Discussion
Predicting range expansions with ecology and evolution occurring on the same timescale is a challenging task. Building on previous ecological range expansion studies (2, 3), we included short-term evolution from standing genetic variation in a simple model parameterised for our laboratory system and confronted predicted evolutionary outcomes with results from experimental range expansions. Both model and experiment show rapid divergence between range core and front treatments, with selection for higher dispersal at the front. The repeated fixation of particular COI genotypes in the experimental lines corresponded to strains identified as most likely winners in the model. This match between predicted and observed outcomes suggest a certain predictability of range expansions, even when evolutionary change occurs. Over longer time scales, experimental range core and front populations continued to diverge, indicating de novo evolution and resulting in the emergence of dispersal syndromes.
Dispersal and growth rate are main targets of selectionIn the context of reaction-diffusion models, dispersal (diffusion) and population growth at low densities are the two key traits for understanding and predicting range expansion dynamics (14, 15). Consistent with this view and previous studies (13, 33), dispersal and population growth were here identified as main targets of selection. Higher dispersal was immediately selected from standing genetic variation at the range front and weakly selected against in the range core in the model as well as in the experiment, where range front populations showed increased dispersal already after the first few cycles. Such strong and fast selection on dispersal in the vanguard front populations has been found in similar experiments (5, 25–29), but also in natural populations (4, 22). Dispersal evolution might therefore accelerate the speed of range expansion already over very short time scales (21).
Contrary to more standard views of range expansion with r- and K-selection (17, 21), growth rate was under positive short-term selection in both range core and front treatments. This can be explained by the fact that populations in all treatments experienced regular bottlenecks, thus imposing general selection for increased growth rate, a trait for which there was ample variation among founder strains (SI Appendix, Fig. S4). Importantly, however, our model shows that dispersal and growth rate can be simultaneously selected in the range front treatment (Fig. 3). Whether one or the other trait has more weight depends on the stochasticity introduced by the dispersal bottlenecks, implemented in the model via the quasi-extinction threshold. With small bottlenecks, even weak dispersers make it into the new patch and can subsequently regrow to high density. Indeed, additional model scenarios show that, when we decrease the quasi-extinction threshold, selection for growth rate overrides selection for dispersal and the strain with the highest growth rate becomes fixed in all treatments (SI Appendix, Figs. S8.1-S8.3). However, the model scenario that fits the observed data indicates a large enough extinction threshold in our experiment, putting equal selective weight on dispersal and growth rate (Fig. 2) and allowing selection to pick the best possible disperser strain that still has a high growth rate.
Predictability of outcomesGenetic analysis indicates that the experimental lines became fixed for single COI genotypes. Despite limited resolution (several strains have the same COI), there is a good correspondence with model predictions: Range core and control treatments were fixed for the b05 COI genotype, and the two b05 strains in the founder population were the most likely winners in the model, due to their particularly high growth rate (Table S1). In the range front treatment, there is more uncertainty (12 strains carry the b07 genotype fixed in this treatment), but among the possible candidates only the predicted most likely winning strain (goe_14) has both high dispersal and high growth rate (Fig. 3). Additional sequencing would be required to determine whether these lines are fixed for the same or different (combinations of) strains.
Although our model seems to correctly identify the most likely winner strains, it nonetheless predicts the frequent fixation of strains with alternative COI genotypes (Fig. 3). Indeed, according to the model, our exclusive finding of b05 strains in all 9 range core and control lines is highly unlikely (0.289 < 0.0001). Similarly, even in the front treatment, the expected probability of exclusive fixation of b07 strain(s) is well below 5% (0.486 = 0.01). In this sense, our experiment was more deterministic than the model. Possibly, when we determined dispersal and growth of the individual strains, a large measurement error was added to biologically relevant variation (Fig. S4). This additional noise then cascades through the model, from the strain posterior distributions (making them wider) to the phenotypic composition of the founder population (making strains more similar) to model outcomes (making them more variable). Alternatively, our model may be missing additional factors, such as direct strain-strain interactions or density-dependent dispersal, which potentially amplify among-strain variation in performance. In addition, our experiment allowed relatively long periods (6 days) of population build-up in-between dispersal events, thus resembling a pushed-wave expansion scenario (36). Under such conditions, advancing range fronts are fuelled by dispersal from large populations. This prevents drift from becoming too important and can therefore explain the highly repeatable outcomes observed here for independent replicates of the same treatment.
Regardless, one main conclusion from this model-experiment confrontation is that evolution from standing genetic variation can be fairly predictable, at least in the short term. As already shown for ecological models (3), realistic predictions can indeed be made about range expansion dynamics, at least in highly controlled laboratory settings. Here we infer trait change from knowledge of standing genetic variation in only a few parameters, suggesting that such models can be readily extended to include evolution.
Long term evolution of dispersal syndromes and emergence of trade-offsExperimental evolution studies show that adaptation to novel conditions may reduce performance in other environments (34). The emergence of such trade-offs depends on underlying biochemical and life-history constraints (35), but also on historical contingency, determining the composition and genetic architecture of the ancestral population, and thus the available trait space for selection. Generally, natural populations present high levels of standing genetic variation, on which selection can act and produce rapid adaptation (36–40). Our case mimicked this situation and generated short-term response to selection, but no clear signs of trade-offs. In the long run, however, range front and core populations continued to diverge in multiple traits (Fig. 4D), and the increase in dispersal in the front treatment was associated with a decrease in growth rate (Fig. 4A). Such coupled responses in dispersal and life-history traits are referred to as dispersal syndrome (8, 9). Typically, they involve the emergence of a competition-colonisation trade-off, where dispersal evolution coincides with selection for opportunistic growth strategies (r-selection). Theoretical and empirical studies have demonstrated the importance of dispersal syndromes in generating eco-evolutionary feedbacks and accelerating the pace of range expansions and biological invasions (4, 17, 20, 21). Dispersal - growth trade-offs were previously reported for this (41) and another ciliate species (25). In these systems, growth rate is a good indicator of competitive ability, and the trade-off with dispersal likely reflects a true life-history constraint, mediated through energy costs of foraging activity (25). The evolved differences between core and front lines are stable, even after switching core and front treatments for multiple cycles (SI Appendix, Fig. S6.1). Moreover, mixes of core and front lines readily respond to dispersal selection (SI Appendix, Fig. S6.2), making new evolutionary experiments possible, where phenotypic measurements change can easily be combined with the tracking of COI genotype frequencies.
Advantages and limitations of an asexual reproduction scenarioIn this study, we consider asexual reproduction in both model and experiment. Hence advantageous allele combinations are not broken apart or reshuffled by sex and recombination (42, 43), such that strains with favourable trait combinations rapidly increase in frequency in our range and core treatments. Similar results were reported for experimental range expansions of the plant Arabidopsis thaliana , where the fastest-dispersing clonal genotype became predominant in multiple replicate lines, all starting from the same initial mix of clones (5). Thus, asexual reproduction narrows down the variability in the range expansion outcomes and, as we show here, makes predictions possible with relatively simple models. Clearly, recombination will make predictions more difficult, and replicated range expansion experiments with sexually reproducing organisms already showed higher variability and uncertainty in final outcomes (26, 28, 29). For example, recombination may slow down range expansions in the short term, but speed up longer-term responses by creating novel trait associations not previously available. In our system, sex may have immediate and strong fitness consequences due to the nuclear dimorphism typical of all ciliates. Aside from creating novel genetic variants (in the germline micronucleus), sexual reproduction also involves the recreation of a new somatic macronucleus and thereby the loss of any (somatic) adaptation acquired during asexual life (44).
ConclusionsPredicting evolution is arduous because of the intrinsic tension between determinism and contingency (45), and it demands an adequate theoretical representation of the eco-evolutionary processes in the biological system in question and reliable information on the genetic variation in the relevant traits (31), as we describe in this work. At least in simple settings as ours, accurate predictions of the evolutionary outcomes of range expansions require surprisingly few parameters, and independent biological realisations can be highly repeatable. Future studies will need to consider, for example, more realistic landscape scenarios and interactions with other species occurring during range expansion. This would imply a more systems-biology approach, with simulations calibrated on the empirical knowledge of the specific ecological scenario and biological players (46, 47). More generally, increasing our capacities to make reliable quantitative predictions of invasive eco-evolutionary processes is critical to a variety of issues, from conservation and biocontrol strategies to antibiotic development and disease management.