Introduction

Anthozoan cnidarians such as anemones and corals are sessile or semi-sessile marine invertebrates. Their complex life cycle includes a motile larval stage which serves dispersal, an important feature that ensures the resilience of the species by enhancing genetic diversity and facilitating inhabitation of novel ecological niches.1,2 During the larval phase, the animals display state-specific behaviors, with planktonic larvae exhibiting dispersal behavior early in development. Animals can disperse and explore the environment passively (e.g., by ocean currents) or actively (through swimming). The second type of behavior occurs when the larvae become competent, a physiological state in which behavior centers around the choice of a permanent settlement site. Many aspects of larval dispersal and settlement remain enigmatic. It is not well understood which sensory modalities cnidarian larvae employ to guide swimming or settlement, how such information is processed, how it modifies behavior, and which cellular subsystems are involved.

Anthozoan larvae are diverse in developmental time and anatomy,3 and consequently larval behavior too is multifaceted and species-dependent, suggesting that both these traits contribute to adaptations to a specialized ecological niche with its physical peculiarities. Swimming in the larvae can encompass a variety of modes and planes, and some cnidarian larvae even exhibit other complex behaviors such as active predation.4,5,6 Within the same species, larval behavior is highly plastic throughout development, with animals transitioning from near motionless to active swimming and later to crawling and gliding or probing the substrate immediately prior to settlement.7,8,9

For both dispersal and settlement, it can be assumed that sensory information is relevant in guiding the larvae. Indeed, both behaviors have been linked to sensory perception in cnidarian planula larvae.6,10,11,12,13 Cnidarian larvae have not been found to perform active taxis but rather phobic responses which nevertheless result in specific distribution of the adult polyps with respect to sensory stimuli.13 During these phobic behaviors, the animals respond to sensory stimuli by modifying ciliary beating and body shape, which result in slow positional changes.6,8,11 Using phobic responses rather than active tactile swimming is a sensible and energy-efficient strategy, considering that, with the size of the larvae and the vastness of the ocean, active swimming toward or away from a stimulus would often exceed the physiological capacity of the animals.14

How cnidarian larvae with their small size and relatively simple body plan regulate these diverse behaviors is unclear. The motile planula larva is polarized and swims forward with its aboral end propelled by cilia, which cover the whole animal, forming metachronal waves that improve the efficiency of locomotion.15,16 In addition, some anthozoan cnidarians possess a sensory organ, the apical organ, with long protruding cilia at the aboral end.17,18,19 Furthermore, many other specialized cell types, such as secretory, neurosecretory, putative sensory, and support cells are enriched in the apical region of the larvae, suggesting that this area combines several sensory effector centers that may be involved in guiding specific behaviors during the larval life.19,20,21,22 This idea is also supported by the underlying prominent aboral nerve plexus, of which some parts might be larval-specific.17,20,23 In addition to the above-mentioned cell types, muscles become more prevalent and increasingly organized over the course of development,24 and late-stage planula larvae can contract using their muscular-hydraulic system.6,25

Knowledge about cnidarian behavior is critical to our understanding of the basic biology of this vast group of marine invertebrates that form the foundation of many rich ecosystems. Cnidarian planula larvae also occupy a particularly interesting evolutionary position, being one of the most basal extant animal lineages with a nervous system that might influence their behavioral repertoire already during early life stages. While even single-celled eukaryotes can perform sophisticated cilia-mediated behaviors26,27 and sponge larvae use both sensory and ciliated cells to modify their swimming pattern in response to light,28 it is an exciting endeavor to study the exact function of the nervous system and other larval-specific structures, such as the apical organ, to shed light on the complex interplay between cilia and developing nervous and muscular systems.

Here we describe the basic swimming behavior of the Nematostella vectensis planula larvae over the course of development from hatching to metamorphosis into a primary polyp. To study if swimming behavior indeed changes over development and reflects life stage-specific behavioral interests that are mediated by specialized anatomical features, we analyzed these aspects concomitantly in carefully selected stages. Nematostella is a burrowing sea anemone, commonly found in shallow estuaries and tidal pools. Its life cycle includes a motile larval stage with an apical tuft that transitions into a primary polyp (metamorphosis) within a few days (Figures 1A and 1B). Our data show that swimming behavior changes drastically over the course of development. Nematostella planula larvae become faster and swim further and are less spatially confined until metamorphosis. This advance in swimming abilities occurs simultaneously with an increase in sensory capabilities, in the form of a higher response to mechanical perturbation, a longer apical tuft, and the enhanced expression of sensory receptors, which were found predominantly in the ciliated epithelial cells. Furthermore, while the length of the motile cilia and the body remain relatively stable during the active phase, the animal’s body shape changes both over development and as the animals behave, correlating with specific swimming modes. We find that the nervous system controls the shape-shifting abilities of the late larval stages but has little impact on sensory perception or the ciliary beating per se, suggesting that Nematostella larvae use two independent systems to control swimming behavior: a ciliary sensory-motor system and a neuronally controlled muscular contraction system, aimed at optimizing body shape to swimming mode.

Figure 1: Swimming behavior is changing over the course of larval development(A) Nematostella vectensis adult, (B) its life cycle, and (C) the behavioral arena (5 × 5 cm) used in the horizontal swimming experiments and an example of a swimming track obtainable in this assay.(D) Overlaid tracks produced during a 5-min video from four ages of N. vectensis larvae showing swimming behavior. 48 hpf (n = 123, N = 3), 72 hpf (n = 167, N = 3), 96 hpf (n = 158, N = 3) and 120 hpf (n = 187, N = 3), scale bar, 1 cm.(E–H) Distributions of track mean speed, total distance, maximum distance and confinement ratio. Dashed line at the mean; stars denote statistical significance of ns p > 0.05, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 by Kolmogorov-Smirnov test, x axes on log10 scale.(I) The defined thresholds for categorizing larvae as either stationary or moving, scale bar, 5 mm.(J) Ratio of moving and stationary larvae per age (48 hpf: 44% moving, 72 hpf: 80% moving, 96 hpf: 89% moving, 120 hpf: 52% moving).(K) Example trace (scale bar, 1 cm) and (L) calculations of age-specific average distances the larvae swam per minute. Lines show a smoothed conditional mean line ± SEM. See also Tables S1–S6.