Contents:
1. Introduction
2. Rhizome architecture
3. Adventitious rooting as an adaptation to disturbance
4. Adaptations to waterlogged soil and anaerobic conditions
5. Soil preferences and nutrient cycling
6. Silicon
7. Mycorrhizae and root hairs
8. Nitrogen fixation
9. Effects of humidity on growth
10. Sexual reproduction
11. Allelopathy
We have relatively little direct knowledge of the ecology and physiology of the giant horsetails. So far, only two published papers that have dealt specifically with aspects of giant horsetail ecology, one conducted in Costa Rica ( Hauke, 1969a ) and one conductd in Cuba ( Álvarez de Zayas, 1982 ). These studies were limited to some qualitative, though very interesting, observations on natural history . One recent study has investigated the biomechanics of aerial stems in E. giganteum 1 ( Spatz et al., 1998 ). However, considerably more is known about the ecophysiology of several temperate Equiseta.
Giant horsetails are pioneer (early-successional) obligate wetland plants and are poor competitors ( Hauke, 1969a ; Ĝllgaard, 2000, personal communication). Abundance of groundwater supply and lack of competition are key habitat requirements of these plants. Hence, they are often associated with rivers and alluvial soils ( Hauke, 1969a ; Álvarez de Zayas, 1982 ). The limitation of giant horsetails to higher altitudes in the tropics is probably due to their poor competitive abilities and their inability to tolerate shade. Hauke ( 1969a ) observed that the giant horsetails stop producing cones when shaded by other vegetation and are subsequently displaced by other plants. Hence, giant horsetails do not tend to persist in a given site unless the disturbance regime or other factor prevents shading-out of colonies. The lower competitive pressure in the cooler high altitudes combined with increased light intensity may allow the horsetails to "hold their own" against other vegetation ( Hauke, 1969a ). This would explain why the genus is absent from the lowland Amazon basin where temperatures are warm and plant competition is especially intense. Hauke ( 1969a ) described several cases wherein giant horsetail colonies in Costa Rica had disappeared, presumably due to the process of succession at once-suitable sites. Ĝllgaard (2000, personal communication) has observed that giant horsetails are frequent pioneers on land and mud slides in the valleys of the Río Pilaton and Río Pastaza in Ecuador. Professor Ĝllgaard suspects that these pioneer stands can probably persist until sufficient forest regeneration occurs to shade out the horsetails (after ~25-50 years).
Rhizome architecture in relation to ecology
The giant horsetails, like other Equisetum species, develop extensive underground rhizome systems. Unfortunately, there has been no study of the rhizome architecture of the giant horsetails. However, it is known that the rhizomes of E. telmateia (the largest member of the subgenus Equisetum ) can extend more than 4 m deep into wet clay soil ( Page, 1997 ). Anthony Huxley reported in the book "Plant and Planet" (1975 , p. 243) that "field bindweed is recorded at a depth of 7 meters and horsetails in light soil two or three times as deep again". This report suggests that Equisetum rhizomes may penetrate to the extraordinary depth of 21 m in certain situations! Unfortunately, Huxley did not mention his source for this report or the species and location involved, so it would be proper to remain skeptical of this claim. Golub and Whetmore ( 1948 ) excavated the rhizome system of a colony of E. arvense (the "weedy" member of the genus) to a depth of 2 m and found five successive horizontal layers of rhizomes connected by vertical rhizomes. This rhizome system extended below 2 m, but the investigators did not excavate further. I suspect that this "tiered" rhizome architecture may be unique in the plant kingdom. Indeed, other rhizomatous plants generally have but a single horizontal rhizome system layer ( Bell and Tomlinson, 1980 ). It would be most interesting to determine how deeply the rhizomes of giant horsetails penetrate and whether the rhizomes form a tiered architecture like that of E. arvense . Interestingly, the deep rhizome system of E. arvense allows it to access saturated soil layers even in situations where it appears to be growing in relatively dry surficial soil (Hauke, 1966 ). For a discussion of how the deep penetration of Equisetum rhizomes contributes to their ability to acquire nutrients, see " Soil preferences and nutrient cycling " below.
The rhizome systems of giant horsetails allow these plants to colonize disturbed areas quickly via vegetative growth and to survive episodes of erosion and sediment accumulation, especially along rivers (Hauke, 1963 ). In addition, rhizome segments that are exposed by erosion and broken-off can be carried downstream to establish new clones ( Hauke, 1969a ). Equisetum species generally invest a large proportion of resources in rhizome growth. Borg ( 1971 ) found that E. palustre may produce more than 100 times more rhizome biomass than aerial stem biomass. Equisetum arvense also allocates the larger proportion of it's dry matter to rhizomes and tubers (thought not to such an extreme extent as E. palustre ) ( Marshall, 1986 ). The large pool of resources stored in Equsietum rhizomes facilitates aerial shoot regeneration if a distrubance destroys the aboveground stems. Hence, this growth strategy is adaptive for the types of disturbance-prone habitats favored by many Equiseta , including giant Equiseta . Although there have been no studies of the ratio of above to below ground biomass allocation in giant horsetails, I suspect that the giants allocate a considerably higher proportion of resources to the aerial shoots than do the more diminutive northern species. It would also be interesting to know whether the two subgenera of the genus Equisetum differ overall in their biomass allocation patterns.
Vegetative reproduction probably accounts for the widespread occurrence and persistence of Equisetum hybrids, such as E. x schaffneri, even where one or both parents is absent (e.g. in Mexico and Venezuela for E. x schaffneri ) ( Hauke, 1963 ). This situation may occur when a hybrid persists vegetatively, yet one or both parent species does not. Hybrid horsetails are for the most part sterile and without means of sexual reproduction. However, Krahulec et al. ( 1996 ) have documented the occurrence of occassional viable spores in hybrids of the temperate species within the subgenus Hippochaete . This finding adds support to Hauke's ( 1963 ) hypothesis that the occurrence of E. x schaffneri in the absence of one of its parent species may be due to occassional viable spores being produced by E. x schaffneri. The plants resulting from such spores may then persist vegetatively.
Adventitious rooting as an adaptation to disturbance
All Equiseta have preformed bud and root primordia ( photo of an emerging adventitious root below a branch bud of E. giganteum ) at each node of both the aerial stems and underground rhizomes ( Gifford and Foster, 1988 ). This allows Equisetum stems to quickly put forth new roots and shoots on aerial stems when the stems are partly or wholly buried in sediment. Hence, even if the deeper parts of a stem or rhizome become crushed or smothered by sediment, the upper parts may be able to survive and reestablish the clone ( Gastaldo, 1992 ). This ability is clearly advantageous for enhancing survival of Equisetum species in the wake of disturbance events in riparian and other wetland habitats. The preformed primordia can also facilitate vegetative propagation and dispersal via stem pieces ( Wagner and Hammitt, 1970 ; Hauke, 1963 ). Schaffner ( 1931 ) and Praeger (1934 ) utilized the adventitious rooting capabilities of Equisetum stems to successfully propagate many species from aerial stem cuttings.
Gastaldo ( 1992 ) gives evidence for similar stem regeneration abilities in the large extinct Equisetum relative Calamites (see also this computer graphics reconstruction of Calamites) and discusses the ecological importance of these abilities. Similarly, Kelber et al. ( 1998 ) found that the extinct close relative of extant horsetails, Equisetites arenaceus, could propagate vegetatively via the adventitious rooting of shed branches.
Adaptations to waterlogged soil and anaerobic conditions
Like nearly all organsims, plants require oxygen (O2) for efficient cellular respiration. Plants that grow in water-saturated soil often have to cope with anoxic conditions around their underground organs ( Blom and Voesenek, 1995 ). This is because O2 diffuses 10,000 times more slowly through liquid water than through air ( Grable, 1966 ). Under waterlogged conditions, cellular respiration by plant roots and soil microorganisms often quickly depletes the available O 2 , leading to anoxic soil conditions ( Drew and Lynch, 1980 ; Kludze and DeLaune, 1995 ). These conditions lead to a large decrease in plant nutrient availability (Ernst, 1990 ) and to buildup of phytotoxins produced by anaerbic soil microbes or by anaerobic respiration in plant roots ( Koch and Mendellsohn, 1989 ).
To deal with anoxic conditions, wetland plants have several morphological and physiological adaptations to maintain aerobic respiration by facilitating transport of O 2 from the atmosphere to underground and underwater organs ( Allen, 1997 ). In many wetland plants, gas spaces (lacunae) in specialized tissue (called aerenchyma) provide pathways for O 2 and carbon dioxide (CO2) to move from one part of the plant to another much more quickly than would be possible through tissue without lacunae ( Allen, 1977 ). Rhizomes and stems of wetland Equisetum species, including the giant horsetails, have large canals that are thought to function like aerenchyma tissue in facilitating O 2 transport ( Hauke, 1963 ; Hyvonen et al., 1998 ). Oxygen movement through aerenchyma occurs either by diffusion alone or by diffusion combined with convection ( Allen, 1997 ). These mechanisms and their relative effectiveness have many important implications both for wetland ecology and for the growth and productivity of crop plants, such as rice (Oryza sativa L.), that typically grow in waterlogged soils ( Wassmann and Aulakh, 2000 ; Allen, 1997 ). The most efficient known mechanism for oxygen transport to submerged plant parts is via pressurized convection ( Allen, 1997 ).
Up to this time, studies of pressurized O 2 transport in wetland plants have focused exclusively on angiosperms, with the exception of one study that investigated the gymnosperm Taxodium distichum L. ( Grosse et al., 1992 ). However, the rhizomes of Equisetum species often concentrate more deeply than the roots and rhizomes of accompanying vegetation ( ; Borg, 1971 ). Marsh et al. ( 2000 ) found that, in an Alaska wetland, Equisetum rhizomes were concentrated in the deeper C soil horizon whereas the roots and rhizomes of other species were concentrated in the surface O horizon. The especially deep penetration of waterlogged sediments by Equisetum rhizomes suggests that existence of efficient mechanisms for rhizome aeration.
The only studies of gas transport in Equisetum to date have dealt with E. fluviatile, a species that frequently grows as an emergent aquatic plant ( Hauke, 1978 ) and one of the most anaerobiosis tolerant Equisetum species (at least in the temperate zone) ( Page, 2002 ). An early study by Barber ( 1961 ) found a diffusion gradient from high concentrations of O 2 and low concentrations of CO 2 in aerial stems to the reverse condition in submerged rhizomes. In addition, Barber ( 1961 ) found that diffusion along excised aerial stems was relatively efficient. However, this study did not provide information that would indicate whether or not a pressurized ventilation mechanism might be active in E. fluviatile . A study by Hyvonen et al. ( 1998 ) of methane release from an E. fluviatile stand suggested that this species does not have a pressurized ventilation flow system because there was not a discernable diurnal pattern of methane efflux. Similarly, Strand (2002) found that E. fluviatile had a "low or non-detectable" air flow rate in its stems, yet was found in "unexpectedly deep water".
Page ( 2002 ) mentioned that Equisetum species in the British Isles vary in their tolerance of anaerobic soil water conditions. Equisetum fluviatile appears able to tolerate the greatest degree hypoxia in soil water whereas E. telmateia is least tolerant of anaerobic soil water and occupies sites with continually flowing groundwater ( Page, 2002 ).
Soil preferences and nutrient cycling
Alvarez de Zayas ( 1982 ) observed that in Cuba Equisetum giganteum is associated with mineral rich, acidic, alluvial soils. Correspondingly, my experiences with E. giganteum in cultivation (and, to a lesser extent, other Equisetum species) suggest that this species has a high requirement for micronutrients. To date, no other field or laboratory studies have investigated the soil preferences or nutrient needs of giant horsetails. However, a recent study of nutrient acquisition by temperate Equisetum species provides interesting insight into the possible roles of giant horsetails in mineral nutrient cycling ( Marsh et al., 2000 ).
Members of the genus Equisetum have the ability to extend their rhizomes deeply into saturated soil ( Marsh et al., 2000 ; Borg, 1971 ). The rhizome system has generally been found to comprise most of the plant’s biomass ( Borg, 1971 ; Marshall, 1986 ). The ability of Equisetum rhizomes to penetrate deeply into wetland soils plays an important role in their recently discovered role as nutrient pumps. In an Alaskan shrub wetland, ( Marsh et al., 2000 ) found that Equisetum species can acquire and accumulate substantial amounts of phosphorus (P), potassium (K), and calcium (Ca) from lower soil layers and transport these nutrients to the surface where they are available to other plants. Remarkably, these investigators found that although Equisetum species made up only 5% of the total biomass of the wetland community, the Equisetum tissues had 16% of the total phosphorus and 24% of the total potassium ( Marsh et al., 2000 ). Furthermore, Equisetum species contributed disproportionately to soil nutrient inputs in the shrub wetland. During the two year study period, Equisetum litter provided 75% of the calcium, 55% of the phosphorus, and 41% of the K input to the soil ( Marsh et al., 2000 ). The nutrient pumping of Equisetum species in the shrub wetland probably contributed to the unusually high primary productivity of the ecosystem ( Marsh et al., 2000 ). This nutrient pumping function of the shrub wetland Equisetum species appears to be at least partly due to the ability of Equisetum rhizomes and roots to penetrate more deeply into the soil than roots and rhizomes of other wetland plants. While Equisetum roots and rhizomes were concentrated in the deeper C horizon of the soil, roots and rhizomes of other wetland plants concentrated in the surface O horizon ( Marsh et al., 2000 ).
To date, there has been no clear evidence that Equisetum species are mycorrhizal (either in the gametophyte or the sporophyte stage) and most studies have found essentially no mycorrhizal colonization of horsetails ( Read et al., 2000 ). Although Koske et al. ( 1985 ) found fungal structures in roots of Equisetum species growing in a sand dune habitat, the close association of the Equisetum roots with roots of characteristically mycotrophic plants raises the possibility that the observed fungal structures represented "simply the penetration [of Equisetum roots by] a 'non-host" ( Read et al., 2000 ). Hence, the role of mycorrhizae in Equisetum ecology remains controversial. Overall, however, Equisetum species clearly appear to do quite well in many situations without mycorrhizal associations ( Reat et al., 2000 ). For example, Marsh et al. ( 2000 ) found no mycorrhizal colonization of Equisetum roots in the Alaskan shrub wetland they studied. Although enhanced phosphorus acquisition is often a major contribution of mycorrhizal associations to plant nutrition ( Orcutt and Nilsen, 2000 ), the mycorrhizae-free Equisetum species studied by Marsh et al. (2000 ) absorbed soil nutreints, including phosphorus, very effectively.
Schaffner ( 1938 ) and Page (2002 ) have observed that Equisetum species have exceptionally long root hairs and Page (2002 ) has noted that these hairs are "unusally persistent", at least in water culture. Page ( 2000 ) hypothesized that these root hairs may function to enhance the absorptive capacity of Equisetum roots in a manner similar to mycorrhizae. Marsh et al. ( 2000 ) noted the presence of root hairs on Equisetum roots in the O horizon but not on roots in the C horizon of the Alaskan shrub wetland they studied. The investigators hypothesized that nutrient concentrations were lower in the O horizon, necessitating greater roots surface area for absorption, whereas nutrient concentrations were high enough in the C horizon to inhibit formation of root hairs ( Marsh et al., 2000 ).As in other pteridophytes, sexual reproduction in Equisetum occurs by means of spores. Although, as far as I know, sexual reproduction of giant horsetails has not been observed in the wild, successful spore germination and gametophyte development have been observed in the laboratory (Hauke, 1963 ; 1969b ). However, what is known of the ecology of sexual reproduction in temperate horsetails likely applies to giant Equiseta as well.
Equisetum spores are short-lived and can germinate within 24 hours of release from the cone. After 5-17 days, depending on ambient humidity, spores are no longer viable ( Hauke, 1963 ). In the tropical giant horsetails, cones and spores may be produced throughout the year ( Hauke, 1963 ), but in temperate species spores are generally produced over a short period of time during the growing season ( Duckett, 1985 ). Hauke ( 1969a ) observed, however, that shading causes a cessation of coning in giant horsetails. Equisetum gametophytes appear to require a substrate of recently exposed bare mud to establish themselves ( Duckett and Duckett, 1980 ). Behaving as pioneer species, the gametophytes rapidly attain sexual maturity and are adversely affected by competition from bryophytes and vascular plants ( Duckett and Duckett, 1980 ; Duckett, 1985 ). The inefficiency of spore germination and gametophyte development in non-pioneer situations probably limits gene flow and leads to the high degree of genetic divergence found between Equisetum populations ( Korpelainen and Kolkkala, 1996 ). Therefore, sexual reproduction in Equisetum is limited to rather narrow ecological conditions and this limits the dispersal ability of Equisetum via spores.
The uniform chromosome number throughout the genus (n = 108)2 facilitates hybridization between Equisetum species (Scagel et al., 1984 ). Hybridization is also favored by the relatively narrow ecological requirements of gametophytes that encourages the formation of mixed populations of gametophytes on suitable sites ( Hauke, 1978 ). These mixed populations increase the probability of cross fertilization between gametophytes of different, but compatible, species. In areas where environmental conditions are especially conducive to spore germination and gametophyte establishment, Equisetum hybrids are particularly frequent and widespread. In Britain and Ireland, for example, Equisetum hybrids are especially successful. This success appears to be due primarily to the moist temperate oceanic climate and relatively little competition from other plants, conditions that favor both gametophyte and sporophyte generations of Equisetum ( Page, 1985 ). Equisetum hybridization is especially frequent within the subgenus Hippochaete, within which five common hybrids are known (including E. x schaffneri). Within the genus Equisetum , there is only one common hybrid ( Hauke, 1978 ). There are many more known hybrids within each subgenus, but these hybrids tend to be much less common. No hybrids between the two subgenera have yet been reported and this provides further evidence that the two subgeneral are naturally distinct ( Krahulec et al., 1996 ).
