Introduction
Complex life cycles, whereby organisms pass through distinct life-history stages before reaching the adult stage, are ubiquitous across the tree of life. Life-history stages can be quite different from each other in terms of morphology, trophic mode and even habitat—but they are also linked, both ecologically and evolutionarily. Ecological studies of complex life histories first focused on numerical links across the life cycle—the number of individuals entering and exiting each phase determines the dynamics of the adult population . More recently, the role of phenotypic links between life-history stages, termed “carry-over” or “latent” effects have also been recognised as ecologically important . Links among life-history stages have evolutionary significance as well—the popular “adaptive decoupling hypothesis”, argues that metamorphosis reduces the strength of genetic links across the life history, such that different stages can evolve independently to their particular optima . Accordingly, an increasing number of studies explore the phenotypic and genetic links among life-history stages, both at the macro- and microevolutionary scales . As the number of studies grow, it becomes possible to identify the aspects of complex life histories that have been relatively well studied, and we can also locate any important knowledge gaps that remain.
It is unclear whether all life-history stages have been studied equally or if biases with regards to particular stages exist. For example, a cursory reading of literature implies that links between the larval and juvenile stages are of particular interest in studies of marine invertebrates , insects , frogs But other stages can regulate populations—for example, the egg stage may determine adult densities . Identifying emphases in the literature, and the extent to which some stages/links are less examined, would allow future studies to address these gaps, and provide a more complete understanding of complex life histories, such that no one stage remains a ‘black box’.
Different experimental designs for studying life histories provide access to different inferences—how we study the life cycle determines whether we can make inferences about links between stages. Broadly, there are three experimental design approaches to studying life histories: (1) single stage, (2) cohort longitudinal—following one cohort through multiple stages, and (3) individual longitudinal—following specific individuals through multiple stages. Single stage studies are useful for characterising a particular stage of the life history, but they can tell us little about phenotypic links, because they preclude the estimation of covariances between stages—using different cohorts for each stage means that within-species covariances (be they genetic or phenotypic) cannot be estimated. Single stage studies are also useful for estimating covariances among species but increasingly, the focus is on within-species links of the life history. To make inferences about these links, estimating the (co)variances in traits following single cohorts (cohort longitudinal), or individuals (individual longitudinal) through the life cycle is essential. Even then, the two longitudinal approaches have different strengths and weaknesses. Cohort longitudinal approaches are susceptible to Simpson’s Paradox whereby trait relationships observed across cohorts may not reflect the trait relationships for individuals; the trend at the individual-level could even be opposite of that at the cohort-level (Figure 1; note that among-species comparisons are also vulnerable to Simpson’s Paradox). As such, if one wishes to make inferences about phenotypic links among life-history stages, then individual longitudinal studies are most appropriate, as they estimate trait (co)variances at the appropriate scale and avoid the potential for Simpson’s Paradox. On the other hand, individual longitudinal studies are potentially laborious—rearing individuals can be much harder than rearing cohorts. If one is interested in quantifyinggenetic links among life-history stages, and a quantitative genetics breeding design is used, the scale of replication is cohort from a single sire, and therefore cohort longitudinal approaches are appropriate (e.g. .
The nature, strength and variability of links between life-history stages vary systematically between laboratory-based study and those done in the field . Controlled laboratory conditions make experimental manipulations easier and are sometimes the only way to examine certain life-history stages . Nevertheless, field experiments provide information that cannot be gained from laboratory experiments alone . Identifying the relative abundance of field and laboratory studies should allow us to identify where and when field studies should be priority.
Here, we use a systematic map of marine invertebrates to describe the state of our knowledge and outline the field’s strengths and knowledge gaps. Systematic maps use a repeatable methodological framework to quantify what has been studied. Unlike systematic reviews and meta-analyses, systematic maps do not statistically analyse combined data from empirical studies . Instead, a systematic map collates, catalogues and describes studies, outlining the current state of knowledge for a particular topic . Marine invertebrates are an ideal model system because of their numerous phyla, diverse life-history modes and long history of study from the perspective of complex life cycles . We collected methodological data for studies of life histories across the following six stages: (1) F0 adult; (2) embryo; (3) larva; (4) metamorphosis; (5) juvenile; and (6) F1 adult (Figure 2a) and recorded the experimental design used for each study.