Ecology of metal hyperaccumulationстатья из журнала
Аннотация: Research into the accumulation of extraordinary levels of metals by plants – hyperaccumulation (Brooks et al., 1977) – has moved from exploration through investigations of mechanisms to exploitation. During the past decade, in addition, the ecological consequences of metal hyperaccumulation began to attract attention. Ecological studies of metal hyperaccumulation have been designed to: provide insight into how and why metal hyperaccumulation has evolved by determining its adaptive value; and to examine how the extraordinary metal concentration of hyperaccumulators impacts species relationships in the habitats in which hyperaccumulators have evolved. In this issue, Hanson et al. (pp. 655–662) make important contributions to our understanding of the ecology of Se hyperaccumulation. Plant bodies contain many elements, some in relatively large amounts, some in trace quantities, and others in between. The familiar lists of plant macro- and micronutrients, derived from centuries of research on plant nutrition, show the importance of various elements to plant growth. Thus, we have extensive data that define the 'normal' concentrations for many elements found in plant tissues. Some plants, however, contain unusually large concentrations of certain elements on a dry mass basis. Termed hyperaccumulators because of their extraordinary level of accumulation (Brooks et al., 1977), the threshold that defines hyperaccumulation depends on the element involved. Hyperaccumulation thresholds for the best studied metals are: 10 000 µg g−1 of Mn or Zn; 1000 µg g−1 of Ni, Cu or Se; and 100 µg g−1 of Cd, Cr, Pb or Co (Reeves & Baker, 2000). Other less studied metals, including Al (Jansen et al., 2002) and As (Ma et al., 2001; Meharg, 2002), have hyperaccumulation thresholds of 1000 µg g−1. The unusual elemental composition of hyperaccumulators has attracted considerable scientific interest (Cobbett, 2003; Kraemer, 2003; Macnair, 2003). Some researchers have focused on identifying new species of hyperaccumulators through botanical explorations. For example, research in Cuba (Reeves et al., 1996) has discovered more hyperaccumulators of Ni than have been found in any other location to date. Most hyperaccumulators are found on soils that are unusually rich in metals, such as ultramafic soils, and Cuba contains a number of ultramafic soil areas (Reeves et al., 1996). Other researchers have explored the physiological mechanisms whereby hyperaccumulators acquire, process, and sequester these normally toxic metals (Salt, 2001). Still others are examining the usefulness of these species for phytoremediation, cleaning up metal contaminated sites (McGrath & Zhao, 2003), or for phytomining, using them to mine metals from high metal soils (Anderson et al., 1999). During the past decade, a fourth area of interest came to the fore: ecology. Hanson et al. have looked at Se. Most plant species contain little Se, usually less than 1.5 µg g−1 (Reeves & Baker, 2000). However, plants containing unusually elevated levels of Se have been known from deserts of the United States since at least the 1930s (Rosenfeld & Beath, 1964). Originally attracting attention because of their toxic effects on livestock (Rosenfeld & Beath, 1964), they were later studied by physiologists interested in the mechanisms of Se metabolism (Shrift, 1972). Recently, these plants have received attention for their potential use in phytoremediating Se-contaminated soils (Bañuelos, 2001). Despite this attention, study of the ecology of Se hyperaccumulation has languished until now. As recently reported in New Phytologist, Hanson et al. (2004) showed that Se hyperaccumulated by Brassica juncea was toxic to an invertebrate herbivore and to fungal pathogens. In their current paper, they extend these results to include defensive effects against phloem-feeding aphids. Together, these papers illustrate how studies of hyperaccumulator ecology can illuminate the ecological function and evolutionary value of metal hyperaccumulation. Hyperaccumulation has been hypothesized to perform several functions in hyperaccumulator species. Tests of these hypotheses seek to demonstrate an adaptive value of hyperaccumulation and to explain how hyperaccumulation may have evolved. As with any discussion of function in an evolutionary context, it must be kept in mind that the current function of a trait may not be that for which the trait originally evolved (the latter being adaptation sensu stricto). If the former is the case, then the term exaptation may be applied to the trait (Gould & Vrba, 1982). Separating adaptation from exaptation is extraordinarily difficult and would be greatly aided by study of the phylogeny of hyperaccumulation, which is only just getting under way (Broadley et al., 2001; Mengoni et al., 2003). An early review (Boyd & Martens, 1992) summarized four postulated benefits of metal hyperaccumulation in plants, which during the last decade have begun to be tested experimentally. First, the tolerance/disposal hypothesis suggested that hyperaccumulation is a mechanism that allows sequestration of metals in tissues (tolerance) and, in some cases, elimination of metals from the plant body by the shedding of those high metal tissues (disposal). A recent genetic analysis showing that Zn tolerance and Zn hyperaccumulation are decoupled (Macnair et al., 1999) argues against the tolerance portion of this hypothesis, and there is also little evidence to support the disposal portion (Boyd & Martens, 1998). Second, the interference hypothesis suggests that perennial hyperaccumulator plants enrich the surface soil under their canopies by production of high-metal litter to prevent establishment of less metal tolerant species. This hypothesis, re-named the elemental allelopathy hypothesis by Boyd & Martens (1998), remains untested, although Boyd & Jaffré (2002) showed that surface soil metal levels are elevated under canopies of the long-lived Ni hyperaccumulator tree species Sebertia acuminata (Fig. 1a). Third, the drought resistance hypothesis states that hyperaccumulated metal may help hyperaccumulators withstand drought. Evidence bearing on this hypothesis is scarce, but elegant manipulative experiments by Whiting et al. (2003) found no evidence that Ni or Zn hyperaccumulator status provided increased drought resistance to the hyperaccumulators Alyssum murale and Thlaspi caerulescens. Finally, the defence hypothesis suggests that elevated metal concentrations in plant tissues protect plants from certain herbivores or pathogens. It is this hypothesis that has attracted the most research and about which there is both current debate and continuing exploration. An Ni hyperaccumulating plant and a high Ni herbivorous insect. (a) the New Caledonia Ni hyperaccumulator Sebertia acuminata produces a latex with extraordinarily elevated Ni concentration (up to 260 000 µg Ni g−1 dry mass: Jaffréet al., 1976). This large quantity of Ni gives the latex its bluish cast. (b) The high Ni mirid bug Melanotrichus boydi on a flowering stem of the California Ni hyperaccumulator Streptanthus polygaloides. This specialist insect, found only on this hyperaccumulator species, was unknown to science prior to insect surveys by Wall & Boyd (2002), illustrating the potential of studies of hyperaccumulator ecology to yield unique discoveries. Plant chemical defences (e.g. alkaloids, terpenes, glucosinolates) are generally derived from photosynthate, but hyperaccumulated metals represent a suite of defences, termed elemental defences (Martens & Boyd, 1994), derived from soil minerals. Martens & Boyd (1994) suggested that elemental defences differed from secondary (organic) chemical defences because: their toxic principle is an element taken up from the soil rather than one derived from photosynthate; and they could not be biochemically degraded by the chemical counterdefences of herbivores because of their elemental nature. The vast majority of experimental investigations regarding this hypothesis have focused on hyperaccumulators of Ni or Zn (Boyd, 1998). Defensive effects have been shown in some cases (Boyd & Martens, 1994; Boyd et al., 1994; Pollard & Baker, 1997; Jhee et al., 1999; Boyd et al., 2002) but not others (Boyd et al., 1999; Huitson & Macnair, 2003). In two cases, hyperaccumulation status actually promoted attack: in one case by a pathogen (Turnip Mosaic Virus; Davis & Boyd, 2001) and in another by herbivorous snails (Hanson et al., 2003). These results show that, like other plant defences, elemental defences are not absolute and provide protection against only some of the myriad enemies that plants face in natural situations. The paper by Hanson et al., in this issue, along with earlier work by these authors (Hanson et al., 2003), explores a third element hyperaccumulated by plants and shows evidence of defensive effects against two invertebrate herbivores and two fungal pathogens. Unlike earlier work with Ni hyperaccumulation, which failed to show defensive effects against phloem-feeding aphids (Boyd & Martens, 1999), Hanson et al. show that aphids are susceptible to Se-based elemental defences. This illustrates that elements may differ in their defensive outcomes against herbivores with particular feeding strategies. Realization of this point should stimulate additional investigations into the postulated defensive nature of each of the metals hyperaccumulated by plants, using herbivores that represent each of the various modes of herbivory (e.g. folivores, phloem feeders, xylem feeders) as well as examination of specialist and generalist herbivores. Most studies of the defence hypothesis have contrasted herbivore/pathogen response to hyperaccumulating and nonhyperaccumulating plants. However, tantalizing evidence hints that defensive effects of metals may extend to concentrations far below the minima used to define hyperaccumulation (Boyd & Martens, 1998). For example, in this issue Hanson et al. show that Se is toxic to aphids at a leaf concentration of 125 µg g−1 and that a sublethal effect (reduced aphid population growth rate) was observed at only 1.5 µg g−1 Se. This latter effect occurred for an Se concentration that is at the upper boundary of the normal range of Se in plant tissues (Reeves & Baker, 2000). This evidence suggests that elemental defences may be effective at levels below those used to define hyperaccumulation. If this is indeed the case, then elemental defences may be more widespread in natural communities than previously suspected (Boyd & Moar, 1999). In an interesting twist, studies of the ecology of metal hyperaccumulation may in turn influence the threshold values used to define hyperaccumulation. In the case of Se, Reeves & Baker (2000) stated that reports of Se toxicity to grazing animals at levels less than 1000 µg g−1 are one reason for their suggestion that the threshold value for defining Se hyperaccumulation be lowered to 100 µg g−1. The possibility of a selective effect of metals at concentrations below the hyperaccumulation level is exciting because these effects may reveal mechanisms whereby metal hyperaccumulation has evolved. Any ecological benefit of elevated metal concentrations, whether herbivore defence, elemental allelopathy or another, could provide a basis for the progressive evolution of still greater metal concentrations. Thus, studies of hyperaccumulator ecology may help illuminate the evolutionary basis of metal hyperaccumulation. It is important to realize that the question of why hyperaccumulation has evolved is separate from questions regarding the consequences of this trait for other species in hyperaccumulator habitats. Regardless of why plants hyperaccumulate, elevated metal levels in plant tissue will have consequences for other organisms that share those habitats. The elemental defence hypothesis predicts that metal hyperaccumulation will negatively affect at least some herbivores/pathogens in a given habitat. However, by various means (Boyd & Martens, 1998), others will be unaffected by elemental defences. A field study of the California Ni hyperaccumulator Streptanthus polygaloides showed that mammalian herbivores damaged high-Ni plants, probably because their generalist diet diluted hyperaccumulator foliage with that of low-Ni species (Martens & Boyd, 2002). Specialist herbivores have also been found that evolved in association with hyperaccumulators and feed on these plants without harm. Examples discovered to date include the mirid bug Melanotrichus boydi (Fig. 1b) feeding on S. polygaloides (Schwartz & Wall, 2001) and the chrysomelid beetle Chrysolina pardalina feeding on the South African Ni hyperaccumulator Berkheya coddii (Mesjasz-Przybylowicz & Przybylowicz, 2001). Specialist herbivores that consume metal hyperaccumulators without harm are of great scientific interest for three reasons. First, their ability to consume high metal tissues implies they possess adaptations that may give insights into metal tolerance mechanisms in animals (Przybylowicz et al., 2003). Second, coevolution between hyperaccumulators and metal tolerant herbivores could have accelerated the evolution of metal hyperaccumulation by plants. The stepwise interplay between plant defences and herbivore counterdefences, considered a biochemical 'arms race' (Kareiva, 1999), helps to explain the diversity of plant chemical defences. In a similar fashion, we can hypothesize that coevolution mechanisms may have fostered the extremely elevated metal concentrations that characterize metal hyperaccumulators. Pollard (2000) suggested that our ability to address coevolutionary questions about metal hyperaccumulators and their specialist herbivores constituted a valuable model system. The third reason is that elevated metal levels in adapted herbivores may have consequences for organisms at other trophic levels in the food webs of ultramafic communities. Some hyperaccumulator-specific herbivores have relatively elevated whole body metal concentrations. Melanotrichus boydi (Fig. 1b) averages 780 µg Ni g−1, and work in progress in South Africa by our lab has documented an insect herbivore containing an astounding 3500 µg Ni g−1. Such elevated metal levels may have consequences for organisms at other trophic levels in food webs that contain hyperaccumulators and their specialist herbivores. One ecological consequence of the existence of high metal herbivores is that metal may influence the herbivore's interactions with other species, such as predators or pathogens. Organic plant defences sequestered by some tolerant herbivores can chemically defend them against predators (Termonia et al., 2002). We investigated this question using M. boydi, with mixed results. Tests involving three arthropod predator species fed either high Ni M. boydi or low Ni prey showed no effect in two cases but significantly higher mortality for a thomisid spider when fed M. boydi (Boyd & Wall, 2001). Another test, using entomopathogenic bacteria and nematodes as example pathogens, showed no beneficial effect of the high Ni concentration of M. boydi on insect mortality rates (Boyd, 2002). Further research with insect herbivores containing greater amounts of Ni or other hyperaccumulated metals should be conducted to determine the generality of these results. The potential toxicity of high metal insects to predators may also raise environmental concerns about applied uses of hyperaccumulators for phytoremediation or phytomining. For example, toxicological studies have shown that vertebrates can be negatively affected by a diet containing > 500 µg Ni g−1 (Outridge & Scheuhammer, 1993), so that consumption of high-Ni insects could harm birds or other animals if large numbers were consumed. Hyperaccumulators may also affect ecosystem level processes by mobilizing metals into food webs. Hyperaccumulators are generally minor vegetation components in most European and North American habitats, but can be relatively abundant in some locations in New Caledonia, Cuba and South Africa. In these situations, significant quantities of metals may be mobilized into food webs. A pioneering study (Peterson et al., 2003), performed in a Portuguese ultramafic community containing only a single Ni hyperaccumulator species (Alyssum pintodasilvae), showed that hyperaccumulators can act as gateways for metals into food webs. Predator assemblages from communities containing the hyperaccumulator showed significantly elevated Ni concentrations compared to those from communities lacking the hyperaccumulator. Other ecosystem level consequences have scarcely been articulated, much less studied. For example, the influence of high metal (hyperaccumulator) litter on decomposer communities and nutrient cycling rates (Boyd & Martens, 1998) is as yet virtually unexplored. Study of metal hyperaccumulator ecology is in its infancy, yet this area of research promises to reveal exciting biological phenomena. Research to date has focused mainly on Ni and Zn hyperaccumulation, but the ground-breaking research of Hanson et al. has begun to explore the ecology of Se hyperaccumulation. It is clear that much research remains to be conducted: we have multiple functional hypotheses to test regarding multiple hyperaccumulated metals. It is likely that some hypotheses will be validated for one metal but not another and, for a given hypothesis (such as the defence hypothesis), the result may be dependent on the ecological specifics of both the plant species and the herbivores involved. These questions are important because they provide an ecological context for molecular/physiological studies as well as for studies that are exploring the uses of metal hyperaccumulators for phytoremediation/phytomining. Research on hyperaccumulator ecology promises more fully to illuminate a fascinating plant trait and to extend our knowledge of hyperaccumulation across the biological hierarchy from molecules to ecosystems.
Год издания: 2004
Авторы: Robert S. Boyd
Издательство: Wiley
Источник: New Phytologist
Ключевые слова: Heavy metals in environment, Aluminum toxicity and tolerance in plants and animals, Arsenic contamination and mitigation
Открытый доступ: closed
Том: 162
Выпуск: 3
Страницы: 563–567