title for amphipod section of A SNAIL'S ODYSSEY
  Locomotion & tidal rhythms, burrowing, & celestial navigation
  black dot
  Celestial navigation
  Studies on celestial navigation are considered here, while studies on LOCOMOTION & TIDAL RHYTHMS and BURROWING are dealt with elsewhere.
  black dot

Amphipods are known for their ability to navigate using celestial cues, notably, sun and moon.  In fact, species of talitrid amphipods in Italy and elsewhere are thought to have a continuously operating adjustment as the moon transits through the night sky, presumably to help them orientate seawards or landwards.  It may seem surprising that such a complex system for navigation should exist, given the many other potential orientating cues that are available for a beach-inhabiting amphipod, such as moisture gradients in air and sand, beach slope, wave vibration, scent of seaweeds stranded by the tide, and the like. However, west-coast semiterrestrial talitrids inhabit burrows among small-scale dunes and seaweed clumps where distant vision may be obscured, and shifting sands and eddying breezes may disrupt navigation by chemical cues.

Several studies have been done on the orientating ability in Megalorchestia corniculata in southern California.  The results generally agree with respect to solar navigation, but not with respect to lunar navigation. This section begins with some information on vision in amphipods, done not on west-coast species, but on related European ones. This information may be useful in interpreting the Research Studies below, and may generate ideas for work on vision in west-coast species.

  black dot
Research study 1

composite photo/drawing showing fine structure of the eye of a talitrid amphipod MegalorchestiaWhat do amphipods see?  Do they see in colour?  Can they resolve images, or is their vision limited to motion-detection or shadow responses?  Is their vision good enough to resolve landmarks and distinguish between them?  Amphipod eyes are compound as in other arthropods, and sessile, that is, unmovable.  Each eye consists of several hundred individual ommatidia, each of which has its own lens system, light-sensitive retinal cells, nerve leading to an optic ganglion, and each is thought to produce a single image.  Visual fields of adjacent ommatidia overlap, presumably producing good motion detection, but possibly less good resolution.  Light entering a single ommatidium is screened by surrounding pigment from entering adjacent ommatidia, thus maximising visual acuity.  Presumably, as proposed for other arthropods, the multiple images are combined into a single image in the optic ganglia of the brain.  Given the similar ommatidial morphology between amphipod eyes and other arthropod eyes, it is probably safe to assume that they have good resolution and motion detection, and probably see in colour.  Hallberg et al. 1980 Zoomorphol 94: 279.

NOTE  lit. “eye” G.  The several hundred ommatidia of an amphipod eye compares with the approximately 800 in the eye of a fruitfly Drosophila melanogaster, and we know that these flies have good vision.  In comparison, a dragonfly’s eye has 30,000 or more ommatidia and, with a visual field of almost 360o, their vision must be among the best of any arthropod.  Even so, it is anecdotally quoted that for a dragonfly’s eye to produce comparable image-resolution to a human eye, it would have to exceed one meter in diameter. The cross-section and single ommatidium shown here are for Gammarus, a species not closely related to Megalorchestia, but one likely with similar eye structure

NOTE  amphipods have so-called “light-adapted” compound eyes, where the screening pigments extend the entire length of each ommatidium.  Here, only the light entering a given ommatidium passes through to its retinal cells.  Other crustaceans, for example, mysid shrimps, have so-called “dark-adapted” eyes in which the screening pigments are concentrated at either end of each ommatidium, thus allowing light from one ommatidium to pass through to several adjacent ommatidia.  In this type of eye, light-gathering power is enhanced at the cost of visual acuity, because images from adjacent ommatidia will be superimposed, and the resulting image less sharp.  Dark-adapted eyes would be expected, then, in deep-sea and nocturnal crustaceans, where light is at a premium, but the distribution of light- and dark-adapted eyes among crustaceans is actually more haphazard than that.  Most advanced arthropods have the capability of shifting from one state to the other

  black dot
Research study 2

photograph of a head of an amphipod showing approximate location of the ocelli within the brainStudies on several European amphipods, including the terrestrial Talitrus saltator, disclose the presence of paired ocelli within the brain.  There is no visible evidence of their presence from the outside of the head, yet they appear functionally capable of sensing ambient light intensity.   Each ocellus consists of 3 light-sensitive cells.  Based on their dorsal location in the brain and their structure, the authors think that the ocelli may represent remnants of the ancestral naupliar eye.  The authors caution that future interpretation of amphipod behaviour involving light responses should take into account the presence of these ocelli.  Frelon-Raimond et al 2002 Invert Biol 121: 73.

NOTE  present-day amphipods incubate their young in brood chambers, and ancestral remnants of a free-living naupliar stage are visible only briefly during embryonic development

  black dot
Research study 3

schematic showing orientation of amphipods Megalorchestia californiana to the moon at nightschematic showing orientation of amphipods Megalorchestia californiana to the moon at night, but with a mirror redirecting the moon's reflectionResults of an early study in southern California confirms an ability for lunar orientation1 in Megalorchestia corniculata.  Animals are captured from a beach near La Jolla, California and kept in light-proof containers, then transported 40 km inland2 for testing.  In the test results shown on the upper Left, 3 groups of 25 amphipods each orientate at an angle of 120o to the moon.  Were they to be back on their home beach in La Jolla, an orientation of N15oE would take them to the water, so an angle of 120o is more or less in the same direction.  If a mirror is now used to reflect the moon’s image 180o from its true position, the amphipods re-orientate to an angle of about 135o from the moon's image (see schematic lower Left). These data show that Megalorchestia can use the moon for orientation, but is the angle always the same, for example, 120o, or do the amphipods continually adjust to the changing position of the moon as it transits the night sky? The first is the idea of "constant angle hypothesis"; the second, the "lunar navigation hypothesis".

graph showing different hypotheses for lunar orientation in talitrid amphipods Megalorchestia californianaThe author tests these 2 hypotheses (see graph on Right). Note that a “constant angle hypothesis", in this case at 120o would produce the regression shown in purple, while a continuously-adjusting behaviour as exhibited by some European talitrids would produce the regression labelled "lunar-navigation hypothesis". The angle of the actual regression line for the data, shown in blue, does not differ significantly from that predicted by the “constant angle hypothesis", in other words, as slope equal to zero.  Thus, the amphipods orientate at a constant angle with the moon, regardless of its stage or position in the sky, and fail to adjust continuously as shown for other talitrid species3.   Interestingly, if Megalorchestia is given a glimpse of either sunset or moonrise on the night of testing, then they do correctly time-shift their angle of orientation to the moon.  However, this is a single-cycle, orientation rhythm that has to be reset with the appropriate stimulus each night in order to work.  Enright 1961 Biol Bull 120: 148.

NOTE1  the author remarks that he has already documented solar orientation in Megalorchestia, but the relevant paper has not yet been incorporated into the ODYSSEY

NOTE2  use of an inland site reduces effects of city lights and eliminates effects of other cues such as moisture gradients and scents arising off the ocean. A test involves the animals being released in groups of 25, of mixed sexes, into opaque-walled arenas (25cm diameter x 4cm in height, covered in transparent glass) at night during different phases of the moon and during different stages of the moon’s transit.  The animals cluster against the wall of the container and photos are taken.  The compass direction of this clustering is their orientation

NOTE3  this type of behaviour has previously been documented for a species of Talitrus in Italy

  black dot

So, an amphipod comes out of its burrow in late afternoon or at dusk, checks the direction of sunset and/or moonrise, and heads off towards the strandline to find food. After an night's activity on the beach, it checks the position of the moon and heads unerringly back to its burrow. What is wrong with this story? Think about the answers provided then CLICK HERE for explanations.

Well, to start with, it would't work.

If celestial cues are important for seawards navigation at dusk, wouldn't the sun be a better cue than the moon, especially on the west-coast where it sets over the water?

If the moon is so important to navigation, what do the amphipods do for the portion of the month when it is not visible?

On overcast days, when neither sun nor moon is visible, must the amphipods now remain in their burrows?

Food and mates are never far removed from an amphipod's burrow, so an ability to migrate up and down the beach is of little selective value.

  black dot
Research study 4

Another study on Megalorchestia corniculata in Santa Barbara, California fails to confirm the existence of lunar navigation.  The methods used are similar to those of Research Study 1, but with some experiments being done under laboratory conditions and others under field conditions.  The author notes that much of the feeding activity in Megalorchestia is randomly directed, but adds that the amphipods do go generally towards the sea on emergence from their burrows at dusk and do return landwards when high tide or sunrise comes.  In support of these later results the author argues that an ability for lunar orientation is obviated by the fact that the moon is only visible for half the time.  This is a compelling argument. However, as the earlier study has demonstrated that a lunar-orientating ability exists, and as it is shown to be present in other talitrid species, we should probably for now acknowledge that it exists in M. corniculata, at least to some extent.  Craig 1971 Anim Behav 19: 368.

  black dot
Research study 5

A later study by the same author on Megalorchestia corniculata at Santa Cruz, California shows that when placed on the hard moist sand between tide marks during daytime, individuals invariably move landwards (see diagram upper Left). This occurs regardless of sunny or overcast sky conditions, variations in beach slope or moisture gradients, or displacement to a new beach with reversed orientation to the home beach (see diagram lower Left).

schematic showing movement patterns of amphipods Megalorchestia corniculata after being released on a beachIn another location where the beach at point of release actually slopes downwards towards the land, and where there is a cliff and a backshore body of water, the amphipods still move landwards. Thus, in each case, the amphipods move landwards, despite being on a downslope and moving towards water. These results differ from the results in Research Study 1 above in which M. corniculata,  displaced to a new beach with reversed orientation, retains fidelity to the landwards compass direction of the original beach.  However, note that the present study is done in daylight, while the earlier study was done in the dark.

The results emphasise the importance of visually obvious landforms in navigation of the species during daylight.  Note in the 3 diagrams that M. corniculata orientates towards a backshore cliff despite other varying conditions. Craig 1973 Mar Biol 23: 101.

  black dot
Research study 6

histograms showing directions of movements of amphipods Megalorchestia corniculata on wet and dry slopes at different anglesIn an another study the same author shows that Megalorchestia corniculata around Santa Barbara, California have a remarkable ability to discriminate between slopes differing by only 2o.  Moreover, the amphipods tend to move up a slope if it is wet, and down if dry.  The tests are done in the laboratory with the test specimens in a long covered box containing either wet or dry sand.  The animals are left for a time with the box adjusted to a certain slope, and then the cover is removed and number in each half counted.  Results for wet-sand trials show that more animals move up the slope as its steepness increases (see upper histogram).  Thus, on a 1o slope about 35% of individuals go downwards, while 65% go upwards (see blue bars).  On a 9o slope the comparable numbers are about 20 and 80% (see purple bars). 

In comparison, on dry slopes amphipods tend to move downwards more with increasing slope, in about the same proportions (see lower histogram). 

Regardless of position on the beach, then, these behaviours will bring the amphipods to the upper part of the intertidal zone where they normally burrow.  The author notes that beach slopes of 5o are common in the study area. He offers this hypothesis of slope effects in place of a previous hypothesis of lunar orientation (see Research Study 1) for navigation in M. corniculata.  Craig 1973 Anim Behav 21: 699.

NOTE  it is not clear from the author’s description, but the trials seem to have been run in batches of around 20 individuals each, and then the data summed

NOTE  numbers moving up or down differ significantly for all slopes save 1o, and this is true for the dry-sand trials, as well.  In 0o control tests the amphipods distribute themselves more or less evenly within the test box

  black dot
Research study 7

one of several schematics showing orientation of talitrid amphipods Megalorchestia corniculata after being released on beaches in southern California under different conditionsA study on daytime orientation in Megalorchestia corniculata in southern California also shows that local landmarks, such as high land-forms behind the beach, appear to be used in addition to celestial cues to navigate.  The study involves transporting animals to beaches with differing compass orientation to their home beaches, and assessing seaward and landward orientations.

First, landward orientation of amphipods collected at La Jolla, California is tested on their own beach, released on the lower zone of the one of several schematics showing orientation of talitrid amphipods Megalorchestia corniculata after being released on beaches in southern California under different conditionsbeach at low tide (see schematic upper Left).  The results show a compass orientation of 86o, not significantly different from the landward orientation of the beach at 105o, indicating that these individuals have no trouble finding their way to the top of the shore .

However, if Santa Barbara animals are brought to La Jolla and tested, their orientation is to a compass direction of 93o, significantly different from the 10o landward direction for their home beach (see schematic upper Right). The author suggests that cues other than sun-compass ones, possibly landmarks, are being used for navigation. 

one of several schematics showing orientation of talitrid amphipods Megalorchestia corniculata after being released on beaches in southern California under different conditions
What about seaward orientation?  Tests of animals from Del Mar beach released in the back, upper shore area on their own beach show a mean orientation of 261o, close to the actual 251o direction to the ocean (see schematic lower Left). However, tests of landward orientation of Del Mar amphipods on another beach at North Island where landward direction is 15o show that they retain their “home” orientation of 84o and thus head in an incorrect direction (see schematic lower Right).  The special feature of North Island Beach is that it is devoid of obvious landmarks such as cliffs or large sand dunes, and the Del Mar amphipods in this circumstance appear to rely on their sun compass which is still set for their home beach.  By inference, because no experiments have been done specifically to test it, the author suggests that in addition to a sun-compass guidance system, M. one of several schematics showing orientation of talitrid amphipods Megalorchestia corniculata after being released on beaches in southern California under different conditionscorniculata possesses a guidance system dependent upon local beach landmarks. When the two conflict, the landmark system dominates.  Hartwick 1976 Behav Ecol Sociobiol 1: 447.

NOTE  an individual to be tested is placed in the centre of a circle demarcated in the sand.  The compass direction where the animal hops or crawls across the circle is its orientation.  To minimise observer influence during a test, the amphipod is confined within a metal can placed in the centre of a circle drawn in the sand.  A string joining the can to a long pole allows enables the can to whipped away by an observer remaining motionless some distance away. 

NOTE  in landward tests the animals are released at low tide on the moist sandflat; in seaward tests the animals are releaed on the dry sand of the back beach above the high-tide mark

  black dot