Locomotion of sea stars is via multiple tube feet that are part of a larger system of hydraulic ducting known as the water-vascular system.  Hydraulic movement of fluid from storage sacs or ampullae to the tube feet (podia), accompanied by relaxation of longitudinal muscles, act to extend the tube feet.  On contact with the substratum the tube feet attach momentarily to the substratum by means of a sucker. Contraction of postural muscles at the proximal regions of the tube feet moves the body in relation to the tube feet for locomotion.  Contraction of the longitudinal muscles in the tube feet pull them closer to the substratum for anchoring, or pull prey closer to the arms for feeding. There is evidence that the madreporite acts as a conduit for at least some of the fluid present in the water-vascular system.

The nervous system is sited close to the elements of the water-vascular system. It consists of a ring around the mouth and radial nerves running down each arm, with smaller nerves running to each tube foot. There are no ganglionic clusters and nothing equivalent to a brain. Other than moving in the same direction during locomotion, there is no coordination of stepping movements of the tubefeet.  

The number of tube feet in a sea star is large (estimated in an adult sunflower star Pycnopodia helianthoides to be about 15,000), so precise control may in any case be beyond the capacity of the nervous system.  It is common to see fast-moving sea stars, such as Pycnopodia, pulling themselves along with the tube feet on the leading half of the body and letting the arms on the back half of the body stream out behind, with little or no contact of tube feet with the substratum.

This fast-crawling individual P. helianthoides is moving towards
the camera with its back arms trailing (not visible). Not atypically,
the arms at each side are held up and/or curled, indicating that
they are not participating in the action 0.5X

How do the tube feet attach to the substratum? Early thoughts were that suction is created by special levator muscles raising up the centre of the terminal disc of the sucker in combination with secretion of sticky mucus, but focus is now mainly on chemical adhesion by fast-acting glues.  Microscopic examination of tube-foot epithelium of Leptasterias spp. reveals 3 types of cells: adhesive cells, large-granule secreting cells, and monociliated cells thought to be sensory in function. A sea star moving across a clean surface, like glass, actually leaves tube foot-prints, marks left by secretions.  A sea star attaching and detaching to surfaces naturally filmed with bacteria and diatoms has clean tube feet, indicating that adherent material does not accumulate on the tube-foot surfaces.  Nor do the tube feet stick to, or pull on, the substratum as they are lifted away. Through the nature of proteinaceous secretions from the adhesive and large-granule secreting cells the entire surface of the sucker end of the tube foot becomes negatively charged, which also may be involved in attachment to the substratum.  Tube feet attach best to charged substrata, and attach poorly or not at all to uncharged surfaces such as Parafilm, dental wax, epoxy resin, and the like.  Detachment may be effected by secretion of other chemicals, releasing the tube feet cleanly from substrata to which they have attached, but leaving footprint residues behind.  From the results of their study, the authors do not discount the involvement of suction, especially on solid surfaces, but suggest that this is a secondary adjunct to adhesion established by a protein glue.  Thomas & Hermans 1985 Biol Bull 169: 675.

NOTE  although the authors state that the “distal surface of the tube foot is coated by a negatively charged surface which somehow attaches to the substrata”, the way that this might work is not made clear.  Perhaps it will be the subject of a future study 

Ochre star Pisaster ochraceus anchored to a rock with its tube
feet during low-tide exposure. The attachment is so stong that
the the tube feet may be ripped away when force is applied to
the body, leaving the torn-off portions hanging to the rock 2X

Is there a functional anterior part of a sea star as it locomotes along?  Early observations on sunflower stars Pycnopodia helianthoides crawling in Bremerton Bay, Washington suggests that there is a functional anterior, defined by the author as the axis of the oldest arms as the sea star crawls along. Without a description of early development of the post-larval arms in P. helianthoides, however, it is not possible from the author’s description to figure out which of the arms these “oldest” ones might be.  The author states that “the first pair of…post-larval rays…is at right angles to the anterior-posterior rays” (p. 243), which suggests that these first, or “oldest arms”, actually extend out to one side or other, or perhaps both, of the rays that are designated as the anterior ones.  Whether this anterior end has a fixed orientation to the madreporite is not mentioned. Despite this confusion the author is clear that the same set of arms is in the lead during different locomotory bouts and, after turning from an obstacle, the sea star resumes locomoting with this functional “anterior end” in the lead.  Other experiments show that 92% of righting responses are toward this “anterior end”.  Kjerschow-Agersborg 1918 Biol Bull 35: 232.

NOTE  most authors take a contrary view, that while certain arms do take on temporary dominance in locomotion, these change as an individual changes direction or activity. In this view, localised control of movement temporarily resides in a node at the junction of radial and ring nerves on the arm that happens to be leading, and this changes as the leading arm changes

NOTE  46 out of a total of 50 trials; all trials, however, are done on the same specimen over the course of a single day (10cm diameter with 13 arms, at 10^oC).  Of interest, in this regard, is the apparent indefatigability of this specimen, whose righting speed remains essentially constant over the course of the 50 trials (see histogram above)

Sunflower star Pycnopodia helianthoides crawling at high speed
along a piece of kelp in the direction shown by the white line. The
madrepoite can't be seen in this view, so is likely not at the front 0.4X

Another question related to the one above regarding leading arms during locomotion, is whether there a preferred/leading pair of arms used during righting. This is tested in a study in Oregon using Henricia leviuscula and Leptasterias hexactis.   Most asteroids right themselves by somersaulting, or a combination of folding-over and somersaulting, and these 2 species are no exception.  Henricia only rights by somersaulting, while Leptasterias uses both somersaulting and folding-over, sometimes in combination.  The authors report a tendency in Henricia for the arm pair opposite the madreporite to lead during righting, but with a more general pattern being exhibited in Leptasterias.  As expected, size has an effect on righting times, with larger animals taking significantly longer to right.  However, the authors do not specifically investigate scaling effects of size and this might be an interesting project for further work. Neither righting method was significantly faster than the other, suggesting no advantage for survival in the field.  Polls & Gonor 1975 Biol Bull 148: 68. Photographs courtesy Dave Cowles, Walla Walla University, Washington rosario.wallawalla.edu.

NOTE  in somersaulting all 5 arms curl aborally bringing their tips in contact with the substratum.  One adjacent pair of arms then twists so that they face one another.  These 2 arms begin to move to the sides causing the remaining arms to rise up and swing over the body.  In the folding-over method is similar but without the curling of the arm-tips.  A third, rarely seen method, involves all arms rising up to form a tulip shape.  Several arms then collapse causing the animal to flop over.  Interest in the subject dates from 1862

NOTE  other west coast somersaulters are Pisaster ochraceus and Patiria miniata, while somersaulter/folding-overs include Pycnopodia helianthoides. The photo series below shows a juvenile P. helianthoides righting itself apparently by the folding-over method, sequence goes Left to Right

Many species of sea stars curl up the tips of their arms when crawling, as shown here for an arm of a sunflower star Pycnopodia helianthoides.  The behaviour can also be induced by a variety of mechanical, chemical, and photic stimuli.  Not only does the behaviour provide maximum exposure for sensory tube feet located at the arm tips, but the light-sensitive eyespot is also exposed.  Sloan 1980 J Nat Hist 14: 469.

NOTE  also known as the compound ocellus or optical cushion.  Information on light effects on the morphology and physiology of ocelli of Patiria miniata, Leptasterias pusilla, and Henricia leviuscula can be found in Eakin & Brandenburger 1979 Zoomorph 92: 191.

Arm tip of a sunflower star Pycnopodia heliathoides showing the
red eyespot nestled within a protective array of spines. Other
features to note are the long, chemotactile tube feet, clusters
of pedicellariae, and the sac-like dermal branchiae 2X

Usually we assume that sea stars either have pointed/non-suckered tube feet (categorized as PNS) useful in mud-bottom habitats as in Luidia spp. in the Order Paxillosida, or they had flat-tipped/suckered tube feet (FS) useful in hard-bottom habitats, as in Pisaster spp. in the Order Forcipulatida. A study by researchers at the University of Alabama, however, shows that this simple notion needs to be reconsidered, as there is considerably greater variation in tube-foot morphology in asteroids than was previously thought.  The authors examine 45 world species in 7 orders and 19 families, with 16 west-coast species being included.  Their observations  require the creation of 4 new categories of tube-foot morphology, including semi-pointed/non-suckered (SPNS), flat-tipped/non-suckered (FNS), semi-flat-tipped/suckered (SFS), and semi-flat-tipped/non-suckered (SFNS).  The authors also find that a consistent relationship exists between  tube-foot morphology and taxonomic order.  Thus, all Forcipulatids have flat-tipped/suckered tube feet (FS), all Velatida have semi-flat-tipped/suckered ones (SFS), all Valvatida have flat-tipped/non-suckered ones (FNS), and so on.  See examples below for a few west-coast species.  Notwithstanding this taxonomic consistency, the authors remark on the variability in tube-foot design in asteroids and suggest that what is needed now is a better idea of how each morphology suits a particular locomotory, feeding, and anchoring need.  Vickery & McClintock 2000 Amer Zool 40: 355. Photograph of Henricia pumila courtesy Dave Cowles, Walla Walla University, Washington.