Some evolutionary consequences of being a tree (extract)
RemyJ. Petit (1) and Arndt Hampe (1, 2)
(1) Institut National de la Recherche Agronomique, UMR Biodiversity,
Genes and Communities, F-33610 Cestas, France; email: firstname.lastname@example.org
(2) Consejo Superior de Investigaciones Científicas, Estacion Biologica de Donana, Integrative Ecology Group, E-41080 Sevilla, Spain; email: email@example.com
Published by The Annual Review of Ecology, Evolution, and Systematics. 2006. 37:187-214
allometric scaling, evolutionary rate, gene flow, growth form, life history
Los árboles no forman un grupo natural, pero comparten atributos tales como gran tamaño, longevidad y alto rendimiento reproductivo que afectan su modo y ritmo de evolución. En particular, los árboles son únicos en el sentido de que ellos mantienen altos niveles de diversidad, mientras que la acumulación de nuevas mutaciones sólo ocurre muy lentamente. También son capaces de una rápida adaptación local y puede evolucionar rápidamente de antepasados no arbóreos, pero la mayoría de los linajes de árboles típicamente experimentan baja especiación y tasas de extinción. Se discute por qué el hábito de crecimiento del árbol debería llevar a estas características aparentemente paradójicas.
La importancia de los árboles para mantener la vida en general y la biodiversidad, en particular, difícilmente puede exagerarse. Se estima que el 27% de la superficie terrestre de la Tierra está (todavía) cubierta por bosques (FAO World Resources 2000-2001), y los árboles constituyen aproximadamente el 90% de la biomasa de la Tierra (Whittaker 1975). No es de extrañar que los bosques también albergan la inmensa mayoría del mundo de la biodiversidad terrestre. Las estimaciones mundiales de la riqueza de especies de árboles van de un mínimo 60,000 (Grandtner 2005) a 100,000 taxones (Oldfield et al. 1998), es decir, tanto como el 15% y el 25% de la 350,000-450,000 plantas vasculares del mundo (Escocia & Wortley 2004). Por desgracia, la deforestación en curso (estimada en 9,4 millones de hectáreas por año en el decenio de 1990) y otros inducidos por el hombre han traído cambios y más del 10% de las especies de árboles están en peligro de extinción (Oldfield et al. 1998). El impacto del cambio global dependerá en gran medida de la reacción de los árboles y los ecosistemas que sustentan (por ejemplo, Ozanne et al. 2003; Petit et al. 2004a, 2005b). La mitigación de estas consecuencias perjudiciales requiere el conocimiento de la biodiversidad de árboles y la evolución. Sin embargo, los árboles no sólo son sobreexplotados sino también subestudiados en muchos aspectos, debido a su tamaño y período de vida que los hace difíciles como temas de las investigaciones experimentales (Linhart 1999).
El hábito de crecimiento de los árboles ha evolucionado muchas veces. Esta es probablemente la razón por la cual pocos intentos se han hecho en los últimos decenios a considerar colectivamente los árboles y discutir su modo de evolución. Esta aparente falta de interés contrasta con una fuerte tradición en años anteriores (por ejemplo, Arber 1928; Clarke 1894; Grant 1963, 1975; Sinnott 1916; Stebbins 1958). El actual interés en la biología comparativa, gracias al desarrollo de filogenias precisas y poderosos métodos analíticos, debería ayudar a revitalizar esta tradición. Lejos de representar un problema, los múltiples orígenes de árboles en realidad facilitarán este trabajo, como cada linaje de árbol puede ser visto como un experimento evolutivo independiente. Análisis comparativos deberían ayudar a dilucidar si características típicas de un árbol como altura, longevidad, fecundidad afectan a su dinámica evolutiva.
Desde un punto de vista evolutivo, los árboles tienen varias intrigantes y aparentemente paradójicas características. En particular, a menudo tienen altos niveles de diversidad genética, pero la experiencia de bajos tipos de nucleótidos y la sustitución de bajas tasas de especiación. También se combinan alta diferenciación local para los rasgos de adaptación con un ampli flujo de genes. Por otra parte, el excepcional mantenimiento de la integridad de especies de cara de abundante flujo de genes interespecíficos parece ser la regla en los árboles.
Trees do not form a natural group but share attributes such as great size, longevity, and high reproductive output that affect their mode and tempo of evolution. In particular, trees are unique in that they maintain high levels of diversity while accumulating new mutations only slowly. They are also capable of rapid local adaptation and can evolve quickly from nontree ancestors, but most existing tree lineages typically experience low speciation and extinction rates. We discuss why the tree growth habit should lead to these seemingly paradoxical features.
The importance of trees for sustaining life in general and biodiversity in particular can hardly be overstated. An estimated 27% of the terrestrial surface of Earth is (still) covered by forests (FAO World Resources 2000-2001), and trees make up around 90% of Earth’s biomass (Whittaker 1975). Not surprisingly, forests also harbor the vast majority of the world’s terrestrial biodiversity. Estimates of global tree species richness range from a low 60,000 (Grandtner 2005) to 100,000 taxa (Oldfield et al. 1998), that is, as much as 15% to 25% of the 350,000-450,000 vascular plants of the world (Scotland & Wortley 2004). Unfortunately, ongoing deforestation (estimated at 9.4 million hectares per year in the 1990s) and other human-induced changes have brought >10% of the world’s tree species close to extinction (Oldfield et al. 1998). The impact of global change will depend to a great extent on the reaction of trees and the ecosystems they sustain (e.g., Ozanne et al. 2003; Petit et al. 2004a, 2005b). Mitigating these harmful consequences requires knowledge of tree biodiversity and evolution. However, trees are not only overexploited but also understudied in many respects, because their size and life span make them difficult subjects for experimental investigations (Linhart 1999).
The tree growth habit has evolved many times. This is probably the reason why few attempts have been made over the past several decades to consider trees collectively and discuss their mode of evolution. This apparent lack of interest contrasts with a strong tradition in earlier years (e.g., Arber 1928; Clarke 1894; Grant 1963, 1975; Sinnott 1916; Stebbins 1958). The current interest in comparative biology, thanks to the development of accurate phylogenies and powerful analytical methods, should help revive this tradition. Far from representing a problem, the multiple origins of trees will actually facilitate this work, as each distinct tree lineage can be viewed as an independent evolutionary experiment. Comparative analyses should help elucidate if typical tree features such as tallness, longevity, and fecundity affect their evolutionary dynamics.
From an evolutionary standpoint, trees have several intriguing and apparently paradoxical features. In particular, they often have high levels of genetic diversity but experience low nucleotide substitution rates and low speciation rates. They also combine high local differentiation for adaptive traits with extensive gene flow. Moreover, exceptional maintenance of species integrity in the face of abundant interspecific gene flow seems to be the rule in trees.
2.1. What Is a Tree?
With one known exception-Prototaxites, a 9-m-high tree-like holobasidiomycete or lichen that dominated the land flora 3 50-400 Myr ago (Hueber 2001, Selosse 2002)- all organisms ever considered to be trees are vascular plants (tracheophytes). As such, they share features such as indefinite and flexible growth, modular structure, lack of clear separation between germline and soma, reversible cellular differentiation, great phenotypic plasticity and physiological tolerance, and presence of haploid and diploid multicellular generations (Bradshaw 1972). Evolution of trees cannot be understood without due consideration of these attributes.
The particular character of the tree growth form has always been recognized and, since Theophrastus (born c. 370 EC), botanists have generally distinguished between trees, shrubs, and herbs. From a functional point of view, trees share a number of features, such as large size, long life span, and a self-supporting woody perennial trunk, but not one is really exclusive. According to Van Valen (1975), a tree is, in the ecological sense, “any tall woody plant.” However, trees are generally distinguished from shrubs and vines, so most researchers prefer to be more specific. For instance, for Thomas (2000), “a tree is any plant with a self-supporting, perennial woody stem”; for Donoghue (2005) the tree growth habit is characterized by “tall plants, with a thickened single trunk, branching well above ground level”; and for Niklas (1997) a tree is “any perennial plant with a permanent, woody, self-supporting main stem or trunk, ordinarily growing to a considerable height, and usually developing branches at some distance above the ground.” The modulations introduced express the need to accommodate situations where plants generally considered to be trees adopt unusual habit or size in some environments. Finally, somewhat arbitrary definitions can be found in the forestry literature, for inventory purposes: “Trees are woody plants with one erect perennial stem, a definitely formed crown, a height of at least 4 m and a stem diameter at breast height of at least 5 cm” (Little 1979).
The presence of wood is sometimes taken as an argument to circumscribe trees to die lignophytes (see Niklas 1997). Interestingly, recent molecular genetic and genomic studies in Populus and Arabidopsis have shown that the genes responsible for cambium function and woody growth are not unique to woody plants: Genes involved in the vascular cambium of woody plants are also expressed in the regulation of the shoot apical meristem of Arabidopsis (Groover 2005). This might explain why woodiness can evolve so readily (as observed in many island radiations; e.g., Bohle et al. 1996, Carlquist 1974) and led Groover (2005) to conclude that forest trees “constitute a contrived group of plants that have more in common with herbaceous relatives than we foresters like to admit.”
According to Arber (1928), one needs to go beyond textbook definitions and acknowledge that the difference between trees and other plants is mostly a question of scale. Below, we provide an account of the prominent features of the tree growth habit from an ecological standpoint. In so doing, we follow Arber (1928) and stress questions of scale and allometry.
2.2. Prominent Tree Features
The woody habit involves a series of ecological benefits and constraints that have contributed to the dominance of trees across many ecosystems worldwide and to their scarcity or complete absence from others. According to Harper (1977, p. 599), the major advantage of a woody growth habit is that “it can give perenniality to height.” These two components are tightly linked, as a high stature can obviously not be attained without the corresponding life span. Tallness and longevity are also the prerequisites for another central feature of trees: their large, sometimes huge lifetime reproductive output, despite a somewhat delayed maturity.
Although it is clear that these characteristics have been molded by selection pressures (Niklas 1997), they are subject to a diversity of anatomical, physiological, or ontogenetic constraints (e.g., Mencuccini et al. 2005, Niklas 1997, Rowe & Speck 2005, Silvertown et al. 2001). Major steps to understand the primary causes of the evolution of the tree growth habit have been made by simulating adaptive walks through the morphospace of early vascular land plants (Niklas 1997). These studies indicate that growing tall is indeed an adaptive process; in particular, “tree-like morphologies bearing lateral planated branching systems or foliage leaves occupy adaptive peaks” (Niklas 1997). Altogether, the tree growth form can be viewed as an integrated ecological strategy involving many trade-offs (Table 1). In the following we discuss implications of the tree habit and outline the major characteristic of trees’ life cycle.
2.2.1. Tallness. Trees grow tall where resources are abundant, stresses are minor, and competition for light takes place (e.g., Falster & Westoby 2003, King 1990, Loehle 2000). Large size enables them to create a physical and chemical environment that influences their own performance and that of interacting organisms (e.g., Boege & Marquis 2005, Herwitz et al. 2000, Ricklefs & Latham 1992). High stature helps mitigate the effects of disturbances that take place primarily at ground level, such as grazing and trampling by large herbivores or fires (Ordonez et al. 2005), butitmakes trees highly susceptible to other disturbances such as wind (Gutschick & BassiriRad 2003, Loehle 1988, Rowe & Speck 2005). Growing tall requires the development of resistant supporting and protective tissues. This generates high costs of maintenance, reduces growth rates, and limits the existence of trees to areas that provide a minimum long-term input of energy, water, and nutrients (Ward et al. 2 005a, Wardle et al. 2004).
Tall stature also tends to increase gene flow. For instance, the height of release is often (yet not necessarily) related to the median transport distance of wind-dispersed pollen and seeds (Nathan et al. 2002, Okubo & Levin 1989, Portnoy & Willson 1993). Tall, conspicuous plants with large flower or fruit displays also tend to attract disproportionately many animal pollinators and seed dispersers (Ghazoul 2005). The latter holds likewise for antagonists such as herbivores or pathogens, however.
Height is probably the plant trait that has most often been included in comparative studies. Westoby et al. (2002) consider it a leading dimension of plant ecological strategies, as it conveys knowledge on many other aspects of species’ ecology. The upper limit to plant height has been the object of many studies and debate that will not be reviewed in detail here. Two major hypotheses coexist, the respiration hypothesis and the hydraulic-limitation hypothesis (e.g., Mencuccini et al. 2005). However, evolution should not lead to a single limiting factor, according to the principle of equalization of marginal returns on alternative expenditures (Westoby et al. 2002). Hence, species are not expected to grow as tall as physically possible because of various trade-offs, for example with reproduction, wood density (and hence longevity), or leaf mass (e.g., Loehle 1988).
2.2.2. Extended life cycle. Although great advances have been made in our understanding of initial recruitment processes in forests and in the modeling of vegetation dynamics (e.g., Clark et al. 1999, Loehle 2000), extremely few studies have considered the entire tree life cycle. A notable exception is the work of Van Valen (1975), who presented the first complete life table for a tree, the tropical palm Euterpe globosa. According to his computations, only one seed in one million produces a shoot that reaches the canopy in this species. The resulting demographic curve, when expressed in logarithmic terms, is highly convex, contrary to that of many large animals, as first pointed out by Szabo (1931) (Figure 1). In this palm species, generation time was estimated to be 101 years, a value intermediate between age at maturity (50 years) and maximum observed life span (156 years). These values underline the need to distinguish between age at maturity, generation time, and life span, which are often inappropriately used interchangeably in the literature.
2.2.3. Seed production. Estimates of lifetime reproductive output for trees are rare (Moles et al. 2004), but it is clear that many trees produce prodigious numbers of seeds. Reproduction is costly and trade-offs with vegetative growth are well-known (e.g., Obeso 2002). Niklas & Enquist (2 003) proposed an allometric model for reproduction in seed plants that shows that the annual reproductive biomass scales with the two-thirds power of the standing shoot biomass; in other words, allocation to reproduction decreases size. Hence, larger plants would produce comparatively fewer seeds if seed size scaled isometrically with plant size. Aarssen (2005) tested this latter relation for 600 North American species and found that seed length increases only at about half the rate of plant height, indicating that the prevalent evolutionary trend (i.e., the deviation from allometric scaling) is toward comparatively smaller seeds, thus maintaining fecundity at the expense of seed provisioning. Moreover, the variability in seed length grows disproportionately with plant height. Similar results were obtained by Moles et al. (2004) for plant and seed mass. Aarssen (2005) argued that the observed patterns might be a simple consequence of the fact that the spectrum of possible seed sizes broadens with plant size.
Contrary to animals, plant fecundity usually increases more or less continually through an individual’s life (Franco & Silvertown 1996). Hence, the lifetime seed production of trees is typically orders-of-magnitude greater than that of herbs, even though decreasing allocation results in a lower annual output of seeds per unit of canopy (Moles et al. 2004). Unfortunately, as for many other relationships, it remains unclear if individual variation in lifetime fecundity is greater in trees than in herbs or if it scales isometrically.
Finally, much attention has been paid to the phenomenon of mast seeding in trees (i.e., the synchronous intermittent production of large seed crops). Overall, it appears that fruit crop size scales positively with its among-year variability (Kerkhoff & Ballantyne 2003). But it remains unclear whether this phenomenon results mostly from weather conditions or represents an evolved plant reproductive strategy to improve pollination efficiency and outcrossing levels (in wind-pollinated species), and/or to increase offspring survival through predator satiation. Recent meta-analyses of extensive data sets indicate that both components may be involved to varying extents (Kelly & Sork 2002).
2.2.4. Establishment. As trees tend to live in comparatively stable habitats and generation turnover is slow, only an extremely small fraction of the seeds produced during an individual’s lifetime will eventually survive to maturity. This has important consequences for trees’ evolution. First, the considerable selection potential during early life stages should favor local adaptation of recruits, particularly for traits that enhance competitive ability (such as early growth and delayed maturity). By contrast, selective culling during trees’ establishment appears to have little influence on population demography (Franco & Silvertown 1996). Second, because much of the density-dependent mortality takes place before maturity in trees, their effective population size should be closer to the actual adult census size compared to herbs, contributing to preserve genetic diversity (Dodd & Silvertown 2000).
2.2.5. Age at maturity. One classical trade-off in population dynamics is that between early growth and age at maturity. Precocity of reproduction has a great influence on the potential growth rate of a population (Harper 1977). Only very stringent competition for resources (e.g., light) during the early life of trees can select for delayed maturity. Among trees, there is a great variation in age at maturity. Woody angiosperms tend to reproduce sooner than gymnosperms [modal class is 1-5 years compared to 6-20 years (Verdii 2002)]. Age at maturity has received some attention by molecular biologists. Genetic manipulations demonstrate that juvenile trees can be induced to flower by modifying the expression of a single gene, e.g., LFFin trans-genie poplars (reviewed in Martin-Trillo & Martinez-Zapater 2002). Hence, as for secondary growth, the evolution of shortened maturity does not require profound genetic changes at the molecular level. (The converse is not necessarily true, however the evolution of delayed maturity might be more complex.)
2.2.6. Longevity. A long life span is favored in stable habitats as long as it remains advantageous to allocate resources to future reproduction. Great longevity provides several obvious advantages. First, once successfully established, plants can endure periods of environmental stress while taking advantage of relatively short pulses of less harsh conditions. In particular, long-lived species can endure periodic reproductive failures without direct negative demographic consequences (Ashman et al. 2004 Calvo & Horvitz 1990). This flexibility might explain why woody plants generally display stronger pollen limitation than herbs (Knight et al. 2000). Second, spreading reproduction over many years boosts lifetime reproductive output. However, a long life span also means that individuals have to cope with variable environmental conditions including catastrophic events (Gutschick & BassiriRad 2003). Hence, allocations to growth, reproduction, and survival need to be adjusted throughout lifetime. Such plasticity would in turn contribute to enlarge trees’ potential habitats (e.g., Hampe & Bairlein 2000, Jonsson 2002), resulting in considerable buffering against extinction (Hampe & Petit 2005).
2.2.7. Senescence. The extreme longevity observed in woody plants makes them useful models for senescence research and trees have actually started to attract the interest of gerontologists (e.g., Flanary & Kletetschka 2005, Lanner 2002, Larson 2001). As pointed out by Williams (1957), the degree of senescence is a function of the lifetime distribution of reproductive effort, so senescence should be far lower in organisms that increase reproduction with age, like trees.
Extreme conditions (e.g., low temperatures, drought or wind) are associated with the occurrence of particularly old and slow-growing trees (e.g., Laberge et al. 2000, Lanner 2002, Larson et al. 1999), suggesting that low metabolism contributes to their delayed senescence. Until recently, it was generally assumed that whole-organism metabolic rate scales with the three-fourths power of body mass in all organisms (Gillooly et al. 2001). Hence, trees would inherently experience reduced metabolic rates simply owing to their size. However, Reich et al. (2006) have shown that the metabolic rate of plants (including herbs, woody plant seedlings and young saplings) instead scales approximately isometrically with plant size, thereby discarding allome-try as a possible source of reduced metabolic rate in trees. Nevertheless, the remarkable amount of resources that woody plants need to invest in supporting structures and defenses (such as a thick bark or defensive chemicals) is generally related to a reduction of growth rate and, hence, of metabolism (Loehle 1988).
Trees dispose of a suite of active and passive mechanisms to repair, isolate, or re-olace deteriorated tissues (Loehle 1988). These can greatly increase life span thanks to the modular structure of plant growth and to the fact that at least some cell lines inside meristems retain the juvenile ability to contribute to new growth (Lanner 7002). Low extrinsic mortality and efficient repair mechanisms would promote resource allocation to repair (especially early in life), resulting in delayed growth rate and maturity, large size, and a dramatic increase in survival and maximum life span (Cichon 1997). Empirical support for this notion comes from a demographic analysis of herbaceous and woody plants (Silvertown et al. 2001) that detected increasing age-specific mortality near the maximum life span (that is, signs of senescence) only in the longest-lived species. So far, however, little evidence exists for whole-tree senescence in terms of changes in gene expression that might indicate genetically controlled aging mechanisms (Diego et al. 2004), although there are preliminary data indicating that both telomere length and telomerase activity could be involved in tree longevity (Flanary & Kletetschka 2005).
Edición Digital: Centro de Estudios de Recursos Bióticos, Universidad de Panamá