“Fitness is a relative measure of evolutionary advantage which is based upon the survival and reproductive success of individuals with different phenotypes. Natural selection is a demographic process.” These sentences come from the introduction of a classic textbook on plant ecology, published in 1982 by Jonathan Silvertown (Silvertown 1982). One of us (TT) read that book 30 years ago and was shocked by the deep implication—that fitness can be calculated using demographic parameters such as survival and fecundity, which are normally used to examine population dynamics in ecology (Fig. 1a). These concepts are at the heart even of Darwin’s thinking on natural selection, which was based on the idea of competition in an overpopulated world leading to differential survival (Darwin 1859). Natural selection optimizes survival and fertility schedules through differential fitness, and these optimal schedules drive changes in population dynamics (Fig. 1b). Therefore, there must exist an interaction between ecology and evolution, and we now know that this interaction is fast enough to proceed at “ecological” timescales. This realization has created a revolution in evolutionary biology, where biologists are now discovering in what fundamental ways eliminating the theoretical separation based on timescale between ecology and evolution changes expectations in both disciplines (Shefferson and Salguero-Gómez 2015).

Fig. 1
figure 1

The interaction between ecology and evolution. a There are two types of demographic parameters, i.e., mortality schedule and fertility schedule. They affect not only population dynamics but also fitness through ecological processes. b Fitness can be calculated using demographic parameters. This concept is at the heart even of Darwin’s thinking on natural selection. Natural selection optimizes survival and fertility schedules through differential fitness, and these optimal schedules drive changes in population dynamics. Therefore, there must exist an interaction between ecology and evolution

Full size image

This important interaction is nowadays expressed by the term “eco-evolutionary dynamics,” defined as “interactions between ecology and evolution that play out on contemporary time scales” (Hendry 2016). According to Hendry, the first use of this term was very recent, by Olóriz et al. (1991). A special issue of the journal Functional Ecology in 2007 popularized the term, although the field itself had gained attention and acceptance through work by David Reznick, Takehito Yoshida, Stephen Ellner, John Endler, and others. Many papers on this topic have been published since then, and we have continued to see an explosion particularly in literature related to life history evolution, population dynamics, and community and ecosystem dynamics (Fussmann et al. 2007; Haloin and Strauss 2008; Genung et al. 2011; Strauss 2014; DeLong and Luhring 2018).

“Evolutionary demography” is a term that describes a related concept, though one that has evolved on its own and has an older tradition. Evolutionary demography uses age- and stage-specific demographic parameters such as survival and fecundity to explore the evolution of life histories. However, unlike life history theory itself, evolutionary demography is rooted specifically in demographic patterns, meaning those related particularly to mortality, though also fecundity. It also includes the traditionally separate field of biodemography, which has primarily focused on humans and senescence (Vaupel 2010), and even incorporates some basic genetic models (Rees and Ellner 2016). The first use of the term was in Wilbur (1975) (H. Caswell, personal communication), where the term was found as the name of a university seminar managed by Donald W. Tinkle at the University of Michigan Museum of Zoology (H. Wilbur, personal communication). Caswell (2001) devoted a chapter in his classic textbook to the subject and wrote, “The connection between demography and evolution follows from the recognition that the life cycle is part of the phenotype.” Several researchers who focus on mortality schedules have studied the senescence of plant and animal populations, including human populations (Rose 1997; Baudisch et al. 2013; Jones et al. 2014), and some who focus on fertility schedule have studied semelparity and iteroparity in plants (Wikberg 1995; Metcalf et al. 2003; Vaupel et al. 2013). Many of us have now integrated evolutionary demography with eco-evolutionary dynamics to re-assess and rebuild what is understood about life history evolution (Cameron et al. 2013; Shefferson et al. 2014; Metcalf et al. 2015). We, the scientists working in evolutionary demography, have walked a long and winding road from demography to evolution and back for these 40 years.

Although the basic motivations in these two emerging research fields are similar, the approach in evolutionary demography is more closely linked with the above statement written by Silvertown than eco-evolutionary dynamics because studies in evolutionary demography have focused on the whole life history. Evolutionary demography provides a more fundamental assessment not only of the life history of a species, but can also be extended to explain population and community dynamics (Griffith et al. 2016), and even the evolutionary origins of species (Shefferson and Salguero-Gómez 2015). Metcalf and Pavard (2007) reasoned that, “Evolution is driven by the propagation of genes, traits…. This propagation depends on the survival and fertility of individuals at each age or stage during their life history. Demography is therefore central to understanding evolution.” She suggested the significance of demographic data throughout the whole life history and the recent development of demographic research. Unquestionably, data throughout the life history of a species is indispensable to study evolutionary demography. To this end, two large-scale databases of plant and animal life history are now available online, the COMPADRE Plant Matrix Database and the COMADRE Animal Matrix Database, which include 4246 population projection matrices from 695 plant species (version 4.0.1), and 1625 projection matrices from 345 animal species (version 1.0.0), respectively (Salguero-Gómez et al. 2015, 2016). The British Ecological Society has even published a joint special feature dealing with this topic across six of their journals in 2015 and 2016 (Griffith et al. 2016). We are now in a revolutionary era in the demographic research of plant and animal populations (including human populations). It is now the right time to address the following two questions using such big data:

  1. 1.

    How do ecology and evolution interact, if we reconsider what is understood about demography from the point of view of potentially rapid evolution?

  2. 2.

    What can we learn in general through the accumulation of demographic data from across the Tree of Life?

Many skills and approaches are needed to answer these inherently interdisciplinary questions, because evolutionary demography covers a wide array of subjects, bodies of theory, and analytical toolkits. This special issue also covers a wide array of subjects, in line with the general topic of evolutionary demography:

  1. 1)

    Demographic analysis of populations (including human populations) from the point of view of evolutionary ecology.

  2. 2)

    Meta-analysis using big databases of populations.

  3. 3)

    Eco-evolutionary studies at the population and/or community level.

  4. 4)

    Theoretical studies and the development of mathematical models of life history evolution.

The 14 collected papers of this special issue attempt to answer a variety of questions using original ideas, new tools, and big data. Surprisingly, 11 of the submitted papers use mathematical models. The first three papers in this special issue are of theoretical studies (DeLong and Luhring 2018; Shyu and Caswell 2018; Takada et al. 2018). DeLong and Luhring (2018) develops a new Gillespie eco-evolutionary model to clarify the role of predation in evolution of the growth rate of individuals (the resultant change in age at maturity) using data from protists and numerical simulations. Shyu and Caswell (2018) present a new framework for two-sex demographic models using the mating-rule approach of Pollak and discuss the outcome of the evolution of mating strategies (polygamy or monogamy). Takada et al. (2018) propose a randomly generated population matrix model to understand why the elasticity vectors in plant populations are located in specific regions of the ternary plot. They propose a formula and a hypothesis for the elasticity vector distribution.

The fourth paper, by Shefferson et al. (2018), examines 16,770 papers on conservation management and indicates a shortage of research dealing with the evolutionary consequences of management strategies, a surprising finding given the interest that conservation biologists have had in basic demography and life history models over the years. The authors advocate a new research agenda to identify and counter this issue, and particularly call attention to the need for the integration of eco-evolutionary dynamic models with existing life history models to predict the scope of evolution and its impact on conservation potential. The remaining 10 papers are studies based on large-scale databases of populations or long-term census data of a population. One of the two papers on plants, by Horvitz et al. (2018), investigates the demography of an invasive tree (Psidium cattleianum) in the rainforest of Hawai’i and focuses on the transient dynamics of the invasive species population. The authors emphasize the importance of the variation in transient dynamics to manage invasive species. Another paper, by Kellett and Shefferson (2018), examines how variation in reproductive cost selects for reproductive delay in a neotropical milkweed (Asclepias curassavica). Simulations using a demographic model show that temporal variation in reproductive costs and payoffs is an important selective force in the evolution of delayed flowering.

The seventh to tenth papers deal with invertebrates and mammals, including hominids (Anderson 2018; Gimenez and Gaillard 2018; Hartemink and Caswell 2018; Nakahashi et al. 2018). Hartemink and Caswell (2018) construct an age–stage classified projection matrix model to detect variance in longevity among individuals from 10 studies of invertebrates. They decompose the variation into the effects of heterogeneity among stage groups and stochasticity, and show that the differences in longevity among heterogeneity groups are on the order of 30% of the overall life expectancy. Gimenez and Gaillard (2018) develop a statistical approach to estimate individual fitness under imperfect detection of life history. They examine the validity of their approach using intensively collected data of a roe deer (Capreolus capreolus) population. Anderson (2018) proposes a mathematical model to integrate three related theories: senescence theory, metabolic theory, and vitality theory. Using this model, he examines the relationship between mammal body mass and survivorship with data for 494 nonvolant mammals. Nakahashi et al. (2018) estimate the interbirth intervals of apes and modern humans based on both intensive literature research and fossil samples, and explore the possibility that hominids had shorter interbirth intervals than the extant apes to enhance fertility. They suggest that the interbirth intervals of australopithecines were significantly shorter than those of extant great apes.

The last four papers were submitted by human demographers (Colleran and Snopkowski 2018; Morita 2018; Nicol-Harper et al. 2018; Tai and Noymer 2018), and are studies based on large-scale human demographic data (including one review paper). Morita (2018) collects 122 papers on the effects of socioeconomic status on fertility decline in human populations and reviews them from the viewpoint of connecting the following two topics: “the association between socioeconomic status and reproductive success” and “an integrated understanding of evolutionarily adaptive behaviors”. Colleran and Snopkowski (2018) focus on differences in fertility decline among countries. They compare similarities and differences in wealth and education across 45 countries using Demographic and Health Survey data, and suggest that associations between wealth and fertility differ substantially across contexts, whereas the association with education is consistently negative. Tai and Noymer (2018) examine five models for determining the best way to estimate Gompertz mortality parameters using 7704 life tables of the Human Mortality Database. Of the five models, they recommend log-transformation as the best way to fit death rates on age with Gompertz mortality functions. Nicol-Harper et al. (2018) emphasize, in the introduction of their paper, the importance of the transient dynamics of populations to efficiently model populations and to best inform policy. They propose a new population metric, the non-normality of a matrix, to evaluate transient population dynamics, and showcase about 1000 human population projection matrices for 42 European countries.

After reading through these articles, we are convinced that they provide a variety of insights and improve understanding for several questions in evolutionary demography. Additionally, they present several useful analytical toolkits such as a Gillespie eco-evolutionary model (DeLong and Luhring 2018), a two-sex demographic model (Shyu and Caswell 2018), a randomly generated population matrix model (Takada et al. 2018), a statistical approach to estimate individual fitness (Gimenez and Gaillard 2018), and non-normality metrics of a matrix (Nicol-Harper et al. 2018). We hope this special issue will provide readers a useful perspective on evolutionary demography.