumulation and phytomining potential of pecan and other Carya species was assessed by comparing accumulated elements within foliar organs (i.e., leaves) from several non-Carya species growing nearby on the same soil type. These are slippery elm (Ulmus rubra), red maple (Acer rubrum), american persimmon (Diospyros virginiana), southern red oak (Quercus falcata), sassafras (Sassafras albidum), and sweetgum (Liquidambar styraciflua), thus representing different families (i.e., Ulmaceae, Ebenaceae, Aceraceae, Lauraceae, Rosaceae, and Fagaceae) than that of the Carya species (i.e., Juglandaceae).

Rare-earth element metallome analysis.

A snapshot of the foliar ionomes and metallomes was derived by randomly sampling mature foliage from 10 long vegetative shoots collected around the lower sun-exposed canopy of individual trees. Foliar sampling was from the middle leaflets of middle compound leaves of each sampled shoot as is the standard practice for leaf analysis sampling in commercial pecan orchards. Sampling was taken during early August within the 15 July to 15 Aug. window as recommended by extension specialists for tissue sampling to assess tree nutrient element health. Foliage from non-Carya species was sampled and processed as described for Carya species; midshoot simple leaves were sampled if the vegetative shoots did not possess compound leaves. Six trees of each non-pecan species were sampled from wild trees growing nearby on non-fertilized Norfolk loamy fine sand soil. To reduce the probability that detected REEs were surface contaminates from dust, sampled leaflets were combined for individual trees and rinsed with agitation 30 s in 2% nitric acid and triple rinsed with agitation in deionized water. Samples were then dried to a constant weight at 55 °C, ground in a Wiley mill, and 500 mg digested in 10 mL of 70% ultralow trace element grade nitric acid using a MarsXpress carousel in a Mars-5 microwave oven (CEM Corp., Matthews, NC). Microwave digestion was as follows: temperature was ramped from room temperature to 210 °C over 12 min and held at 210 °C for 13 min. Cooled digested samples were filtered, brought to 20 mL (by weight; using 2% nitric acid) in a 50-mL polypropylene centrifuge tube, and 0.10 to 1.00 mL (or 0.100 mL added to a 15-mL polypropylene disposable centrifuge tube and brought up to 14 mL using 2% ultrapure nitric acid made with greater than 18 MΩ deionized water.

Inductively coupled plasma quadropole spectrometer analysis.

The concentration of all REEs plus other elements similar to REEs (i.e., Ca, Mg, Mn, Fe, Cu, and Zn) in foliage was determined using an inductively coupled plasma quadrapole spectrometer SCIEX ELAN-9000; PerkinElmer, Concord, Ontario, Canada). Quantitative analysis was facilitated by various similar-mass internal standards (Li⁶, Ge⁷², In¹¹⁵, Ce¹³³, W¹⁸², or Bi²⁰⁹, depending on the elements being analyzed) and by external standards using multielement standard solutions (PerkinElmer Multielement Calibration Standard Sets) diluted to cover three to four orders of concentration magnitude.

Statistical analysis.

Metal analysis was based on two offspring from each of three mother trees of 12 Carya species. The experiment had a randomized complete block design with single tree pots for each of 12 species per block and six replicates (n = 72). Analysis of variance (P ≤ 0.50)] for species and ploidy treatments was determined for foliar elemental parameters. Elemental accumulation relationships were identified using multivariate pairwise correlation analysis. Concentration means for each species were analyzed through multivariate hierarchical clustering using the Ward's minimum variance method for defining distances between clusters to yield dendrograms illustrating species and relationships within ploidy classes.

Rare-earth metallomes.

The elemental composition of the foliar REE metallome of North American Carya did not qualitatively differ among species; however, quantitative differences were substantial with individual elemental concentrations differing severalfold (Tables 2 and 3). The REEs of highest concentration varied among species but were generally Ce > La > Nd > Y with all other REEs present at much lower concentrations (Tables 2 and 3; Fig. 1). Differences in the relative concentration of individual REEs are too numerous to discuss herein but are listed in Tables 2 and 3 for future baseline reference.

Table 2.

Relative prominence of the light rare-earths in foliage of several North American Carya species.

View Table

Table 3.

Relative prominence of the heavy rare-earths (REEs) in foliage of several North American Carya species.

View Table

Relative abundance of the rare-earth element composition of the foliar ionome of pecan and diploid (including pecan) vs. tetraploid Carya species. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)]. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

View Full Size

Relative abundance of the rare-earth element composition of the foliar ionome of pecan and diploid (including pecan) vs. tetraploid Carya species. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)]. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

Relative abundance of the rare-earth element composition of the foliar ionome of pecan and diploid (including pecan) vs. tetraploid Carya species. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)]. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

It is noteworthy that C. tomentosa contains relatively high concentrations of Ce (326 μg·g⁻¹), La (144 μg·g⁻¹), Nd (136 μg·g⁻¹), and total REEs (859 μg·g⁻¹). The concentrations of these three dominating REEs were 83%, 45%, and 37%, respectively, of that found earlier by Thomas (1975). Additionally, these three Lns comprise 71% of all REEs detected, which compares with 80% found earlier in C. tomentosa. Much of this discrepancy is likely the result of differences in soil REE composition and bioavailability and tree genotype. By comparison, the REE concentration of C. aquatica was especially low at 84 μg·g⁻¹ or approximately one-tenth that of C. tomentosa. The relative concentration of REEs in pecan, a diploid, is La > Ce > Nd > Y > Gd > Pr > Sm > Dy > Er > Yb > Ho > Tb > Tm > Sc > Lu. Several REEs (Y, La, Ce, Nd) are present in foliage at approximately the same concentration as are Fe, Zn, and B; and others (Pr, Sm, Gd, Dy) are present at approximately the same concentration as Cu and still others (Tb, Eu, Ho, Er, Tm, Yb, Lu) approximately the same concentration as Ni and Mo. When viewed collectively, the concentration of all REEs in pecan is ≈190 μg·g⁻¹ dry weight, which is similar to that of Mn in many orchard trees and ≈20% as the critical level of phosphorus and sulfur. Total REE concentration in pecan is only ≈4% to 13% as high as that of Ca and Mg, respectively, when these macronutrients are near the concentration that triggers visual deficiency symptoms.

Accumulation of the individual lanthanide constituents of the Carya foliar REE metallome is highly correlated [correlation coefficient (r)] among constituent members (data not included) with r ranging from 0.93–0.99 (P ≥ 0.001) for all Ln pairs, except for Ce, which ranged from 0.77–0.94 (P ≥ 0.001) depending on the Ln. This is true for pecan as well as for diploid and tetraploid Carya. Such tight linkage suggests that Ln uptake mechanism and transporters in Carya taxa possess little or no specificity for the individual Lns, except with the possible exception of Ce. This might be because Ce typically exists as a tetravalent cation in soil solutions, whereas the other REEs typically exist as trivalent cations.

The accumulation of REEs in Carya is such that those with even atomic numbers (Z) are present at greater concentration than that of the two adjacent elements with odd Z (Fig. 1). This pattern is consistent with that found by Watanabe et al. (2007) for Ln composition of several orders of plants. Thus, relative REE accumulation patterns in Carya are consistent with that observed for other plant species being consistent with the Oddo-Harkins rule, which holds that elements with an even Z are more common than elements with an odd Z. This is because elements with an even proton number are balanced in proton spin and less likely to capture an additional proton during formative processes than odd-numbered elements possessing an unpaired proton.

When tetraploid and diploid Carya classes are compared against several non-Carya (from non-Juglandaceae) species, it is apparent that both Carya ploidy classes accumulate REEs at much greater concentrations than do many other species when grown under similar conditions (Table 4). Tetraploid Carya as a group possessed much higher REE concentrations than did diploid Carya but with relative concentrations among REEs being equivalent for taxa within each of the two ploidy levels. Accumulation of G-II REEs, LREEs, IREEs and HREEs, Lns, and total REEs by tetraploids was ≈1.7- to 1.8-fold greater than that by diploids and ≈32- to 38-fold greater than by non-Carya species (Table 4). In terms of specific REE elements, tetraploids accumulated more Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu than diploids; however, diploids accumulated more Sc, the lightest of the REEs, than tetraploids. Both tetraploids and diploids accumulated all REEs at a greater concentration than did non-Carya species found growing in the same or similar habitats.

Table 4.

Comparison of rare-earth (REE) metal composition of the foliar ionome of diploid (2n = 32) and tetraploid (2n = 64) Carya species and non-Carya species.^z

View Table

If REEs act as analogs to certain essential divalent metals [M²⁺ (e.g., Ca, Mg, Fe, Cu, and Zn)] of the Carya metabolome, then the M²⁺:REE ratio likely varies among Carya species and relatively strong correlations with accumulation of these other metals. The M²⁺:REE ratio of key essential metals within the Carya metabolome indeed varies depending whether the foliar ionome originated from tetraploid or diploid Carya or from non-Carya species. In the case of essential nutrient elements that function in planta as hard Lewis acids (i.e., Mg, Ca, and Zn) these ratios were very high for non-Carya species (Table 5) but relatively low for diploid Carya and especially low for tetraploid Carya (i.e., Mg/G-II REEs, Mg/LREEs, Mg/IREEs, Mg/HREEs, Mg/Total REEs, Ca/G-II REES, Ca/LREEs, Ca/IREEs, Ca/HREEs, Ca/HREEs, Ca/Total REEs, Zn/G-II REEs, Zn/LREEs, Zn/IREES, and Zn/Total REEs). In the case of the intermediate strength Lewis acids (e.g., Mn, Fe, and Cu), there were no REE ratio differences between tetraploid and diploid Carya; however, almost all ratios were much greater for non-Carya species than for either Carya ploidy class (Table 5). These REE ratio differences tend to support the possibility that one or more REEs might act in planta because limited substitutes for certain processes usually requiring Lewis acids (Mg, Ca, Zn, Mn, Fe, Cu, Co, Ni).

Table 5.

Ratios of different classes of rare-earth elements (REEs) in the foliar ionome of diploid and tetraploid Carya and in non-Carya species to that of chemically similar metabolome metals.^z

View Table

It is noteworthy that for diploid Carya, the accumulated concentration of Mn within the foliar ionome exhibits substantial positive correlation (r = 0.59–0.72; data not shown) with accumulated concentration of elements comprising the various REE classes, whereas there were negative associations with Cu. For the individual REEs, Mn was positively correlated with all REEs (r = 0.43–0.72) as was Co (r = 0.40–0.57), but Cu was negatively correlated (r = –0.45–0.54). For tetraploid Carya, accumulation of Mg (r = 0.24–0.40), Ca (r = 0.24–0.41), and Fe (r = 0.59–0.67) in the foliar ionomes is positively associated with the concentration of various REE classes. The accumulation of Mn in pecan, and diploid Carya, is positively correlated with each of the 14 individual Lns (r = 0.62–0.72). Also, there was no significant relationship between Mn, Cu, Co, Ni, Zn, and Mo and any specific Ln; however, Fe accumulation is positively associated (r = 0.56–0.65) with all Lns. There is also a weak positive correlation between Mg and La (r = 0.45), Ca and La (r = 0.42), Mg and Ce (r = 0.39), and Ca and Ce (r = 0.41) for tetraploid Carya. Thus, the accumulation of all REEs in diploid Carya is positively correlated with Mn and Co but negatively associated with Cu, and there was no correlation of Mg or Ca with any REE. Conversely, in tetraploid Carya, Fe is the only micronutrient correlated with the individual REEs; however, the macronutrients Mg and Ca are positively correlated with La and Ce.

Cluster analysis of foliar rare-earth metallome.

Hierarchal cluster analysis of the multivariate components of the REE metallome of the foliar Carya ionomes, when the variance within a cluster is minimized (i.e., Ward's method), produced dendrograms with relatively long stems and small distances within the far branches of the trees, hence identifying likely discrete natural grouping within each of the two ploidy classes (Fig. 2). The ses were small in each dendogram but slightly larger for the tetraploid population than that of the diploid population, indicating what might be greater variability within the REE metallome components of tetraploid than diploid taxa. The clustering of C. aquatica through chaining to the C. illinoinensis/myristiciformis cluster, while possessing a short stem-associated distance metric, indicates that although the REE metallome of C. aquatica is distinct, it possesses substantial similarity to that of the other two species. Similar close relationships exist between C. carolina-septentrionalis (of section Carya) and C. cordiformis (of section Apocarya). The REE metallomes of C. ovata and C. laciniosa are also very similar to each other and also similar to C. carolina-septentrionalis and C. cordiformis. These four species possess REE metallomes more similar to each other than to the other three diploid taxa (i.e., C. myristiciformis, C. illinoinensis, and C. aquatica).

Dendogram derived from hierarchical clustering of diploid (2n = 32) Carya species based on the rare-earth metallome of foliage. Scaling reflects a relative distance scale. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

View Full Size

Dendogram derived from hierarchical clustering of diploid (2n = 32) Carya species based on the rare-earth metallome of foliage. Scaling reflects a relative distance scale. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

Dendogram derived from hierarchical clustering of diploid (2n = 32) Carya species based on the rare-earth metallome of foliage. Scaling reflects a relative distance scale. Diploid species are shellbark hickory [C. laciniosa (2n = 32)], southern shagbark hickory [C. carolinae-septentrionalis (2n = 32)], shagbark hickory [C. ovata (2n = 32)], nutmeg hickory [C. myristiciformis (2n = 32)], pecan [C. illinoinensis (2n = 32)], water hickory [C. aquatica (2n = 32)], and bitternut hickory [C. cordiformis (2n = 32)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

The REEs metallomes of the five tetraploid taxa clearly differentiates into two distinct groups with C. tomentosa and C. glabra being most similar of the tetraploids (Fig. 3). Carya floridana and C. pallida are also very similar to each other and much less similar to the REEs metallome of C. texana, which clusters with these two species through chaining and at considerable distance. Thus, the REE metallomes of C. tomentosa and C. glabra are distinctly different from that of the other three tetraploid taxa.

Dendogram derived from hierarchical clustering of tetraploid (2n = 64) Carya species based on the rare-earth of foliage. Scaling reflects a relative distance scale. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

View Full Size

Dendogram derived from hierarchical clustering of tetraploid (2n = 64) Carya species based on the rare-earth of foliage. Scaling reflects a relative distance scale. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

Dendogram derived from hierarchical clustering of tetraploid (2n = 64) Carya species based on the rare-earth of foliage. Scaling reflects a relative distance scale. Tetraploid species are sand hickory [C. pallida (2n = 64)], mockernut hickory [C. tomentosa (2n = 64)], pignut hickory [C. glabra (2n = 64)], black hickory [C. texana (2n = 64)], and scrub hickory [C. floridana (2n = 64)].

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 136, 6; 10.21273/JASHS.136.6.389

• Download Figure

Potential of a physiological role for rare-earth element in Carya species.

Brown et al. (1990) reviewed possible generalized physiological roles for REEs and concluded they could potentially influence membrane stability and ion transport (especially Ca²⁺, K⁺, and Na⁺), nitrogen fixation, disease resistance, hormonal interactions, sensing, and enzyme activity. The hyperaccumulation documented here provides circumstantial evidence that one or more REEs might theoretically benefit Carya metabolism or physiology. However, hyperaccumulation does not necessarily mean that REEs play a biological role. It is possible that the root system of Carya species, especially that of tetraploids, fails to bind up excessive REEs as effectively as that of diploids or of the roots of non-REE accumulating non-Carya species. If one or more REEs possess biological relevance in Carya, then the ion must pass through cell membranes and enter symplastic space. The present study does not discriminate whether REEs are in symplastic or apoplastic space or are bound to cell walls, accumulated in vacuoles, or in free cytoplasmic space. Future research needs to identify their locations if the metabolic or physiological relevance is to be understood.

Many metalloenzymes, especially hydrolytic enzymes, require hard Lewis acids (e.g., Ca, Mg, and Zn) for activation of key elemental bonds requiring cleavage. Because the REEs are hard Lewis acids, have high coordination numbers, exhibit fast ligand exchange rate characteristics, and have relatively inaccessible redox chemistry, they are well suited as potential cofactors in certain metabolically relevant enzymes (Franklin, 2001). They appear to be especially suitable for replacing Ca and certain other alkali-earth and outer transition-metal elements within various enzymatic proteins or for Ca in intracellular cytosolic signaling (Bulman, 2003; White and Broadley, 2003). Lanthanide ions are near ideal Ca analogs and can substitute for Ca in proteins under artificial conditions with minimal structural changes, but binding is likely to be less reversible than Ca²⁺ (Evans, 1983). Thus, if one or more REEs replace Ca at membrane sites, or pass through cellular membranes, then the REE can potentially engage in a host of critical metabolic processes. This raises the possibility of both competitive inhibition of Ca-dependent processes and the existence of Ln-dependent metalloenzyme isoforms and Ca²⁺-like signaling processes in Carya. REEs could also act as Ca²⁺ surrogates to enhance crosslinking of membranes, thus stabilizing membranes under acidic soil conditions, where Ca might be limiting, and there is a need to stabilize ion transmembrane transport (White and Broadley, 2003). Additionally, because Lns have high binding affinities to the plasmalemma and other intracellular membranes, they potentially influence ion channels and therefore signaling and gene expression (Hagenbeek et al., 2000; Rock and Quatrano, 1996).

If REEs act as Ca (or Mg, Fe, Cu, Zn) analogs to the essential Lewis acid metals of the Carya metabolome, then the M²⁺:REEs likely varies among Carya species, and there is likely to be strong correlations in element accumulation. The ratio of key essential elements within the Carya metabolome to the various classes of REEs indeed varied depending whether the foliar ionome originated from tetraploids or diploids or from non-Carya species. These ratios were very high for non-Carya species but much lower for diploid Carya and especially low for tetraploids (i.e., Mg/G-II REEs, Mg/LREEs, Mg/IREEs, Mg/HREEs, Mg/Total REEs, Ca/G-II REES, Ca/LREEs, Ca/IREEs, and Ca/HREEs. Ca/HREEs, Ca/Total REEs, Zn/G-II REEs, Zn/LREEs, Zn/IREES, and Zn/Total REEs). In the case of certain intermediate-strength Lewis acids (e.g., Mn, Fe, and Cu), there were no M²⁺:REE differences between tetraploids and diploids; however, almost all ratios were much greater for non-Carya species than for either Carya ploidy class. These ratio differences suggest that one or more REEs might substitute for Mg, Ca, and Zn in tree processes requiring hard Lewis acids and also possibly for Mn, Fe, Co, and Cu in one or more specific processes where an intermediate-strength Lewis acid-requiring process might be satisfactorily substituted by a hard Lewis acid. These possibilities are supported in that REE ionic radii are very similar to that of Ca and several other essential divalent cations depending on the coordination environment in which ions are exposed (Barbalace, 2011; Grimes and Wenman, 2011).

Calcium homeostasis is critical to the effectiveness of stimulus-response systems operating in planta (Sanders et al., 1999). If certain REEs are replacing Ca, or possess similar signaling roles, then there would be need for a similar tight regulation of cytosolic REE concentration. Thus, a measure of cytosolic REE concentration would provide evidence pertaining to whether REEs are likely to be involved in “signaling” with concentrations similar to the of Ca²⁺ [(100–200 nM (Nayyar, 2003)] being evidence of activity.

The Lns can replace Mg and Zn in certain proteins and enzymes, and Gd and La act as Ca analogs in Ca-channel studies (Lim and Franklin, 2004), and they also bind with Ca-dependent protein kinases involved in key cellular processes (Plieth, 2005). It seems likely that elements nearly identical to Ca (i.e., the Lns) are involved in life processes, yet no critical roles for Lns, or other REEs, are known in any plant species. This is surprising given that Lns are so similar in ion size, coordination environment, ligand preferences, and Lewis acid activity (giving good hydrolytic activity). Many metals are ideally suited for specific biochemical functions; thus, they are commonly used in life processes. Had the Lns been as abundantly available early on, and over evolutionary time, as what are today the essential hard Lewis acid elements (e.g., Ca²⁺, Mg²⁺, and Zn²⁺), then the Lns might today be obviously essential to life processes (e.g., Eu³⁺ replacing many Ca²⁺ functions) in most species (Franklin, 2001). Now that the Lns are relatively available for biological functions, it may be that as plants continue to evolve while exposed to increasingly REE-rich topsoil environments, they are adapting to use certain REEs for specific metabolic and physiological processes, hence the hyperaccumulation in Carya, and especially in tetraploid Carya. Observations that certain REEs can increase photosynthesis, chlorophyll content, and nitrogen metabolism (Cao et al., 2007; Weiping et al., 2003; Xue and Yang, 2005; Yin et al., 2009), replace Ca²⁺ in stomatal regulation (Xue and Yang, 2005), affect abscissic acid (ABA)-induced gene expression (Dagenbeek et al., 2000; Rock and Quatrano, 1996), replace Ca in photosystem-II (Hong et al., 2002a, 2002b; Wang et al., 2003), replace Mg in the porphryin ring of chlorophyll (Hong et al., 2002a; Wei et al., 2005; Weiping et al., 2003), increase activity of nitrate reductase (Song et al., 2003) and glutamine synthase and glutamate dehydrogenase (Song et al., 2003; Wan et al., 2004), and improve plant growth (Boyko et al., 2011) of certain species is evidence that REEs might be beginning to be incorporated into the biology of certain species. The relatively high tolerance of most species to REEs might also be evidence of active positive biological roles. There is also evidence REEs benefit certain crop species in agricultural settings (Tyler, 2004); hence, they are widely used in Chinese agriculture with anticipation of positive effects on crops (Pang et al., 2002; Xu et al., 2002). Conversely, in U.S. and other non-Asian agricultural systems, REEs receive little or no attention, especially within the context of the physiology or horticulture of woody polycarpic perennial angiosperm type crops, although they are relatively common in many agricultural soils [i.e., Y at 16–33 mg·kg⁻¹, LREEs ≈76–448 mg·kg⁻¹, and HREEs of ≈9–55 mg·kg⁻¹ and are commonly found to be high in many manures added to soils (Tyler, 2004)].

REEs are routinely applied to agricultural soils as impurities in phosphate fertilizers, but plant access to REEs in the soil solution is minimal as a result of their rapid precipitation with phosphate (Diatloff et al., 1993; Laveuf and Cornu, 2009). Thus, if biologically relevant, reduced bioavailability of REEs under high soil P or leaf P conditions and dry conditions might be expected to limit crop yield and/or quality. The selective accumulation of REEs, especially by tetraploids, identifies these taxa as a phenomenon of potential physiological and ecological uniqueness and raises questions regarding REEs relevance of to Carya nutritional physiology. The recent discovery that pecan possesses a relatively high Ni requirement during early spring (Wood et al., 2004), and Ni's associated importance in nitrogen, organic acid, and fatty acid metabolism (Bai et al., 2006, 2007) has led to modified nutrient management strategies that include Ni application to trees in commercial orchards.

These findings raise the possibility that one or more REEs confer positive horticultural benefit to pecan orchard enterprises, especially considering that pecan likely has considerably more exposure to biologically available REEs with the rich organic matter topsoils in its native habitats than in commercial orchards. Because absolute REEs concentrations are typically greater in alkaline soils such as those commonly found in the arid southwestern United States than the highly weathered acidic soils of the southeastern United States, it might be that the typically much higher orchard nutmeat yields (approximately twofold greater) in a low P arid soil environments is partially linked to better tree REE nutrition. This possibility merits further investigation.

It is possible that one or more REEs stimulate certain in planta processes as a result of a “hormesis effect” [i.e., a process in which exposure to a low dose of a chemical agent or environmental factor that is damaging at higher doses induces an adaptive beneficial effect on the cell or organism (Mattson, 2008)], where there is a biphasic dose–response in which a low dose stimulates positive outcomes, whereas a high dose is inhibitory or toxic. Examples of positive hormetic outcomes include protection of cells or organisms against environmental stresses (e.g., drought, high temperature, toxic metals, and oxidative inducers), thus benefiting cell or organism health and survival fitness. Over evolutionary time, organisms can incorporate hormetic agents into cellular processes to the point that these elements become either beneficial or essential (Mattson, 2008).

It is noteworthy that nuclear detonations and certain types of malfunctioning nuclear reactors can introduce certain radioactive REEs into the environment. This increased exposure to radioactive REEs might present inordinate risks to REE hyperaccumulating Carya species such as pecan and therefore adversely affect subsequent seed germination and seedling survival as well as the health of animals and humans consuming the nutmeats (Thomas, 1975).

Potential ecological role of rare-earth elements in Carya species and speciation.

If substitution by REEs occurs in certain metalloenzymes or proteins, then there is potential for existence of enzyme isoforms that benefit species adaptation. REE accumulation might confer a degree of metabolic and ecological flexibility absent in many non-Carya species; thus, certain REEs might function as beneficial, or even essential, nutrient elements for trees exposed to certain environmental stresses. This could theoretically be a factor contributing to speciation of tetraploid Carya and their successful colonization of specific habitats not well occupied by diploid Carya. Such is not only relevant to the evolutionary biology and ecophysiology of Carya, but also the horticulture of pecan and other Carya.

Based on distance measures of hierarchal cluster analysis, diploid species possessing the most similar REE metallomes are C. illinoinensis and C. myristiciformis. Both species can grow in habitats that are much drier, on upper terraces and low ridges, during portions of the growing season than habitats where C. aquatica thrives; hence, any putative metabolic advantage conferred by one or more REEs might well be operating in both species’ ability to adapt to drier environments. This grouping pattern is also interesting in that taxa classification based on morphology group C. myristiciformis with section Carya and C. illinoinensis and C. aquatica with section Apocarya. Although C. myristiciformis substantially differs from the other two morphologically, it retains substantial similarity in its REE metallome. The similarity between C. ovata and C. septentrionalis is also not surprising in that both are morphologically similar and considered by some to be biotypes of the same species; however, possession of distinct quantitative different in their REE metallomes indicates that they are at least distinct biotypes and may well be different species after all.

In the case of tetraploid taxa, the REE metallomes of C. tomentosa and C. glabra are distinctly different from that of the other three tetraploids with both species occupying seasonally xeric ridges and hills within their natural habitats. Similarly, C. floridana and C. pallida also occupy similar seasonally xeric habitats, but their native ranges do not overlap. The native range of C. texana does not overlap that of either C. pallida or C. floridana and possesses greater dissimilarities in REE metallome composition. The especially high accumulation of REEs by tetraploid species hint that these elements might play important fitness-associated biological roles. Although REE accumulation may simply be an artifact without selective advantage, it is also possible that one or more of the REEs are “beneficial” to Carya species or may even be “essential.”

Several primitive plant taxa accumulate REEs, regardless of soil REE content, compared with more evolutionarily advanced cycadales, conifers, and angiosperms (monocots and dicots). Examples include many ferns (Polypodiophytina), horsetails (Equisetophytina), and clubmossess (Lycopodiophytina) (Ozaki et al., 2000; Watanabe et al., 2007). Wyttenbach et al. (1996, 1998) noted that REE concentration in most plants is in the parts per billion range with certain fern species accumulating La and Nd as high as ≈635 μg·g⁻¹ dry weight. The present study not only confirms previous reports that C. tomentosa accumulates REEs (Robinson, 1943; Robinson and Edgington, 1945; Robinson et al., 1938, 1958; Thomas, 1975), but that almost all indigenous U.S. species are significant REE accumulators.

Topsoils, especially the humus zone, are generally enriched with REEs through plant recycling with organic matter possessing an especially high capacity to complex, adsorb, or chelate positively charged REEs. The organic matter in topsoils is usually enriched more, in a relative sense, with HREEs and MREEs than LREEs (Laveuf and Cornu, 2009); thus, soils high in organic matter such as those in which most wild Carya species grow in their native habitats potentially have positive HREE and MREE anomalies (Thomas, 1975). Decomposing leaf litter can lead to a center of REE concentration that tends to perpetuate itself, exemplifying the role of Carya taxa as accumulators influencing distribution of elements within their ecosystems. The hickories are centers of REE concentration and therefore reduce losses from the tree-soil system where they are major components. Under soil conditions with relatively high REE availability, it is possible that over evolutionary time Carya species have integrated certain REEs into life processes to such a degree that they are at least beneficial, if not essential. This merits further study, especially within the context of pecan horticulture.

It is noteworthy that the natural habitats of all tetraploid North American Carya are substantially more xeric during portions of the growing season than the relatively mesic habitats of diploid Carya. This suggests that REEs might confer a survival advantage linked to water relations. Nutrient availability from dry soils is minimal; thus, REEs might benefit specific xeric-associated metabolic processes when other nutrients are insufficiently available or, perhaps more likely, benefit water relations and drought tolerance. Additional evidence for a REE role under xeric conditions is their apparent ability to replace Ca²⁺ in stomatal regulation (Xue and Yang, 2005) and elicitor activity for ABA-associated gene expression (Rock and Quatrano, 1996).

Although plant reliance on a particular element requires substantial evolutionary time, there has nevertheless been sufficient time to enable enough evolutionary diversification for integration of one or more REEs into the metabolome of at least a few modern species (Wyttenbach et al., 1998). REE accumulator species such as tetraploid Carya might well be the vanguard taxa for such a role.