Ferns: a xylem success story


The fern clade (monilophytes; Pryer et al., 2004) has a remarkable species diversity that outnumbers that of conifers and all other extant gymnosperms by two orders of magnitude (Fig. 1). The fossil record provides ample evidence of the ecological success and longevity of monilophytes throughout Earth history (Fig. 2), yet, because ferns today are generally out-competed in drier environments by seed plant groups, it is often assumed that ferns are not as ‘successful’ as angiosperms or gymnosperms (Rothwell, 1996). However, palaeoecological studies show that ferns continued to play a prominent role in terrestrial ecosystems for c. 30 million yr after the evolution of the first seed plants (spermatophytes) in the late Devonian (c. 380 million yr ago (Mya)) (DiMichele & Aronson, 1992; Willis & McElwain, 2002). Following the evolution of angiosperms in the early Cretaceous, there is evidence that ferns and other spore-bearing plants retained ecological dominance of some terrestrial ecosystems long after angiosperms attained much higher species richness (Wing et al., 1993). During intervals of extreme environmental stress, ferns are the so-called ‘disaster taxa’, the early pioneers of denuded landscapes (Tschudy et al., 1984; DiMichele & Phillips, 1996; McElwain & Punyasena, 2007). For example, following two of the greatest mass extinction events in Earth history, at the Triassic–Jurassic boundary (200 Mya) and the Cretaceous–Paleogene boundary (65 Mya), it was ferns and not gymnosperms or angiosperms, respectively, that were ecologically dominant in the recovery interval (McElwain & Punyasena, 2007). Despite the undeniably important contribution of over 11 000 fern species to current and past biodiversity and ecosystem function, our understanding of fern physiology is dwarfed by that of more derived seed plant groups. Furthermore, the biological attributes that have enabled ferns to persist as a group for over 380 million yr, and maintain higher modern diversity than their gymnosperm counterparts, is rarely considered. In this issue of New Phytologist, Pittermann et al. (pp. 449–461) have taken important steps in addressing this question. Their study investigates the morphological and functional attributes of fern xylem compared with that of gymnosperms and has elucidated features that are contrary to expectations for a generally wet-loving group of plants.

Figure 1.

Comparison of known species diversity among extant tracheophytes (vascular plants): angiosperms, gymnosperms, ferns and lycophytes. Estimates of species diversity were obtained from a compilation by Palmer et al. (2004).

Figure 2.

A simplified time tree of ferns (monilophytes) redrawn from Pryer & Schuettpelz (2009) showing the origin of ferns in the late Devonian and subsequent divergence times of the five major fern lineages. Leptosporangiates include gleichenioids, heterosporous ferns, schizaeoids, tree ferns, and polypods. C, Carboniferous; Cret., Cretaceous; D, Devonian, J, Jurassic, Ng, Neogene, P, Permian, Pg, Paleogene, Tr, Triassic.

‘… why have ferns evolved such high resistance to cavitation if they are predominantly confined to more mesic environments?’

Pittermann et al.’s study reveals some expected, and some highly unexpected, results. First, with few exceptions the average tracheid diameter in ferns was found to exceed even the largest diameters in conifers. Secondly, as evident in seed plant groups, a wide diversity of tracheid morphology and function was observed across the fern phylogeny (Fig. 2), with maximum conduit sizes observed in species with life-history strategies that require high xylem conductivities, such as in vines (Lygodium) and in pioneers (e.g. Pteridium aquilinum). This mirrors the pattern seen among extant angiosperms. The most unexpected result of the study was that, despite such a wide range of tracheid diameters, with many exceeding those of conifers, and in the absence of a torus-margo pit, the ability of the 16 sampled fern species to maintain xylem function under drought stress (through resistance to cavitation; i.e. seeding of the xylem with air) was found to be on a par with that of conifers. For conifers, an increase in conduit diameter of the tracheid generally results in decreased cavitation resistance. Pittermann et al.’s study shows that ferns have none of these trade-off issues. A simple explanation for these observations is that all extant gymnosperms, without exception, are woody (having secondary growth), implying that the evolution of xylem in gymnosperms must be examined in the context of three competing functions or a ‘trade-off triangle’ (Baas et al., 2004), including the functions of water transport, mechanical support and resistance to cavitation and embolism (breaking of the water column leading to xylem dysfunction). In ferns, which all lack true secondary growth, adaptive solutions are required to support only two primary functions, water transport and cavitation resistance, as a ring of thickened sclerenchyma fibres – the hypodermal sterome – provides mechanical support (Rowe et al., 2004).

The role of stomata?

If many ferns can efficiently transport water within their xylem and maintain stomatal conductance, at levels which match or exceed seed plants, and they can do so with huge conduit diameters without the risk of xylem dysfunction, then why are they typically out-competed by seed plants in drier environments? Pittermann et al. pose this important question but do not offer a fully satisfactory explanation. A possible explanation for this seemingly contradictory observation can be gleaned from emerging research on differences in stomatal function between spore-bearing (monilophytes and lycophytes) and seed-bearing (spermatophytes) vascular plants (Brodribb & Holbrook, 2004; Brodribb & McAdam, 2011). The evolution of xylem and that of stomata are tightly interconnected (Sperry, 2004). Considered simplistically, stomata primarily function to control water loss and CO2 uptake, and the xylem functions to control water supply. The functioning of both has to be optimized and integrated to conserve water loss during drought and maintain xylem function. All other traits being equal, therefore, a species with greater coordination of xylem and stomatal functions will probably out-compete a species with lesser coordination of the two. Brodribb & McAdam (2011) have shown that ferns and lycophytes can only passively control their stomata via the hydration state of the leaf and guard cells, whereas more derived seed plant lineages have evolved active control of stomatal opening and closing using the phytohormone abscisic acid (ABA). The broader ecological significance of this finding is that seed plants may have a greater capacity to actively control water loss during drought than do ferns. The suggestion is controversial, however, with other groups claiming that active control of stomatal opening and closure is ancestral for all land plants with stomata including ferns, because it is evident in the moss Physcomitrella patens (Chater et al., 2011) and the lycophyte genus Selaginella (Ruszala et al., 2011). Further work needs to be carried out to establish whether active ABA control in mosses and lycophytes has subsequently been lost in ferns and some lycophytes. An investigation of stomatal control and response to ABA in the 16 fern species studied by Pittermann et al. would be a good place to start.

Earlier work on the coordination of stomatal and xylem functions in cohabiting ferns and angiosperms showed that fern stomata shut to prevent xylem cavitation and associated dysfunction much earlier than the stomata of angiosperms (Brodribb & Holbrook, 2004). This conservatism inevitably would lead to lost photosynthetic potential among ferns compared with angiosperms, where stomata operated much closer to the critical point of xylem dysfunction. If Pittermann et al.’s study is considered in the light of these results, it suggests that differences in the coordination of stomatal and xylem functions between ferns and seed plant groups may be the reason why today ferns are more ecologically restricted. Of course, differences in reproductive biology, in particular the requirement for liquid water during fertilization in ferns, but not in seed plants, are also an important distinction.

Resistance to cavitation in ferns

Pittermann et al.’s study raises the equally intriguing question: why have ferns evolved such high resistance to cavitation if they are predominantly confined to more mesic environments? They argue that the degree of integration of the hydraulic networks in ferns may play a role in achieving high resistance to cavitation, with the least integrated having the highest resistance to cavitation and visa versa. As to the why, perhaps it is a redundancy issue; because ferns in general have lower functional xylem volumes than do gymnosperms, as demonstrated by the work of Pittermann et al., the fewer and often larger conduits they do have must be highly cavitation resistant. Perhaps it is a developmental constraint of only having passive rather than active control of stomatal function which selects for such high cavitation resistance? Selection for high cavitation resistance could also be simply driven by the fact that many wet habitats are often seasonally transient and prone to drying out at certain times of the year. Bracken (Pteridium aquilinum) and the royal fern (Osmunda regalis) have both adapted to such transience in water availability by being drought deciduous (Pittermann et al.), whereas other fern taxa have adapted through high cavitation resistance.

The results of Pittermann et al. clearly overturn the often oversimplified assumption, which, admittedly, this author and others have made, that functional traits, in this case those of the xylem, follow a trajectory of increasing sophistication across the plant phylogeny – that is, angiosperms > gymnosperms > ferns. Looking back over Earth history, ferns have indeed shown a pattern of decreasing relative abundance compared with gymnosperms and angiosperms; however, within the ecological niches that ferns occupy they have continued to diversify (Schuettpelz & Pryer, 2009) and persist at a rate that exceeds that of gymnosperms, leading to an estimated diversity of over 11 000 species in the modern flora. Pittermann et al.’s study has elucidated the versatility and surprising diversity of xylem morphology and functioning in ferns, which must have contributed significantly to this ancient group’s success.