Understanding the selective pressures affecting the evolution of dispersal strategies is of prime importance for a broad range of biological fields, ranging from conservation biology to research on the evolution of species, host-parasite interactions and communities of species [1-7]. Dispersal strategies of living organisms affect the dynamics, and the demographic and genetic structure of their populations [8,9], and impacts often crucially on the survival and reproductive success of individuals [2,10,11]. For example, dispersal has been theoretically demonstrated to profoundly influence the evolution of sociality [12-17] and dispersal strategies affect selection on other core life-history traits, e.g. reproductive strategies and effort, survival capacity, colonization ability, defenses against predators, parasites and diseases [7,18-23].
Co-variation among life-history traits has been the centre of much research in the past decades in the context of adaptive evolution as well as from the point of view of constraints on this evolution due to trade-offs between traits [24-29], and the existence of syndromes encapsulating dispersal strategies has been hypothesized [21,23,30,31]. For example, Baker and Stebbins  hypothesized that species living in unstable habitats in metapopulations with a high turn-over develop a set of co-adapted traits where high dispersal and colonization ability are linked to high fecundity and short life span (low survival), allowing excellent exploitation of freed-up patches (successful dispersal and colonization followed by rapid growth of the colonist population) . Such hypotheses have been supported by correlative approaches, especially by comparing species in different types of habitats . Theoretical models equally predict the evolution of syndromes in which dispersal ability influences or is influenced by the evolution of other life history traits [18,33], even if sometimes they come to a different conclusion than verbal or correlative approaches . For example, Ronce and Olivieri  predicted a positive association between life span and dispersal while Crowley and McLetchie  predicted the reverse, the direction of the trend seemingly depending on details of the functioning of metapopulations  and also, we note, on potential constraints (trade-offs) on life-history traits and their capacity to evolve [25-28]. Similar complex theoretical results have also been found for the evolution of social traits and dispersal ability .
The evolution of dispersal syndromes through natural selection requires that the traits in question are under genetic determinism, and we must therefore demonstrate this to corroborate the theoretical view on dispersal syndrome evolution. Detailed studies on genetic variation in life history trait associations may also increase our understanding of the process of dispersal syndrome evolution, by allowing an evaluation of the degree to which the directionality of these patterns follow theoretical predictions or might be limited by trade-offs, and can also suggest additional selective factors that may have shaped syndromes. Nevertheless, only quite few studies test for and explore variation in dispersal syndromes apart from species comparisons, although genetic variance for dispersal itself has been demonstrated for some species (e.g. insects, birds, reptiles, mammals, [36-47]). Some correlative studies have found differences in the life-history of dispersing versus philopatric individuals [10,11] or intra-specific differences in dispersal strategies related with the cause of dispersal and landscape structure [48,49]. Other studies have found clear associations between individual propensity to disperse and some morphological, physiological or behavioral characteristics [50-54]. Such associations may, however, be due to maternal or environmental factors (i.e. forms of phenotypic plasticity, review by [27,47,55,56] as well as to genetic variation. Overall, if we except species with a clear dispersal-dedicated apparatus [24,45,47,58], evidence for genetic variation in dispersal syndromes and close dissections of trait associations are scarce (for a few exceptions [39,41,44], see also ). This means that our empirical and experimental insight into the evolution of these syndromes remains relatively poor.
One reason for the scarcity of evidence for genetic variation in dispersal strategies and other life history traits within species is that most studies have been done either on vertebrates, where assessing the genetic determinism of traits associations is difficult, or on invertebrates where an individual following up is often impossible. For this reason, artificial microcosms and clonal organisms with short generation time seem one of the best ways to investigate such syndromes ([23,59], see also ). Microorganisms, characterized by the general ease of maintaining large population sizes in the laboratory under controlled environmental conditions, should therefore be very well-suited organisms for studies on the co-variation between dispersal strategies and other core life history strategies. Surprisingly, however, very few studies have addressed the relationships between dispersal rate and core life history traits in microorganisms (, but see ).
We here present a study on the co-variation of dispersal strategies with other life-history traits in the unicellular, ciliated protozoan Tetrahymena thermophila. This small (60 μm) eukaryote feeds on bacteria and dissolved nutrients in fresh water ponds and streams in America [61,62], and while it is widely used as a model system by molecular and cell biologists , it has been surprisingly little studied by evolutionary biologists (but see [63-65]).
Yet the life history characteristics of T. thermophila make it particularly exciting for studies on dispersal. Firstly, this organism lives in habitats likely resembling those of meta-populations with high turn-over of local patches. Studies on life history trait associations in T. thermophila would therefore allow gaining insight into the degree to which central theoretical models can predict dispersal syndrome evolution in ruderal species (e.g. the classic colonizer syndrome model [23,30,31]). Secondly, genetic variation in all core life history traits can be easily estimated in T. thermophila because separate clonal lineages can be kept in the laboratory. Reproduction remains clonal whenever nutrients are present  and even under conditions inducing sexual reproduction (such as starvation), conjugation is impossible between clone mates because they carry the same mating type (, reviewed by ). Thirdly, T. thermophila shows an intriguing co-existence of apparent short- versus long-distance dispersal strategies (by normal cells versus cells that have transformed into elongated morphs with very numerous ciliae and a caudal flagellum ). T. thermophila of both morph-types swim about and explore their environment (personal observations), but the much greater swim speed (4–5 times [69,70]) and more directional movements (personal observations, see also ) reported for the elongated morph suggest that it is specialized for committed long-distance dispersal. Finally, T. thermophila cells form aggregations and secrete substances favoring the survival of other cells [72-74]; the evolution of dispersal strategies in this organism may therefore also be affected by a balance between kin-benefits and -competition (see also ).
Through experiments in the laboratory, we assessed variation in dispersal rate (in a two-patch system) and differences in the colonization capacity of single cells, and studied growth rate, patch carrying capacity, starvation resistance and concomitant changes in cell shape for ten strains of T. thermophila. These experiments and observations allowed us to (1) test for genetic variation and co-variation in core life history traits, (2) examine whether the directionality of life history trait associations across genetic lineages followed predictions for ruderal organisms (in particular as concerns the classic colonizer syndrome and competition-colonization co-existence) and (3) generate testable hypotheses on how the semi-social life style of T. thermophila may affect the evolution of its dispersal strategies.