- Top of page
- Materials and Methods
Hyperaccumulating plants are defined by the extremely high concentration of metals sequestered within their tissues (Brooks et al., 1977). The threshold metal concentration used to define a hyperaccumulator depends on the particular metal sequestered. For nickel (Ni) hyperaccumulators, plants must contain 1000 µg g−1 dry mass or greater (Brooks et al., 1977). At least 318 taxa hyperaccumulate Ni (Reeves & Baker, 2000).
Several functions of metal hyperaccumulation have been proposed (Boyd & Martens, 1992), including plant defense (Boyd & Martens, 1992; Boyd, 1998). A plant defense increases resistance against the attack of herbivores (natural enemies) (Levin, 1976; Mauricio & Rausher, 1997), where resistance is defined as a plant characteristic that reduces damage inflicted by an herbivore (Rausher, 1992). Some elements sequestered by plants and algae may have defensive functions. These include silicon (Si) (McNaughton & Tarrants, 1983), fluorine (F) (Twigg & King, 1991), calcium (Ca) (Hay et al., 1994), cadmium (Cd) (Jiang et al., 2005), Ni (Boyd & Martens, 1994), zinc (Zn) (Pollard & Baker, 1997; Jhee et al., 1999; Behmer et al., 2005) and selenium (Se) (Hanson et al., 2003, 2004). Termed ‘elemental’ chemical defenses by Martens & Boyd (1994), these defenses differ from organic chemicals because they are elements taken up from soil and sequestered in tissues rather than being produced from photosynthate. Martens & Boyd (1994) also pointed out that, unlike many organic chemicals, toxic elements cannot be degraded into less toxic components, thus eliminating one possible herbivore counterdefense tactic.
Plants are consumed by a diverse array of herbivores, and plant–herbivore interactions can be influenced by herbivore feeding mode (Strauss, 1991; Karban & Baldwin, 1997; Gavloski & Lamb, 2000). Studies that have compared damage caused by herbivores representing different feeding modes on the same plant species have often found differing impacts (e.g. Moran & Whitham, 1990; Strauss, 1991; Meyer, 1993; Gavloski & Lamb, 2000). For example, Moran & Whitham (1990) found plant biomass and seed set of Chenopodium album (Chenopodiaceae) were reduced by a leaf-gall forming aphid but unaffected by a root-feeding aphid. Similarly, Meyer (1993) found that the relative growth rate of Solidago altissima (Asteraceae) was decreased to the greatest extent by a xylem-feeding spittlebug, less by a leaf-chewing beetle and not at all by a phloem-feeding aphid. Defenses can also vary in their effectiveness against herbivores of different feeding modes. For example, Karban & Nagasaka (2004) found that defenses induced by damage to Raphanus sativus (Brassicaceae) increased plant resistance to leaf-chewing folivores (caterpillars) but decreased resistance to phloem-feeding aphids.
It is unlikely that elemental defenses will provide complete protection against all herbivores because some plant enemies circumvent every plant defense (Gatehouse, 2002; Karban & Agrawal, 2002). Therefore, studies that use herbivores representing various feeding modes are needed to establish the boundaries of elemental defenses. Most studies of defense by hyperaccumulated metals have used folivores and almost all have found defensive effects (e.g. Boyd & Martens, 1994; Pollard & Baker, 1997; Boyd & Moar, 1999; Jhee et al., 1999; but see the exception of Huitson & Macnair, 2003). Only two studies have examined the defensive function of hyperaccumulation using nonfolivore herbivores, with contrasting results. Boyd & Martens (1999) examined the effect of hyperaccumulated Ni by Streptanthus polygaloides (Brassicaceae) on the pea aphid Acyrthosiphon pisum (Homoptera: Aphididae), and Hanson et al. (2004) studied the effect of hyperaccumulated Se on the green peach aphid Myzus persicae (Homoptera: Aphididae). Boyd & Martens (1999) reported that Ni hyperaccumulation was ineffective as a plant defense against the pea aphid, but Hanson et al. (2004) found that Se hyperaccumulation was toxic to the green peach aphid.
No study has addressed the effect of a metal hyperaccumulated by a single plant species on arthropods representing a variety of feeding modes. In this study we examined the potential defensive role of hyperaccumulated Ni by S. polygaloides on arthropods of diverse feeding modes including folivores, a rhizovore, xylem- and phloem-feeding insects, and two cell-disrupting arthropods. Our study included eight arthropod species representing four feeding modes. Folivores used were the red-legged grasshopper Melanoplus femurrubrum (Orthoptera: Acrididae) and the cross-striped cabbageworm Evergestis rimosalis (Lepidoptera: Pyralidae). We used the cabbage maggot (cabbage root fly) Delia radicum (Diptera: Anthomyiidae) as a rhizovore. Vascular tissue-feeding insects were represented by the xylem-feeding meadow spittlebug Philaenus spumarius (Homoptera: Cercopidae) and two phloem-feeding insects, the turnip aphid Lipaphis erysimi (Homoptera: Aphidae) and the greenhouse whitefly Trialeurodes vaporariorum (Homoptera: Aleyrodidae). The fourth feeding mode was cell disruption, represented by the tarnished plant bug Lygus lineolaris (Heteroptera: Miridae) and two-spotted spidermite Tetranychus urticae (Acarina: Tetranychidae). To date, no study of hyperaccumulator defense has examined the effect of metals on dipterans or mites, so our choices allowed us to examine examples from previously unexplored taxonomic groups. Our inclusion of xylem-feeding and rhizovore insects also enabled us to examine these feeding modes for the first time in a study of hyperaccumulator plant herbivory. None of these arthropods has been reported as feeding on S. polygaloides in its natural habitat (Wall & Boyd, 2002).
Our experimental design involved a two-step approach. We first tested each arthropod species for differential survival when fed either high- or low-Ni plant material. These no-choice experiments allowed us to investigate elemental defense by examining herbivore response (survival and population growth) to plant metal concentration. If a defensive effect of Ni was observed in a no-choice experiment, we proceeded to a choice experiment in which herbivores were presented with both high- and low-Ni plant material. Choice experiments allowed us to determine if deterrence or selective feeding occurred, resulting in decreased damage to high-Ni plants. Decreased damage to high-Ni plants relative to low-Ni plants may lead to differential plant fitness and in turn could result in evolution of resistance traits (defenses) within a plant population (Boyd, 2004).
- Top of page
- Materials and Methods
Our experiments tested the elemental defense hypothesis with herbivores representing various feeding modes. Because no-choice experiments showed significantly lower survival or herbivore population growth for cabbage maggots, cross-striped cabbageworms, grasshoppers and two-spotted spidermites on high-Ni plant tissues, we conclude that hyperaccumulated Ni defends S. polygaloides against these herbivores (Walling, 2000). Choice experiments showed that Ni hyperaccumulation in S. polygaloides confers herbivore resistance upon high-Ni S. polygaloides through deterrence. Red-legged grasshoppers fed to a significantly greater extent upon low-Ni plants of S. polygaloides than upon high-Ni plants when presented with a choice of the two plant types. Similarly, spidermites had significantly greater population growth on low-Ni plants compared with high-Ni plants when presented with a choice of the two plant types. We also found complete deterrence for cabbage maggot and cross-striped cabbageworm larvae. No cabbage maggots were found in high-Ni roots when a choice was presented, thus suggesting that larvae discriminated strongly between high- and low-Ni roots. Cross-striped cabbageworms were observed to rest on or crawl over high-Ni leaves but we found no visible damage to those leaves. A complete deterrent effect of hyperaccumulated metals is rare. To our knowledge, only Pollard & Baker (1997) have reported it before, for P. rapae given a choice of Zn hyperaccumulating or nonhyperaccumulating Thlaspi caerulescens (Brassicaceae).
As postulated by Boyd (2004), herbivores with different feeding modes may respond differently to elemental defenses. The present study, as well as our previous work (e.g. Boyd & Martens, 1994; Martens & Boyd, 1994; Boyd & Moar, 1999; Boyd et al., 2002), has shown that leaf-chewing folivores (in the present study, red-legged grasshoppers and cross-striped cabbageworms) are especially susceptible to hyperaccumulated Ni. Although the location of Ni in tissues of S. polygaloides is unexplored, other Ni hyperaccumulators store Ni within epidermal and subepidermal tissues (Mesjasz-Przybylowicz et al., 1996; Krämer et al., 1997; Kupper et al., 2001) and that Ni may be complexed with organic acids such as citrate or malate and sequestered within cell vacuoles (Reeves, 1992; Anderson et al., 1997; Krämer et al., 2000; Salt & Krämer, 2000). Leaf-chewing grasshoppers and caterpillars in the present study may have been negatively affected by Ni hyperaccumulated in S. polygaloides within subepidermal vacuoles. As tissue-chewing herbivores ruptured those vacuoles the Ni would be liberated and could then be toxic.
We found that high levels of Ni did not affect phloem-feeding herbivores. Experiments with turnip aphids and greenhouse whiteflies (Table 3) showed no defensive effect of hyperaccumulated Ni. Previous experiments with the pea aphid (Acyrthosiphon pisum) found no effect of hyperaccumulated Ni in S. polygaloides (Boyd & Martens, 1999). Boyd & Martens (1999) suggested that phloem-feeding insects may be unaffected because phloem tissue of S. polygaloides is relatively low in Ni. While xylem sap of Ni hyperaccumulators may contain Ni bound to amino acids (Krämer et al., 1997), little information exists regarding the Ni concentration of phloem fluid. Our conclusion that hyperaccumulated Ni is ineffective against phloem-feeding herbivores may be specific to this element, however, as Hanson et al. (2004) showed that Se-hyperaccumulating Brassica juncea (Brassicaceae) was both toxic and deterrent to aphids.
Another vascular-feeding insect unaffected by hyperaccumulated Ni was the meadow spittlebug. Hyperaccumulated Ni may be complexed with amino acids within xylem sap (Krämer et al., 1997). Using 63Ni, Anderson et al. (1997) demonstrated that Ni absorbed by roots in the South African hyperaccumulator Berkheya coddii (Asteraceae) travels through the xylem sap. Spittlebug nymphs produce a mass of bubbles during their instar development, which may aid in water retention and protection from predators (Hamilton & Morales, 1992). Filter paper saturated with DMG has been used as a colorimetric assay for hyperaccumulated Ni in plant tissues based on a pink color change (Reeves et al., 1999). In our experiments, nymphs feeding on high-Ni plants produced spittle that yielded a positive (pink) reaction when tested with DMG paper. Spittle produced by nymphs is excreted from a filtering chamber where some substances may pass through the gut unchanged or unassimilated (Ponder et al., 2002). The complex excretory system of spittlebugs may allow them to tolerate elevated dietary Ni, as it was clear from the DMG test that Ni was present in the xylem sap consumed by spittlebugs.
We also found no significant effect of hyperaccumulated Ni on the cell-disrupting tarnished plant bug Lygus lineolaris. A field survey of arthropods associated with S. polygaloides (Wall, 1999; Wall & Boyd, 2002) reported several hemipteran herbivores, including Lygus hesperus, feeding on this species. These L. hesperus contained approx. 31 µg Ni g−1 dry mass (Wall & Boyd, 2002) and fed on high-Ni plants with no apparent adverse effects. We suggest that L. lineolaris may have been able to feed on hyperaccumulating S. polygaloides and tolerate high plant Ni concentrations in the same (unknown) manner as L. hesperus in the field. Although the mechanism by which they can feed on high-Ni plants is unknown, we suggest that tarnished plant bugs are relatively tolerant of Ni, as it seems unlikely that this cell-disruptor was able to avoid ingesting Ni when feeding on high-Ni plants. We base this suggestion on two lines of evidence. First, the other cell-disruptor we tested (the two-spotted spidermite) was negatively affected by Ni, indicating that its feeding mode exposed it to elevated Ni. Secondly, field studies of S. polygaloides have discovered two other cell-disruptor hemipteran species that have relatively high whole-body Ni concentrations, thus indicating dietary exposure to Ni. One of these, Melanotrichus boydi (Miridae), is a specialist herbivore on S. polygaloides and was found to contain 800 µg Ni g−1 (Schwartz & Wall, 2001). The other, Coquilletia insignis (Miridae), contained 500 µg Ni g−1 when collected from S. polygaloides (Boyd et al., 2004).
Although considerable evidence has now been garnered for the defensive effectiveness of hyperaccumulated Ni (Boyd, 2004), it is unclear how many other hyperaccumulated elements may also have defensive functions. To our knowledge, only Zn, Se and Cd have been explicitly tested to date. Hyperaccumulated Se defends against some folivores (Hanson et al., 2003) and phloem-feeders (Hanson et al., 2004) but results for Zn have been mixed. Pollard & Baker (1997) and Jhee et al. (1999) showed high-Zn plants were defended against some folivores, and yet Huitson & Macnair (2003) and Noret et al. (2005) found no effect of hyperaccumulated Zn on snails (Helix aspersa Müller), suggesting that some folivores are insensitive to Zn. Boyd et al. (2002) showed that H. aspersa is not unaffected by all hyperaccumulated metals, as they demonstrated toxicity and deterrence for high-Ni leaves of the Ni hyperaccumulator Senecio coronatus (Asteraceae). Jiang et al. (2005) reported that Cd hyperaccumulation decreased feeding damage caused by the thrip Frankliniella occidentalis (Thysanoptera: Thripidae), which feeds by cell disruption. There is some tantalizing evidence, however, that all hyperaccumulated metals can defend plants against at least some herbivores. Coleman et al. (2005) used an artificial insect diet amended with eight metals [Cd, cobalt (Co), chromium (Cr), copper (Cu), magnesium (Mn), Ni, lead (Pb) and Zn] hyperaccumulated by plants. They showed that all tested metals were toxic at hyperaccumulator concentrations to larvae of an insect folivore, the diamondback moth Plutella xylostella (Lepidoptera: Plutellidae). Furthermore, toxicity extended far below hyperaccumulator concentrations for all the metals tested, suggesting that plants other than hyperaccumulators may benefit from elemental defenses. Additional studies are needed to reveal the extent and effectiveness of elemental defenses against the myriad natural enemies of plants.
We conclude that Ni in S. polygaloides defends against tissue-chewing insect herbivores (leaf-chewing and root-feeding) and some cell-disruptors but is ineffective against vascular tissue-feeding insects. Although the defensive function of hyperaccumulation has been investigated before (see summary in Boyd, 2004), our study is notable in several ways. First, we test the defensive effect of a metal against arthropod species representing a diversity of feeding modes. Besides representatives of the leaf-chewing folivores (Lepidoptera and mollusks) and aphids that have been tested by other studies of elemental defenses (e.g. Boyd & Martens, 1994; Pollard & Baker, 1997; Boyd & Moar, 1999; Jhee et al., 1999; Huitson & Macnair, 2003; Hanson et al., 2004), we used representatives of two previously untested feeding modes. Ours is the first study to use a xylem feeder (meadow spittlebug) and the first to use a rhizovore (cabbage maggot). Secondly, we used two arthropod orders previously untested with metal hyperaccumulating plants. We demonstrated a defensive effect of Ni hyperaccumulation on a noninsect arthropod, the two-spotted spidermite (Order Acari), as well as the cabbage maggot (Order Diptera). Thirdly, this study broadens our knowledge of the effects of Ni-based defense because the other arthropods used in our experiments were previously untested species from insect orders that have been tested before. Finally, our work helps refine the defense hypothesis by exploring herbivore traits (e.g. feeding mode) that can influence the effectiveness of a resistance trait (in this case Ni hyperaccumulation). This can predict which suite(s) of herbivores is (are) affected by hyperaccumulated elements and guide the design of much-needed field studies of the defensive effects of elemental hyperaccumulation (e.g. Martens & Boyd, 2002).