Reductions in photosynthetic and growth rates, enhanced foliar photo-oxidative stress and decreased vigour of flower organs are processes associated with aging at the organ level in perennial plants. However, such events are not necessarily indicative of physiological deterioration at the organism level, as occurs in humans and other animals. The combination of modular growth and dormancy in perennial plants, two features individually shared with some animals (Fig. 1), increases their plasticity and strongly reduces the potential damage caused by aerobic life, to the extreme that the likelihood of dying from aging is, at best, negligible.
Modular structure and meristem dormancy
In contrast to most animals, plants have a tremendous plasticity in the form and function of their organs (Walbot, 1996). This is particularly relevant in perennials, in which the apical meristem of at least one of their shoots remains indeterminate beyond its first phase of growth and development. Perennial plants organize their body growth in the vertical plane, based on the division and differentiation of meristems. This structure allows perennial plants to explore the environment vertically in the search for light, and enables some of them to be very large and to have a very long life span. Some trees such as sequoias and pines, among others, can grow up to 100 m in height and survive for centuries and even millennia (Peñuelas, 2005; Munné-Bosch, 2007, 2008). Other species, such as some herbaceous perennials, can also survive for centuries by keeping their meristems intact (García et al., 2008). This lifestyle is determined already at the embryo stage, in which apical (root and shoot) meristems are formed but remain inactive. After some days, weeks, or even years, seeds break dormancy and the seedling germinates. During postembryonic development, modular growth is combined with long periods of dormancy, during which growth from meristems (and other organs and cells with the capacity to resume growth) is not initiated under favourable conditions (Rhode & Bhalerao, 2007). A long life span in perennials is therefore possible.
Regeneration of organs and organisms in animals and plants
A single haematopoietic stem cell can reconstitute the entire haematopoietic system of an irradiated mammal. Progenitor cell populations have been identified within most mammalian organs, including skin, blood, bone, reproductive tissues, skeletal muscle, kidney, lung, liver, intestine, heart and brain (Wagers et al., 2002; Shizuru et al., 2005). Nevertheless, mammals do not possess the capacity to regenerate entire normal organ function after major damage to, or removal of, structures. By contrast, some animals, such as zebrafish, can restore normal organ (heart) function after severe injury (Jopling et al., 2010), and a few species of animals, such as some marine clonal invertebrates (e.g. many cnidarians) (Carter et al., 2010; Bely & Nyberg, 2010), can even regenerate their entire body as a result of their modular development, a trait shared with plants.
Despite these similarities between the animal and plant kingdoms in their regeneration mechanisms, the cellular plasticity in plants seems to be more responsive to extracellular signals. Plant cells can rapidly switch fate and be reprogrammed for regeneration in a process that is a survival strategy for plants (Costa & Shaw, 2006). Furthermore, in the invertebrate crinoid echinoderms and ascidians, which are phylogenically more closely related to mammals, the capacity to regenerate is limited to specific organs and tissues (Bimbaum & Sánchez-Alvarado, 2008). This underlines the fact that only the invertebrate cnidarians and other phylogenically basal members of the animal kingdom show whole-organism regeneration capacity, representing an important link in the interpretation of the common regeneration mechanisms between animals and plants.
Modularity seems to increase organism plasticity in terms of whole-organism regeneration, a property that has been maintained with increasing complexity during evolution of species over time within the plant kingdom, but not within the animal kingdom. However, maintaining organism integrity with increased complexity in an evolutionary timescale has some costs. In a meta-analysis of plant life-history tables, data provided by Silvertown et al. (2001) support the idea that organisms which remain integrated in modules are more likely to senesce than organisms whose ramets separate into unconnected entities.
Agricultural and gardening practices continuously show that not only are entire plant shoots restored, but also that their growth and fruit production is increased after pruning. The only secret is to leave some vegetative meristems alive. Then, less competition for resources within the plant does the rest. The organization of plants in a modular structure that is based on producing a set of vegetative meristems above- and belowground not only optimizes light capture from the sun and water/nutrient absorption from the soil, but also permits the complete restoration of fully damaged structures. Furthermore, while some meristems sustain vegetative or reproductive development during a growing/reproductive season, others are kept dormant. Similarly, some animals, such as turtles, have periods of extended dormancy accompanied by substantial metabolic depression. This prevents age-inflicted damage over long periods of time. In perennial plants, entire modules can be dormant for long periods of time and then emerge after an incident of environmental stress, pathogen attack or mechanical injury to other plant parts. In that sense, Shefferson (2009), while evaluating different features of long-lived and short-lived plants in relation to plant size and reproduction, provided evidence for a relationship between vegetative dormancy and life span. In particular, he proposed that dormancy is an age-related response to stressful environmental conditions, increasing the life span of plants.
Senescence vs ageing; organ vs organism
As plants age, and despite modular development and dormancy, organs such as the leaves and flowers of the oldest individuals can show signs of senescence and of physiological deterioration, such as reduced photosynthetic rates, reduced growth, increased photo-oxidative stress or loss of flower bud vigor (Mencuccini et al., 2005; Munné-Bosch & Lalueza, 2007; Vanderklein et al., 2007; Oñate & Munné-Bosch, 2010). With increasing age and size, plant growth tends to slow and one may expect that, as a result, a tree is more likely to die. This reduced growth could result from intrinsic damage to meristems, under direct genetic and age-related control. It seems, however, that potential damage to meristems as a result of age-associated mutations is negligible, because no evidence of accumulation of somatic mutations in meristems was found in a study of bristlecone pines (Pinus longaeva) that ranged from 23 to 4713 yr of age (Lanner & Connor, 2001). Special cases of even longer life span have been reported in clonal species, such as Lomatia tasmanica, an endangered species that exclusively reproduces asexually and whose clonal reproduction has been dated to an age of at least 43 600 yr (Lynch et al., 1998).
Another explanation for reduced growth with age comes from the physiological burdens conferred by size, such as the demands on water and nutrient supply associated with increasing height and girth. Results of grafting studies add experimental support for this second view, by showing that age-related increases in size limit growth rates in trees (Mencuccini et al., 2005; Peñuelas, 2005; Vanderklein et al., 2007). It appears therefore that the reduced physiological activity at the organ level, associated with aging, is mainly the result of increased size and the subsequent enhanced competition for resources between modular structures (Mencuccini et al., 2005; Munné-Bosch & Lalueza, 2007; Vanderklein et al., 2007; Oñate & Munné-Bosch, 2010).
Even if these size effects are not the only reason for age-related changes at the organ level, such changes do not necessarily lead to physiological deterioration with aging at the organism level. In fact, only a few studies have reported increased mortality rates associated with senescence at the organism level (Picó & Retana, 2008; Herrera & Jovani, 2010). Moreover, in these studies the aging effect might be masked by changing climatic conditions or plant size over the years (Kirkpatrick, 1984; Vaupel et al., 2004). In fact, it has recently been reported that regional warming and consequent increases in water deficit, together with biotic stress, rather than aging, are probable contributors to the increases in tree mortality rates in the western USA (van Mantgem et al., 2009) in the southern European regions (Peñuelas et al., 2001) and globally (Allen et al., 2010). In addition, the changes that take place during typical succession at the ecosystem level, for example in soil nutrient cycles, with the locking of nitrogen (N) and phosphorus (P) in biomass and the consequent gradual reduction in available N and P in soils over time, could be confounded with senescence at the individual level. Furthermore, no study has demonstrated thus far that increased mortality at old age is associated with a physiological deterioration of meristem function in plants. Some studies of flower and seed production even suggest that fecundity increases with age (size-related increases) or is kept constant at advanced developmental stages, and that it very rarely decreases with age (Kirkpatrick, 1984; Hansen et al., 1992;Vaupel et al., 2004; Oñate & Munné-Bosch, 2010).
In most cases, mortality in perennials seems therefore not to be caused by the progressive deterioration of physiological functions associated with age, as occurs in humans. The unique combination of modular development and dormancy evolved by perennial plants allows them to defy aging. The field of aging and regenerative biomedicine could find inspiration in this combination, which makes the probability of death owing to external factors higher than death caused by physiological deterioration associated with aging at the plant organism level.
Support for the research was received through grants CGL2006-04025/BOS, BFU2006-01127, BFU2009-07294, BFU2009-06045 and Consolider-Ingenio Montes CSD2008-00040 from the Ministry of Science and Innovation of the Spanish Government; grants A/019625/08 and A/016255/08 from the AECI; and the ICREA Academia prize given to S.M. and the grant SGR2009/458 given to J.P., both funded by the Generalitat de Catalunya. We are very grateful to three anonymous reviewers for their insightful comments on the manuscript.