Senescence as a trade-off between successful land colonisation and longevity: critical review and analysis of a hypothesis

Survey methodology

The analysis included in this review was carried out on the basis of kingdom Animalia. To avoid bias in the selection of species, we referred to comprehensive studies that summarised life expectancy patterns across animal clades. We obtained data on vertebrates from a study by Carey & Judge (2000). Data on the life expectancy of invertebrates were obtained from multiple reviews (Giraldo et al., 2016; Wanninger, 2015; Ward, 1992). Additionally, we used data from AnAge, which is a comprehensive database of animal longevity records and related life-history traits (available online: http://genomics.senescence.info/species/). AnAge was compiled from the literature and large-scale datasets (De Magalhaes & Costa, 2009), and currently features 4,244 entries. Data on maximal lifespan available at https://www.demogr.mpg.de/longevityrecords/ were also used. The records of the Maximum Lifespan (MLS) values of animal species belonging to the main clades were verified against previous studies and used to construct a database for further analyses. We split MLSs into three classes: MLS<5 years (short), 5 ≤MLS<35 years (intermediate), and MLS ≥35 years (long). Each clade was assigned to an MLS category based on their predominant MLS phenotypes (Table S1). This maximum lifespan categorisations into three classes is arbitrary.

Next, we distinguished subgroups within the clades containing secondary aquatic species. Accordingly, mammals were divided into aquatic mammals and other mammals, and reptiles were divided into aquatic reptiles and other reptiles. In order to take into account possible differences between typical soaring birds with reduced energy costs of flight and more land-bound birds, we also created subgroups of soaring birds and other birds. The percentage of long-living species among all analysed species in these different groups of animals is shown in Fig. 1. Studies of evolutionary correlations commonly use phylogenetic regression (i.e., phylogenetic generalized least squares) to assess trait covariation in a phylogenetic context. However, there are opinion that while this approach is appropriate for evaluating trends in one or a few traits, it is incapable of assessing patterns in highly multivariate data, as the large number of variables relative to sample size prohibits parametric test statistics from being computed (Adams, 2014). In this article, we use a statistical procedure for performing non-parametric ANOVA (Kruskal-Wallis test), that can accommodate our datasets. Average MLS was compared across animal groups using the non-parametric ANOVA and Dunn’s post-hoc tests (Fig. 2). MLS effects (MLS categories) were compared separately for each pair of animal clades in thirty six 2 ×3 contingency tables from which chi square values were calculated (Zar, 2010; Fig. 3).

We reviewed literature (1950–2020) on the life expectancy, the existence of aging symptoms in animals, the described causes of death, as well as the adaptive characteristics of the main groups of animals. We searched the following publishers and databases: Elsevier’s ScienceDirect, PubMed, Springer’s SpringerLink, Wiley Online Library, Google Scholar, and NCBI Bookshelf (Table S2).

Publications where the searched term sets were found in the title, keywords, or abstract, met the inclusion criteria, and these articles were then analysed at the full-text level. All relevant studies that met the search criteria were included in this review. Studies using the same data from various publications were eliminated to avoid redundancy.

Three sets of words were used in the literature search. The senescence set consisted of words related to senescence, and included ageing, aging, damage, death, decrease in fertility, life, life program, life-span, lifespan, longevity, loss of function, mortality, non-senescence, repair, regeneration, rejuvenation, and senescence. The animal features set consisted of words related to the adaptive features of animals, including ability, activity, adaptation, agametic, body size, breathing, burrowing, development, clonal, colonial, ecology, extinction, environment, evolution, factor, feature, fecundity, flight, flying, gas exchange, growth, land, locomotion, maturity, metamorphosis, offspring, oxygen, physiology, process, reproduction, sexual, strategy, survival, swimming, temperature, terrestrial, walking, and water. The animal clades set consisted of the names of clades and groups of animals, and included animals, amphibians, annelids, birds, chordates, cnidarians, crustaceans, dinosaurs, echinoderms, fish, gastropods, hexapods, insects, invertebrates, jawed fish, jawless fish, lancelets, mammals, metazoans, molluscs, myriapods, porifers, ribbon-worms, reptiles, sauropsids, slugs, snails, spiders, synapsids, tetrapods, tunicates, and vertebrates.

Literature search was first carried out using all binary combinations of each of the words in the senescence and animal clades sets to identify the general life expectancy patterns of the main groups of animals, the occurrence of senescence symptoms in animals, and the reported causes of death. Next, all binary combinations of each of the words in the animal features and animal clades sets were searched to identify the most important adaptive features of the animal clades.

We then synthesised the data to present a generalised relationship between senescence and life cycles of animals. This may be connected to the ability of cell replacement and regeneration, while considering the biology of the animals and the corresponding environmental factors. Furthermore, we constructed a model showing phylogenetic relationships between animal clades. In the context of the major stages of evolution, we identified the main “senescent” and “biologically immortal” clades of animals. Topology was based on the Tree of Life (ToL) Web Project (http://www.tolweb.org) in addition to many studies from the literature that were used to resolve uncertainty in the ToL project phylogeny (Fig. 4).

Finally, we juxtapose our synthesis of the literature review with the most important concepts and theories explaining the origin and mechanisms of senescence. We present our conclusions and hypotheses that provide an explanation of the causes of senescence symptoms and attempt to indicate their genesis in the main clades of terrestrial animals and the relationship with life expectancy and maximal lifespan.

We discuss mainly three clearly senescing groups of terrestrial animals, that is, insects, birds, and mammals. This choice was based on their dominance in terrestrial environments. Hence, we tried to address the following questions: Why do these three clades, which are so successful on land and yet so different, show symptoms of senescence? Could senescence be adaptive? In addition, can the secondary aquatic reptiles and mammals live longer than their relatives that have remained on land?

Animals with different degrees of senescence

No evolutionary mechanisms propose the programmed death of the individual; instead, senescence is thought to be a consequence of reduction the effectiveness of natural selection with age (see ‘the selection shadow theory’ (Flatt & Partridge, 2018; Gavrilova et al., 1998; Medawar, 1946). Partridge and Barton advocated that the distinction between parent and offspring is a necessary condition for senescence (Partridge & Barton, 1993).

Reproduction and growth vs. senescence

Animals can be divided into two main groups based on whether they are terminally differentiated in terms of sexual reproduction. The first group contains non-senescent, “biologically immortal” forms (at least in theory, because it can be found papers showing senescence in the asexually reproducing marine oligochaete Paranais litoralis, where no distinction between soma and germline (Martínez & Levinton, 1992), which include the simplest animal forms such as cnidarians that reproduce agametically (Reitzel, Stefanik & Finnerty, 2011). The second group contains forms that are irreversibly specialised in sexual reproduction (Sawada, Inoue & Iwano, 2014). Among these animals, some forms (birds) may exhibit symptoms of senescence while others do not (fish). The senescence is observed in the case when an organism switches off the expression of existing growth and regeneration programs (e.g., imago formation in insect development) or when particular programs of growth and regeneration of progenitors are irreversibly lost (e.g., in mammals and birds). In turn, maintaining regenerative capacity prevents or reduce appearance the symptoms of senescence, what can be observed in species, which show ability to continuous growth (Bilinski, Bylak & Zadrag-Tecza, 2016; Bilinski, Bylak & Zadrag-Tecza, 2017). The analysis shows that aging patterns which includes increasing, constant, and decreasing mortality with aging differ between species depending on their ability to determinate or continuous (indeterminate) growth. Species with indeterminate growth, including some animals and most higher plants, seem to escape the senescence, most probably due to their capacity of these species to continuous growth (Jones et al., 2014; Munné-Bosch, 2015). It is connected with Weismman conception of soma and germ-line and disposable soma theory proposed by Kirkwood (Kirkwood & Holliday, 1979). Active program of cell replacement and regeneration allows for the maintenance of the soma and by that occurrence the negligible or negative senescence.

Symptoms of senescence

Symptoms of senescence include gradual deterioration in function (Niccoli & Partridge, 2012), morphological and physiological changes at cellular (Guerville et al., 2020; Herranz & Gil, 2018; Munoz-Espin & Serrano, 2014) and organismal levels (Dodig, Čepelak & Pavić, 2019), phenotypic changes (McHugh & Gil, 2018; van Deursen, 2014), a reduction in fecundity (Ericsson et al., 2001), and increased risk of age-related mortality (Bilinski, Paszkiewicz & Zadrag-Tecza, 2015).

However, senescence symptoms are not observed in groups of animals such as fish or reptiles, which continue to increase their body size after reaching sexual maturity (Bilinski, Bylak & Zadrag-Tecza, 2016; Speakman, 2005). Although there is some evidence of senescence in fish and reptiles (Gerking, 1957; Shefferson, Mizuta & Hutchings, 2017), negligible senescence phenotypes predominate in these clades (Carey & Judge, 2000; Froese & Pauly, 2019; Woodward, Horner & Farlow, 2011). In turn, considering the lifespan among vertebrates, fish represent the most numerous species that live over 100 years (Reznick, Ghalambor & Nunney, 2002).

On the other hand, symptoms of senescence are encountered in groups such as birds (Ricklefs, 2010) and mammals (Broussard et al., 2003), in which body growth ceases as soon as the animal reaches sexual maturity. Alternatively, senescence may emerge in insects and related species during the fundamental reconstruction of an organism such as during complete metamorphosis (Abdelwahab et al., 2018; Nöthiger, 1972).

Regenerative potential vs. senescence

Continuous growth is accompanied by high regeneration and cell replacement capabilities (Bilinski, Bylak & Zadrag-Tecza, 2016). However, even if animals do not grow continuously, some species retain the ability for continuous growth and regeneration in individual parts of the body (Alvarado, 2000). For example, female tarantulas moult throughout their whole life (Costa & Perez-Miles, 2002) and some parts of the human body such as ears, grow continuously (Niemitz, Nibbrig & Zacher, 2008). The possibility of continuous growth reduces the occurrence of senescence and promotes longevity. The crucial role of continuous growth in promoting longevity is best manifested in extremely long-lived species such as the giant tortoise (Aldabrachelys gigantea) (Quesada et al., 2019), orange roughy (Hoplostethus islandicus) (Andrews, Tracey & Dunn, 2009), and Greenland shark (Somniosus microcephalus) (Nielsen et al., 2016). In other long-lived forms, such as termite queens, substantial increases in body size are observed in imagos, a rather unique phenomenon among eusocial species of insects (Heinze & Schrempf, 2008; Keller, 1998).

Surprisingly, the most successful clades of terrestrial animals, in terms of total biomass, population size, or numbers of species, are mammals, birds, and insects, which show clear symptoms of senescence (May, 1988; Monaco, Silvestre & Silveira, 2012; Mora et al., 2011). Hence, senescence coincides with the ability of animal clades to adapt to a wide variety of terrestrial habitats.

Main physical factors hindering the colonisation of land by animals

As a frame for the paper, we explore the evolutionary constraints imposed by colonization, which may help explain the onset of senescence. We try to not develop the side threads and avoid the intricacies of the mortality question. We begin with a brief overview of the animal adaptations needed to make the transition from the aquatic environment to the terrestrial environment.

Respiratory system

When invading land, animals had to overcome various constraints (Bray, 1985) such as supplying tissues with oxygen and removing carbon dioxide. Gills were useless on land as they dried out easily (Hsia et al., 2013). This problem was solved in insects through the direct supply of air to tissues by the system of capillary structures (trachea).

As insects increased in size, this oxygen supply system became less efficient (Nation, 2005), making a significant increase in the body size of insects nearly impossible (Hanken & Wake, 1993; Polilov, 2016). Thus, the miniaturisation of body size (Table 1), frequently observed in insects, significantly improved oxygen supply (Molnar & Gair, 2015; Nation, 2005). Accordingly, the maximum body size of insects (Nation, 2005) is strongly dependent on the oxygen concentration in the air (Burggren & Infantino, 1994; Whitford, 1973). Therefore, insects could grow to large sizes when the oxygen concentration in the air reached 38% in the Carboniferous and Permian ages (Harrison, Kaiser & Vanden Brooks, 2010; Molnar & Gair, 2015).

The first amphibians used the circulatory system already existing in fish (Snyder & Sheafor, 1999), modified by the formation of structures for gas exchange (lungs) (Thomson & Bossy, 1970), with haemoglobin as the oxygen carrier. Amphibians, like insects, are highly dependent on the oxygen concentration in the air, and their maximum sizes were also found in the Carboniferous and Permian hyperoxia periods (Harrison, Kaiser & Vanden Brooks, 2010). Some animals exchange gases not only through internal systems but also directly through the skin (Brainerd, 1999), and some species of amphibians have no lungs (Hutchison, 2008). Since increasing activity on land required greater efficiency of the respiratory system, reptiles underwent a substantial reconstruction (Garwood & Edgecombe, 2011; Little, 2009). Similarly, the air sac system and breathing movements evolved in birds (Hanken & Wake, 1993; Maina, 2015; Polilov, 2016).

Problem of desiccation and movement

The impact of other adverse physical factors, such as UV irradiation and desiccation (Hessen, 2008), was much easier to resolve. For instance, the presence of gas impermeable cuticles (insects) (Moussian, 2010) or dry scaly skin (reptiles) (Rutland, Cigler & Kubale Dvojmoč, 2019) prevented desiccation. However, the problem of movement was the most difficult to solve (Padian, 2016). Insects evolved from Pancrustaceans that had hard, crusty exoskeletons and locomotory limbs (Tamone & Harrison, 2015), which were adapted for both swimming and walking, and therefore, could be easily implemented on land (Rolfe, 1985). Terrestrial insects also developed flight capabilities, making them more dispersible than their arthropod or mollusc competitors (Yanoviak, Kaspari & Dudley, 2009). The problem of movement in tetrapods was solved later by the evolution of fins with a fleshy base in lobe-finned fishes (Sarcopterygii). Fins that were previously used for swimming were thus transformed into limbs for walking on land (Little, 2009; Long & Gordon, 2004; Oster et al., 1988). The flight ability of terrestrial vertebrates (Tables 2 and 3) appeared much later than that of insects (Garwood & Edgecombe, 2011; Hunter, 2007).

Temperature in the environment

Adapting to temperature fluctuations in the terrestrial environment is another major challenge. While the temperature of marine waters is relatively stable, temperature varies dramatically in terrestrial environments according to geographic location, depth, ocean currents, and seasons (Archer et al., 2004). Moreover, variation in temperature from below the freezing point to the values at which proteins become denatured (Bischof & He, 2006) is common in terrestrial habitats. Therefore, animals living under extreme temperatures had to develop appropriate protective molecular and behavioural solutions to survive (Angelini & Ghiara, 1984). Accordingly, the activity of ectotherms depends on external temperatures at moderate temperatures, while they are mostly inactive during cold seasons (Costanzo & Lee, 2013). In contrast, endothermy, which is seen in extant birds and mammals, assures constant temperatures under a broad range of air temperature values, enabling continuous activity (Hill, Wyse & Anderson, 2012).

Thus, the transition to land and the subsequent evolution of insects, birds, and mammals was costly. We propose that these costs led to constraints, and the key constraint was the cessation of growth at maturity. This cessation of growth may have, in turn, led to senescence.

Senescent animals dominate in terrestrial environments

Senescence may have appeared in animals over the course of evolutionary changes associated with key adaptations to life on land. Therefore, the relationship between successful land colonisation and senescence merits further analysis. Tables 1–3 present the factors that may have contributed to the loss of the ability for indeterminate growth and the reduction/loss of the potential for regeneration in adults, and thus, the appearance of senescence. We must examine the physical aspects of the environment (Hackett et al., 2008) to explain such changes in the life history of these animals.

Regeneration abilities and continuous growth—differences between taxonomic groups

Some groups of aquatic animals, such as certain species of crustaceans, molluscs (Gruber et al., 2015), fish (Nielsen et al., 2016), amphibians (De Magalhaes & Costa, 2009), and reptiles (Cayuela et al., 2019; Sparkman, Arnold & Bronikowski, 2007), do not show clear signs of senescence in contrast to terrestrial taxa, such as insects (Clapham & Karr, 2012; Davies, 1974), terrestrial mammals (Gaillard & Lemaitre, 2017), and birds (Møller, 2007), despite their phylogenetic relatedness. Although the body plans, physiology, and developmental strategies of the three terrestrial groups differ substantially, they have one important feature in common: their sexually mature forms are unable to maintain continuous body growth (Miquel, Economos & Bensch, 1981; Monaco, Silvestre & Silveira, 2012; Moore, 2006).

The loss of continuous growth upon reaching sexual maturity leads to the elimination of characteristic ancestral traits, such as high regeneration ability and the inability to replace worn out somatic cells in imagos of insects, or the limitation of this ability to certain body parts in birds and mammals (Bilinski, Bylak & Zadrag-Tecza, 2016). Hence, only hairs and nails of humans (Lipner & Scher, 2017; Wilson & Tobin, 2010), incisor teeth of rodents (Klevezal, 2010), or eagles’ beaks grow continuously (Galis, Wagner & Jockusch, 2003).

The ability to continue growing after reaching sexual maturity was almost irretrievably lost due to evolution of highly adaptive traits under environmental selection pressures. The convergence of senescence and the ability to colonise terrestrial habitats is consistent with the general theory of trade-offs (Stearns, 1992). The loss of growth is often accompanied by a loss of regenerative capacity.

Insects, as imagos, are incapable of continuous growth and do not exhibit regeneration. Adaptation of insects to the terrestrial environment could have favoured the emergence of senescence in this clade. Insects seem to be the best (and rare) examples of the “disposable soma” theory (Kirkwood, 1977). The loss of growth and regenerative capacity in adult insects appears to be due to metamorphosis. Senescence rate varies significantly across birds, mammals, and insects (Jones et al., 2014; Keller & Genoud, 1999). However, senescent clades such as insects likely evolved from a non-senescent group of Pancrustaceans (Tamone & Harrison, 2015), while birds and mammals evolved from sauropsids and synapsids that probably exhibited negligible senescence (Frýdlová et al., 2020). Since sauropsids and synapsids can only be studied on the basis of fossil remains, it is difficult to determine the degree of senescence in these animals, hence the lack of certainty that they did not senescent, but also there is no data confirming with certainty that senescence was visible in them. Woodward, Horner & Farlow (2011) examined the femoral bone microstructure of captive American alligators (Alligator mississippiensis) for the presence of an external fundamental system, a form of bone microstructure present in the outermost cortex of long bones in animals that have attained skeletal maturity. The results of this study have important implications for both extinct and extant members of Sauropsida (ancestors of birds). Since captive alligators are not senescent, this external fundamental system should not be associated with senescence when interpreting the history of extinct animals such as dinosaurs (Woodward, Horner & Farlow, 2011).

Maximum life expectancy—differences between taxonomic groups

Our analysis of the maximum life expectancy of species belonging to the groups of animals revealed distinct groups (Figs. 1, 2 and 3). The first group was insects, with a 99% share of short-lived species. However, the maximal lifespan has been significantly extended in some insects, up to several dozen years, but this occurs via extension of only the larval phase of life. The second group consisted of amphibians and terrestrial reptiles, with an average maximum life expectancy of about 10 years, though some animals, such as olm (Proteus anguinus), tuatara (Sphenodon punctatus), and Galapagos tortoise (Chelonoidis niger) show extremely high life expectancy. The third group consisted of land mammals and some birds. The average life expectancy of animals in these groups is approximately 15 years. Only humans (Homo sapiens) and few other species can be classified as long-lived (Fig. 2). The next group comprised of about 20% long-lived species that live for over 35 years (Fig. 2). This group included fish with short-lived species, but also a large number of very long-lived species, such as Greenland shark (Somniosus microcephalus) and deep-water orange roughy. The term Fish is used to refer to the five modern Classes, i.e., Myxynii (hagfishes), Petromyzontida (lampreys), Chondrichthyes (cartilaginous fishes, e.g., chimaeras, sharks, and rays), Actinopterygii (ray-finned fishes), Sarcopterygii (includes coelacanths, and lungfishes). High average maximum life expectancy was also observed in soaring birds. However, aquatic mammals were completely different from the other groups, with over 40% of them being long-lived (Figs. 1, 2 and 3).

Senescence as the trade-off between successful land colonization and longevity