Portal de Periódicos da CAPES

Life history variation in Thlaspi caerulescens

2006; Wiley; Volume: 173; Issue: 1 Linguagem: Inglês

10.1111/j.1469-8137.2006.01948.x

1469-8137

Mark R. Macnair,

Botany, Ecology, and Taxonomy Studies

Thlaspi caerulescens has been identified (Assunção et al., 2003b) as a key model species in which to study the phenomenon of hyperaccumulation, where a small number of species accumulate a few metals to extraordinary concentrations in their leaves and shoots. Populations of this species are known to accumulate zinc, nickel and cadmium, and there have been many recent papers studying the biochemistry, molecular biology and physiology of zinc and cadmium hyperaccumulation in this material. The evolutionary advantage of hyperaccumulation has also attracted attention, particularly testing the hypothesis that the metals in the leaves act as an antifeedant, deterring a range of small invertebrate terrestrial herbivores (Noret et al., 2005). The basic ecology of the species has attracted less attention, however; this deficiency is at least partially rectified by two papers in this issue of New Phytologist, by Dechamps et al. (pp. 191–198) and Jiménez-Ambriz et al. (pp. 199–215). ‘This shows that the two phenomena – accumulation and tolerance – are independent of each other’ Thlaspi caerulescens is widely distributed in disjunct populations throughout northern Europe on both contaminated and uncontaminated soils. It is highly variable morphologically, and there has been some controversy about the correct taxonomy of this and other related species (Meyer, 1973). Some authors have suggested that the metallophyte populations should be distinguished at the subspecific level from the populations on more normal soils, but molecular analysis has not supported this distinction (Koch et al., 1998). The authors of these two papers have compared the life history of T. caerulescens growing on either contaminated or uncontaminated soils. Jiménez-Ambriz et al. compared three metallicolous populations with three nonmetallicolous populations in the south of France; and Dechamps et al. compared two populations from each habitat from Belgium and Luxembourg. The populations in both studies were grown in common garden experiments in order to identify the genetic differences between them, and under contrasting soil zinc concentrations to study the tolerance of the populations to this metal and the sensitivity of the various life-history characters to increasing pollution. Both studies have found the life history of this species to be very variable under these conditions, but some common themes can be identified. Both Dechamps et al. and Jiménez-Ambriz et al. found that the nonmetallicolous populations are much less tolerant of zinc contamination than those from contaminated environments, with the metallicolous plants growing very similarly in both contaminated and uncontaminated substrates, whilst the nonmetallicolous populations show a marked drop in performance with increasing metal. Reproductive characters were also found to be particularly sensitive to the metal. This might be because toxicity, acting continuously throughout life, will have a greater cumulative effect on late expressed characters (such as reproduction) than on the earlier vegetative characters. The lower tolerance of nonmetallicolous populations of T. caerulescens has been noted before (Escarréet al., 2000; Assunção et al., 2003a) in studies which only looked at vegetative characters in short-term experiments in soil or hydroponics. It seems, therefore, that T. caerulescens evolves locally a degree of tolerance dependent on substrate, since molecular evidence does not support a genetic structure associated with substrate (Koch et al., 1998; Jiménez-Ambriz et al., this issue).This implies that being tolerant in a noncontaminated environment is a disadvantage, the so-called cost of tolerance. Circumstantial evidence for such a cost has been found regularly in the study of metal tolerance (Macnair, 1997); it is frustrating that it has not been possible to identify the cause of this cost (Harper et al., 1997; Macnair, 1997). The difference in tolerance according to substrate in this species contrasts with what has been found in the other model zinc hyperaccumulator, Arabidopsis halleri, where little difference in tolerance is found between populations collected from different habitats (Bert et al., 2002). However, these studies involved vegetative growth in hydroponics; perhaps a much bigger difference would be found if A. halleri were also to be studied throughout its life cycle. Despite the relative lack of tolerance to zinc in nonmetallicolous populations, all populations of T. caerulescens hyperaccumulate this metal; indeed, nonmetallicolous populations accumulate greater concentrations of the metal under controlled conditions than metallicolous populations (Escarréet al., 2000; Dechamps et al., this issue). This shows that the two phenomena – accumulation and tolerance – are independent of each other, as was also shown in A. halleri (Macnair et al., 1999): tolerance is a phenomenon in which the plant interacts with the external metal concentration, with the internal concentration regulated by plant homeostatic mechanisms that are maintained more successfully in tolerant plants than in nontolerants. Accumulation is a phenomenon primarily about the distribution of metal between roots and shoots, though an increased uptake at the root may also be found (Lasat et al., 1996). Thus the life history variation revealed by Dechamps et al. and Jiménez-Ambriz et al. tells us about the variation found in respect of tolerance, and that it actually has little to do with hyperaccumulation per se. Both papers find that there is substantial variation within and between populations in critical life-history parameters. T. caerulescens is self-compatible, and it is generally assumed that selfing is the predominant breeding system in most populations. The French populations studied by Jiménez-Ambriz et al. do show the predicted molecular variation if this were true, with decreased heterozygosity and higher FIS. Thus genetic variation between populations is expected through a combination of reduced gene flow, genetic drift and local selection. Consistent effects associated with geography or substrate might, however, suggest adaptation to a particular factor. Thus most of the plants in the French populations (Jiménez-Ambriz et al., this issue) flowered in the first season, with at most 5% being biennials. Unfortunately, the French plants were not grown beyond one season, so it is not known whether they were iteroparous perennials or annuals. In the more northerly populations (Dechamps et al., this issue), however, nearly 50% of the nonmetallicolous populations exhibited a biennial lifecycle under low zinc conditions, whilst, in the metallicolous populations, bienniality was rare, with most plants being annuals or iteroparous perennials. These observations would be consistent with T. caerulescens responding to a drier environment by evolving greater annuality. The French populations will suffer greater drought, and endure more heat, than those in more northerly climes; one of the general features of metallicolous populations is that the soil structure is worse than in more normal soils, and this produces greater drought (Macnair, 1997). The evolution of annuality in species or populations of species adapting to the drought induced in these metalliferous soils has been shown previously. For instance, in Mimulus, the perennial, hydrophilic ancestral species M. guttatus has produced a number of derived species that are adapted to serpentine (M. nudatus, M. pardalis) or copper-contaminated (M. cupriphilus) soils (Macnair & Gardner, 1999). These species are all annual and flower early to avoid the Californian summer drought, which is much more severe on the freely draining soils of these metalliferous sites. Jiménez-Ambriz et al. found a consistent difference in the reproductive strategy of the metallicolous populations they studied in France compared with those from the normal soils. The plants flowered later, allowing a greater vegetative phase, so that the rosettes were bigger at flowering, producing more floral stalks but fewer fruits and smaller and less numerous seeds. The observation of later flowering in response to metals is more unusual: Agrostis capillaris and Anthoxanthum odoratum on the edge of the Trelogan lead/zinc mine were found to flower earlier than those in the adjacent pasture (McNeilly & Antonovics, 1968). M. guttatus has also evolved earlier flowering on copper mines, even where it remains a perennial species (Macnair & Gardner, 1999). Interestingly, although the populations from the different habitats showed different basic life histories, Dechamps et al. found that the overall fitness of the metallicolous and nonmetallicolous populations did not differ at low zinc concentration, although as zinc concentration increased, the nonmetallicolous populations did show reduction in fitness as their lack of metal tolerance was manifested. In France, Jiménez-Ambriz et al. found that total plant seed yield did not differ between population types in the field, despite the underlying differences in life history displayed in the common garden experiment. This indicates that T. caerulescens may be able to achieve essentially equal seed productions through a number of different strategies. Thus we now know more about the pattern of life history variation in this model hyperaccumulator. We can see some consistent trends in the diversity of life-history components, and some potentially adaptive characteristics can be seen paralleled in other metalliferous species. However, much variation in life history remains unexplained, and it is not going to be easy to disentangle random, neutral variation from the effects of selection and history.