C)BSERVATIONS ON COLORATION IN REFERENCE TO BEHAVIOR IN TIDE-POOL AND OTHER MARINE SHORE FISHES

C. M. BREDER, JR.

BULLETIN OF THE AMPRICAN MUSEUM OF NATURAL HISTORY VOLUME 92: ARTICLE 5 NEW YORKE 1948 OBSERVATIONS ON COLORATION IN REFERENCE TO BEHAVIOR IN TIDE-POOL AND OTHER MARINE SHORE FISHES OBSERVATIONS ON COLORATION IN REFERENCE TO BEHAVIOR IN TIDE-POOL AND OTHER MARINE SHORE FISHES

C. M. BREDER, JR. The American Museum of Natural History

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME 92 ARTICLE 5 NEW YORK: 1948 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTOR . Volume 92, article 5, pages 281-312, text figures 1-5, plates 20-23, tables 1-3 Issued November 22, 1948 Price: $.55 a copy

OBSERVATIONS ON COLORATION IN REFERENCE TO BEHAVIOR IN TIDE-POOL AND OTHER MARINE SHORE FISHES

C. M. BREDER, JR. The American Museum of Natural History

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME 92: ARTICLE 5 NEW YORK: 1948 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY Volume 92, article 5, pages 281-312, text figures 1-5, plates 20-23, tables 1-3

Issued November 22, 1948

Price: $.55 a copy CONTENTS INTRODUCTION .287 TIDE-POOL CONTENTS .288 Typical Species . 288 Casual Species . 289 Accidental Species ...... 289 General Remarks...... 290 TRANSFER TO AQUARIA ...... 292

FISHES OF OPEN BEACHES ...... 296 FISHES OF THE CONCH FLATS ...... 302 DISCUSSION.304 SUMMARY 309 LITERATURE CITED ...... e . 310

285

INTRODUCTION IT HAS LONG BEEN KNOwN that tide pools in The opportunity alluded to above occurred many parts of the world are characteristically between July 15 and August 28, 1947, during inhabited by definite types of small fishes, the construction operations of the Lerner some of which are mature while others are the Marine Laboratory on the island of North young of species that spend their adulthood Bimini, Bahamas, British West Indies. It elsewhere. In addition to the characteristic was necessary for the author to give rather faunas of such locations there are also found close attention to building details but there occasiona.l individuals of species that are were intervals when it was possible to give typical of other habitats and can only be some attention to biological matters. The looked upon as casuals or accidentals in tide- establishment of the Lerner Marine Labora- pool environments. The briefest observation tory is a field operation of the Department of of the behavior of the, ge'nerally, several Fishes and Aquatic Biology of the American species demonstrates that the attitudes of Museum and is intended primarily to ex- the various forms differ considerably and pedite the researches of that department and that 'evidently there is some interspecific to serve as a permanent base for studies in hierarchy in existence. marine ecolog and related matters which The opportunity to make some preliminary require long continued or often repeated ob- observations on the subject presented itself, servations in a single locality under more and these are here presented. They are sug- satisfactory working conditions than can be gestive of the desirability of making more ex- usually found in less than permanent field tensive studies from both an observational stations. and an experimental approach. Most of the incidental collecting was under- In addition to these tide-pool observations taken by Mr. Benjamin Dontzin to whom related phenoinena in other environments credit is due. Dr. R. F. Nigrelli, although were given some attention. They consisted occupied with other studies, also gave useful chiefly of observations of the behavior of help in regard to some collecting matters- certain fishes on the sandy beaches that in- Thanks are due to Miss Francesca LaMonte terrupt the. outcropping rocks, forming the and Mr. John Armstrong for notes on the basis of the tide pools. Fishes in more pro- fate of the aquarium and the tide pools sub- tected places and sandy tidal flats were also sequent to August. The author is also grate- the subjects of further correlated observa- ful to Mr. J. T. Nichols for advice on certain tion. All, are compared and considered in the taxonomic matters and to Miss Priscilla light of the present general knowledge of Rasquin and Mr. James W. Atz for reading behavior and pigmentary reactions and serve the manuscript and for giving various help- *to form a basis for further studies on the ful suggestions. ecology of the environiments concerned.

287 TIDE-POOL CONTENTS THE TIDE POOLS on North Bimini from En- Bathygobius soporator (Cuvier and Valenciennes) trance Point to Paradise Point along the SalaricAthys textilis (Quoy and Gaimard) western shore were examined from time to No was found to contain more at the lower pool than time, especially of the low tides. three or four species, and many of them were Certain of these tide pools are shown in plate completely empty of fish life, especially those 20. From these pools the following species higher up the beach. were collected. Nomenclatural usage is dis- The fishes fell clearly into three categories: cussed with other matters under the later the with evident consideration of the individual typical, specializations as- species. sociated with the peculiarities of such en- Jenkinsia lamprotaenia (Gosse) vironments; the casual, with no obvious Mugil trichodon Poey specializations directed towards responses to Euscinostomus gula (Cuvier and Valenciennes) such a and the which Eques pukher Steindachner habitat; accidental, Pomacanthus aureus (Bloch) were seldom encountered and which gave Acanthurus hepatus (Linnaeus) evidences of not being well suited to the tide- Eupomacentrus leucostictus (Muller and Troschel) pool environment and were typically found Eupomacentrus adustus (Troschel) in much larger numbers elsewhere. This divi- Abudefduf saxatilis (Linnaeus) sion, which is purely a classification con- Abudefduf analogus (Gill) venience, places the individual species to- Thalassoma bifasciatum (Bloch) gether with pertinent notes as follows.

TYPICAL SPECIES Baithygobius soporator (Cuvier and Valenciennes) differences in temperature to which indivi- The scale counts were characteristic of dual clutches of eggs are undoubtedly ex- this form as discussed by Longley and Hilde- posed. brand (1941). None was seen with the higher Individuals in the tide pools ranged from scale count ascribed to B. curacao (Met- 15 to 45 mm. in standard length. They were zelaar). Ginsburg (1947), in a revision of the ubiquitous and much more numerous in the American species of Bathygobius, has at- tide pools than any other species. Some were temptedtheveryformidableundertakingofdi- mature, but none were as large as the breed- viding them into subspecies. As only one of his ing fish discussed by Breder (1943). No evi- forms of B. soporator, Bathygobius soporator dence of reproduction was encountered, but soporator (Cuvier and Valenciennes), is from it may be that less "difficult" environments the Bahamas, evidently the present material are resorted to for this purpose or it may have should be referable to that subspecies. It been too late in the season. The large num- would, moreover, be pointless, in present con- bers of very small fish would tend to support nections, to attempt to work out such a sub- the latter view. In agreement with Beebe specific analysis. In any case many investiga- (1931), it was found that this species could tors would prefer to consider such a condition be induced to wave its free pectoral rays a matter of local populations not suitable for under slightly suffocating conditions. The formal recognition in trinomial nomencla- respiratory advantage this evidently pro- ture. The present author, at least, before vides is probably one of the factors which undertaking a detailed study of the matter, make this species so free ranging in these would want to find out experimentally how tide pools, which vary widely as to tempera- much of the observed variations in propor- ture, salinity, and oxygen content. tions and merestic counts is referable to genetic constitution and how much is en- Salarichthys textilis (Quoy and Gaimard) vironmentally superimposed by incidental Individuals ranged from about 15 to 30 288 1948 BREDER: COLORATION IN MARINE SHORE FISHES 289 mm. in standard length. Since the species conditions at no time was the goby noted to attains about twice the size of the largest wave its free rays while the blenny kept the found here, it is probable that all were im- equivalent ones in incessant motion. The mature. This blenny, the only other species physical differences between the pectoral here designated as typical, parallels the goby fins of the two species are shown in plate 21. in a large number of ways, including general Evidently the blenny has, in any case, higher over-all pattern and coloration as well as be- oxygen requirements since it was never found havior. One item which evidently has not so far up the beach as the goby. It showed been previously reported is that its upper no hesitancy in leaping out of any uncovered pectoral rays, although not free but merely aquarium and about the tide pools it was widely spread, are used similarly to those of notably more active and more prone to move Bathygobius. It differs in the use of them, overland. While the goby tended to seek however, in that at all times noted they were shelter in holes and under shells on attempts seen to be in regular and rapid motion. The at capture, the blenny was much more apt two species when kept in a single aquar- to leave the pool and make off overland down ium showed this difference nicely. Under these beach, individuals often reaching open water.

CASUAL SPECIES Eupomacentrus leucostictus mon than E. leucostictus. Both species were (Muller and Troschel) confined to the lower pools. The most prominent of the brightly Abudefduf analogus (Gill) colored casuals were the young of what is evidently this species. Individuals ranged This was the most common species of the from about 10 to 30 mm. in standard length. casuals. It reached higher up the beach than vivid blue and any other of this category. Individuals They were mostly in their ranged from about 10 to 25 mm. in standard yellow phase, the larger individuals gener- found not smaller. length. This species was previously ally being quite so bright as the at Bimini by Dr. W. Beebe and reported by Breder (1933). Eupomacentrus adustus Troschel Abudefduf saxatilis (Linnaeus) This form with the brick red back is what Next to the preceding, this species was the Beebe and Hollister (1931) called E. rubri- most abundant of the casuals, which form it dorsalis and what Longley and Hildebrand followed closely as to numbers, although it (1941) considered the young of E. adustus. did not reach so far up the beach, Individuals Individuals ranged from about 20 to 30 mm. ranged from about 10 to 20 mm. in standard in standard length and were much less com- length.

ACCIDENTAL SPECIES Jenkinsia lamprotaenia (Gosse) in a large sandy depression which could hardly be called a tide pool as herein dis- This species was found in only one pool. cussed. This is evidently an accidental trap- Evidently a small portion of a school became ping. The fishes seemed ill at ease and dashed accidentally trapped in one of the larger what seemed to be an excited state 35 to about in basins. Individuals ranged from about most of the time. 40 mm. in standard length. Etucinostomus gula (Cuvier and Valenciennes) Mugil trichodon Poey Three individuals of about 30 mm. in A school of about one dozen fish, of about standard length were taken from one tide 55 to 60 mm. in standard length, was found pool. This is evidently an accidental trapping. 290 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 92, Eques pucher Steindachner identical description for the young of P. A single specimen about 10 mm. in paru (Bloch), which equals P. arcuatus. standard length was found in one pool. It is (Linnaeus) of Longley and Hildebrand. shown in plate 23. This is evidently an ac- Either the young of these two forms are cidental trapping. This illustration shows almost indistinguishable on a color and pat- what appears to be the smallest size on tern basis or the two descriptions actually record, about 10 mm. in standard length. refer to a single species. Our specimens are Longley and Hildebrand (1941) describe the referable to the species indicated on a basis patterns of young up to 50 mm. without of dorsal spines and the rather difficult but specifying their smallest. Their description recognizable scale counts. agrees quite well with the present individual, Acanthurus hepatus (Linnaeus) the dark pattern of which was jet black. A very few individuals of this species Fowler (1947) figures a drawing of one of were found in pools exposed only at the ex- about 23 mm. in standard length which also treme low tides. They ranged from about 10 agrees, showing only slight differences that to 20 mm. in standard length. While it is are apparently referable to size. The only impossible to be absolutely certain, these fish other species with which this form might be could hardly represent A. caeruleus Bloch confused at these small sizes is E. la-nceolatus and Schneider. They are clearly not A. (Linnaeus). A description of a 15-mm. speci- bahianus Castelnau on a basis of both Breder men by Longley and Hildebrand (1941) in- (1925) and Longley and Hildebrand (1941). dicates that it is very different, closely re- This material raises some doubt on the sembling its adult pattern. synonymization of A. pawnee (Breder) with Pomacanthus aureus (Bloch) A. hepatus as has been done by Longley and Hildebrand (1941). The present material In the lower pools, exposed only on very should be in the "pawnee" stage on a basis of low tides, young specimens of about 15 to size, but it is not, and resembles larger 20 mm. in standard length were notably con- hepatus out of that size range. Either hepatus spicuous because of their intensely black shows remarkable variation in its size at the bodies crossed by brilliant yellow bars. This time of this transformation or there are two form is here called P. aureus on the authority forms confused. Further study of this prob- of Longley and Hildebrand (1941), which lem is being held in abeyance pending the equals P. arcuatus (Linnaeus) of authors and accumulation of more material. not of Longley and Hildebrand. The present specimens agree closely with the color de- Thalassoma bifasciatum (Bloch) scription of the very young of this species This species was found in small numbers in given by the above students except that the pools which harbored either of the preceding fifth (peduncular) bar is missing. Presumably two species. All were of very small size, rang- our material is yet to develop that mark. ing from about 10 to 15 mm. in standard Jordan and Evermann (1896) give an almost length.

GENERAL REMARKS The tide pools of Bimini are typical of variously eroded, often have a sand bottom, many localities in the West Indies, as to some algal growth, and a few invertebrates both formation and contents. Those espe- (generally a few gastropods, chitons, and cially studied were located at such alevel as to small crabs, with an occasional echinoderm) be emptied of water at each spring tide. That and little else of macroscopic size. is to say, any populations contained therein When undisturbed, a given single pool were either destroyed or escaped in each would usually show the same contents tide complete tidal cycle. The pools themselves, after tide, except as to the forms indicated cavities in the old uplifted coral rock which is under the listing of accidental species. Whlen 1948 BREDER: COLORATION IN MARINE SHORE FISHES 291 a pool had been relieved of its contents, it did cal tide-pool forms, is the most abundant not become refilled with fishes on the next species that was encountered. tide, and many went on to the drying of the Another point in connection with the typi- following spring tide without reestablishment cal species is that they could spend their of a new population. Thus it follows that the entire lives in these environments on a size fishes of these tide pools are not merely ac- basis, while the casuals are all young of cidentally trapped but are in regular resi- species sufficiently large when adult to pre- dence, not ordinarily interrupted for one tidal vent their permanent residence in these small cycle. This appears to be true for both the bodies of waters. blennies and the gobies despite the fact that The typical residents are strictly carni- they travel easily overlanid when disturbed, vorous. Their diet includes a few casuals but and is in agreement with Beebe (1931) who seems to be based chiefly on very small in- found that individuals of Bathygobius re- vertebrates. That typicals and casuals were moved from one pool to another generally found in continued residence in certain tide returned to the original pool. pools for long periods and that their return In this connection it is notable that the to emptied pools was slow woulId indicate species here considered typical of the tide that casuals are not an important part of the pools studied, in addition to their respiratory diet of the typical forms. Beebe and Tee-Van specializations and ability to leave the pools (1928) examined the stomach contents of and travel overland, are able to and do Bathygobius and fouxid only small blennies. match the bottom to a remarkable degree. On the other hand, the casuals are all non- Their pattern and pigmentation are changed predatory, and feed almost exclusively by rapidly, but, in addition, when engaged in picking minute bits of whatever animal and combat or other social behavior, their colora- vegetable matter becomes available. This tion may run counter to the background and method of feeding is true of the accidentals render them for the time very conspicuous. with the exception of Eucinostomus and The casuals on the other hand show no color Mugil which grub in the sand for food to a matching ability at all, and with their fre- marked extent. quently bright colors are usually conspicuous Because the behavior of the fishes in the objects. The least colorful of them, Abudefduf pools could be studied only during the brief analogus with its dark brown and light intertidal periods and because only the lower pattern, is distinctly conspicuous when pools contained any assortment of Kshes (the swimming freely over most of the tide-pool upper pools usually had none), some individ- backgrounds, but when alarmed it often hugs uals were transferred to an aquarium. closely to a rock or shell and rests quietly. At The hurricane that passed close to Bimini these times it becomes fairly inconspicuous. on September 16, 1947, stripped the tide. This is not true of the other species listed in pools of all fish life, except for Bathygobius, this category. It should be noted in this con- and considerably modified other shore en- nection that A. analogus, except for the typi- vironments. TRANSFER TO AQUARIA FOR THE FURTHER STUDY of these fishes an the sea surface over a clean sand bottom. The aquarium 2 feet by 1 foot by 1 foot was em- tide-pool fish paid no attention to them ployed. It was fitted with the local light- whatever, and they remained near the sur- colored coral sand and six shells of various face of the water. kinds and sizes, one of which was that of a The following day the peck-order was young Strombus gigas well covered with a found to be the same, but there was much growth of Batophora (see pl. 22). Standing less evident friction between the various in- water was used which was refreshed by chang- dividuals. Later the same day the following ing a single bucket full, usually every third fishes were added to the aquarium: day. The aquarium was placed on a veranda Eucinostomus gula 1 example in such a position that it received direct sun- Eupomacentrus adustus 1 example light in the early morning and late evening Thalassoma bifasciatus 2 very small examples but not in the middle of the day. Neverthe- less the temperature rose on occasion to as The red-backed Eupomacentrus was barely high as 360 C. without any evident inter- smaller than the dominant fish and im- ference with the activity of the specimens. mediately took issue with the dominance of This should be expected because of the vast the first, and the order was challenged for a temperature changes to which small tide time, but eventually the newcomer took pools are subjected and to which the fishes second place, relegating the former second found within them are evidently well ac- fish to third place. commodated. Although the Thalassoma were very much After the aquarium was well established smaller than the others, they were little the following fishes were placed within it: bothered and seemed to be outside the peck- order Abudefduf saxatilis 3 examples hierarchy. Eupomacentrus leucostictus 2 examples The Eucinostomus held to open water and Abudefduf analogus 3 examples was picked on chiefly by the semi-open-water A. saxatilis. By the next morning both its Almost immediately a peck-order was caudal tips had been removed, and the fish established. The two largest fish were about was taken out of the aquarium in order to the same size: one A. analogus and the other avoid its destruction. The other fishes, E. leucostictus. The latter quickly dominated principally the A. saxatilis, had risen high the former, which became clearly the "num- in the water in order to attack it, the most ber two" fish. The rest, all considerably vigorous attackers being not over half its smaller, were not clearly recognized as to length. It is to be especially noted that at this peck-order, and there was evidently no time not one other fish showed the slightest "omega" fish. All clearly showed a close as- caudal or other damage and that at no time sociation with the shells, except A. saxatilis was the Eucinostomus seen to attempt to which was evidently only "half" in the peck- defend itself. Although the Eucinostomus order system and kept relatively up from the was actually taken from a tide pool, this bottom. The rest were all given to hiding in species is not typical of such places and must, the various shells. The dominant fish took as previously noted, be looked upon as ac- over all of the shells, busily circulating from cidental. This species is more nearly typical one to another and rooting out any possible of white sand flats, and its behavior in such occupants. The arrangement of the sand at localities is discussed farther on. the entrance to one shell was modified from It is notable that the peck-order in this time to time by vigorous fanning with the aquarium was nicely graded, in addition to tail, just as the adults arrange their home the usual size order, not as to pattern but as sites. to darkness of pigmentation. Under the To this population was added two very same conditions of illumination and back- tiny postlarval Atherina stipes Miiller and ground, the order of the various species, ar- Troschel, not from the tide pools but from ranged according to darkness of coloration, 292 1948 BREDER: COLORATION IN MARINE SHORE FISHES 293 becomes identical with the peck-order as to pursue any other fish and was only oc- established by observation of behavior. The casionally pecked by others. largest fish of each kind-and all these were Fourteen days later the peck-order among very nearly the same size-may be arranged the free-swimming fishes had become modi- in the order of both peck relationship and fied so that the Acanthurus was definitely darkness of pigmentation, as follows: number one, but there was little fighting,

Eupomacentrus leucostictus Darkest Eupomacentrus adustus Nearly as dark Abudefduf analogus Large dark blotches Abudefduf saxatilis Narrow vertical lines Rucinostomus gula Light and nearly unmarked

The Thalassoma, which were quite dark, and all kept clear of this tang which "showed" the very small but brilliantly yellow Eupo- its peduncular spine to the more timorous. macentrus, and the transparent and tiny The blue yellow Pomacentrus was number atherinids were evidently not exactly in- two and the red-back was number three. The cluded in the peck-order on a basis of in- latter had lost its red back and was a general significance of size. If a larger fish ap- washed-out gray. It was clearly not sickly, proached they simply "scampered" out of and this is evidently a normal pigmentary the way, but there never seemed to be any change. attempt to peck them on the part of the The sequence of the events in this aquarium larger fish. is given in table 1 together with the moon Three days later six Bathygobius soporator phases which during this period had twice were added to the aquarium. Two of the emptied the majority of the natural tide gobies were a little over twice as long as the pools from which the fishes came. prior residents, two about as long, and two The fish were fed every day on bits of fish, about half as long. The pomacentrids took craw-fish, or similar material, which all took considerable notice of this new introduction. freely, the gobies bolting great chunks of The red-backed one caught one of the smaller meat and the others picking it apart in small and shook it as a dog might shake a rat, but bits. By this time a considerable growth of it escaped, evidently unharmed. This same algae had accumulated on the shells, and the pomacentrid later backed up to one of the rear glass was purposely never cleaned. The larger fishes and attempted to "fan" it away free-swimming fish all spent considerable in very much the fashion that this species time picking on this algae so that it clearly moves sand in modifying its retreats. showed their tooth marks. Although they fed From here on the Bathygobius took charge freely on the animal matter offered, they of the situation. The two atherinids and the spent most of the day working on this algae small Thalassoma quickly disappeared, pre- and their excrement became largely a bright sumably being eaten by the gobies. The peck- green. Simultaneously, the largest blue yellow order previously described for the free- pomacentrid became much more intensely swimming species, while still in some evidence yellow, and it is suggested that this type of if looked for critically, was all but destroyed food is directly connected with the intensity by the presence and interruptional activities of the yellow shown by these fishes. Such yel- of the dominating gobies. low as is displayed by the other forms re- Seven days later a medium-sized Salarich- mained brilliant, a condition which generally thys textilis and a small Acanthurus hepatus rapidly disappears in aquaria where only were added to the aquarium. The latter was animal foods are available. Longley and about the size of the largest Pomacentrus. Hildebrand (1941) indicate that Acanthurus Although it was not much bothered by the " . . . is almost entirely herbivorous," while others it seemed to be low or lowest on the Beebe and Tee-Van (1928) wrote, "The peck-order, that is to say, it was never noted powerful gizzard-like stomach contained un- 294 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VIOL. 92 identifiable vegetable and animal matter." night. Some individuals would be in their The latter wrote, in reference to P. aureus, closest sand-matching pattern while others "Alimentary canal crammed with algae, would be conspicuously dark, in some cases hydroids, etc.," while the former indicated nearly black. Since at least four of these fish that "The diet includes algae and a great were not sexually mature, this is evidently varety of sedentary animals." something not directly connected with sex or A sharp differential was noted in the night sex behavior. The night patterns were no- behavior of the fishes in this aquarium, nearly tably bolder, with a considerable amount of every night that it was in maintenance. The black-and-white, checkerboard effects in bold following observations were made well after blotches, while the daytime patterns ran more TABLE 1 CHRONOLOGY OF THE TIDE-POOL AQUARIUM 1947 Date Species and Data Moon Phase July 24 First quarter 28 Aquarium established with Abudefduf analogus 3 Abudefduf saxatilis 3 Eupomacentrus keucostictus 2 Atherina stipes 2 30 Added the following: Eucinostomus gula 1 Thalassoma bivittatus 2 Bupomacentrus adustus 1 31 Removed Bucinostomus Aug. 1 Full moon 3 Added the following: Bathygobius soporator 6 9 Added the following: Last quarter Acanthurus hepatus 1 Salarichihys textilis 1 16 New Moon 23 Last observation First quarter Acanthurus no. 1 dark when a flashlight was played suddenly to fine pepper-and-salt effects. The darker on the aquarium. fish tended to dominate the bottom-matching Batlhygobius soporator was in and out of light-colored fish in the daylight. These pat- shells both by night and by day, apparently terns were in general agreement with the indifferently. The color pattern which they notes of Beebe (1931), but they are evidently showed is difficult of interpretation and can- much more complicated than he indicated. not be clearly defined without further de- Both species of Abudefduf and the Acan- tailed study. At any one time a variety of thurus rested quietly near the proximity of patterns would be noted, both day and shells for the most part after dark but in no 1948 BREDER: COLORATION IN MARINE SHORE FISHES 295

case did they hide in the cavity of a shell At no time were any respiratory difficulties itself. noticed in any of the fishes in this tide-pool After dark both species of Eupomacentrus aquarium. Bathygobius, which waves the free disappeared completely within the deep pectoral rays in water of low oxygen, content recesses of the shells, and only on one occa- as an evidently accessory respiratory mech- sion was one even seen, and this one was anism, was never noted to have recourse to well within a shell cavity. Although by day this behavior, while the blenny was seen to these fish will retreat deep into a shell when wave its upper pectoral rays at all times. alarmed, they spend most of their time in the The arrangement of the aquarium used in open, but evidently at night they rest com- these studies is shown in plate 22. This pletely hidden. aquarium was maintained for a period after The Thalassoma either disappeared into the departure of the author. Little by little shells or buried in the sand on the coming of the fishes disappeared until, about the middle darkness. The Eucinostomus rested quietly of October, the largest Bathygobius was the in open water well up from the bottom at sole survivor. Just how the loss of each fish night in about the same place in which it occurred cannot be determined, but it is sus- spent the day. The atherinids spent all their pected that with the growth of the fishes time at the surface of the aquarium, both the largest goby either harried or directly night and day. The Salarichthys was both in ate the rest. At least it was seen eating some and out of shells, day or night. of the individuals. FISHES OF OPEN BEACHES REFERENCE HAS ALREADY BEEN MADE to the Young Sphyraena barracuda are found in behavior of Eucinostomus gula in tide pools the same places and show a similar consider- and of its more typical occurrence on the open able ability to change their mottling or re- sandy beaches. Here great numbers in more place it entirely with a silvery vesture. This or less loose aggregations of various sizes are species acts almost exactly like Eucinostomus common. Fishes of like size are usually found over a plain sand bottom in respect to pig- together and range all the way from large, mentation. Both show a black tip to the first fully mature individuals to the smallest dorsal and in addition Sphyraena shows a juveniles. The bottoms over which they are black tip to the second and to either caudal most frequently encountered are either lobe (see fig. 1). These two species present mottled, owing to the presence of trash, a remarkably similar effect, in an over-all shells, or other objects darker than the sense. From above the surface both are nearly white sand, or large patches of per- much more conspicuous than when viewed fectly clean, light-colored sand. In the former laterally through goggles. Since the water in localities the fish show a mottling which which this phenomenon takes place is 3 feet renders them very inconspicuous. Over the or less in depth, such a view is possible if one clean sand they lose this mottling and be- swims in towards the shore and "walks" along come equally inconspicuous in a silvery on one's fingertips while floating face down- vesture which takes on the color of the bot- ward. Under such conditions it is not easy to tom itself. The only markings noticeable are associate the black markings with a fish. It the black pupil and a black tip to the dorsal would seem that these dark marks make up fin. some kind of recognition pattern for the Many if not all of the members of the Eucinostomus since they are a well-marked, family Gerridae root out invertebrate or- aggregating form, groups frequently showing ganisms from such sandy places and they themselves as indicated in figure 2. This may be frequently seen in such activity. Of kind of recognition could operate similarly E. havana (Nichols), Longley and Hilde- for the solitary Sphyraena, in a negative brand (1941) had the following to say: sense. Since, however, the latter preys to a "Nothing more is required to bring together considerable extent on the former and may be such as are near than to wade out into the seen stalking them over these open sandy water a little more than waist deep and stir beaches, where there is nothing to hide be- up sand, which calls them to feed. With them hind, it is not inconceivable that the general comes Eucinostomus lefroyi." In this connec- similarity of these markings may be a de- tion it is noted that with the species here ception practiced on the Eucinostomus. In under discussion more or less numerous in- fact a lone barracuda could look not unlike dividuals would be found to be staying rather a small group of Eucinostomus of a size ap- near a bather or a group of bathers on the propriate for the Sphyraena to prey upon. outside beach. On August 27 the following The way in which this conceivably could note was made in reference to a bathing party work out is indicated in figure 3. The only late in the day: "An eight-inch Eucinostomus disruption would be the tip of the lower became curious about our feet and followed caudal lobe. Groups of fishes about as shown us about for the entire time we were in the in figure 2 must be frequently in the visual water. It was evidently interested in the field of solitary individuals that have become materials our feet scuffed up and was en- separated from their companions. To a tirely fearless." It is well known that many human observer it is possible to see such species of fishes will follow the rooting be- separation at virtually any time. It is to be havior of other species. The above-noted especially noted that a group of three as behavior clearly places Eucinostomus among shown in the diagram have only one eye these, although its protrusile premaxillaries looking outward on the side resembling a and observed behavior show that it can and barracuda and three on the side in which this does forage independently. resemblance is modified by the very presence 296 BULIN AMER. Mus. NAT. HIST. VOL. 92p XA,M20

40-!9RTl"i! I" --,F- - "'.. . wo-s_____ r _T

At. ' - .-

f.4. -. -.

2 Point, 1. The tide-pools at Entrance Point, North Bimini, Bahama. 2. A tykl tide-pool at Entrance showing one Abudefduf analogus Buuzr Amim. Mus. NAT. Hisr. VOL. 92, PLATE 21 :ioI ..

5 1. Quartering view of Sakirickihys kxtiis, showing membrane connecting upper pectoral rays and positiOR in which held when used for apparently supplementary respiratory efforts 2. Lateral vew of pose of * larichihys textiis. 3. Semi-latera view of Bawygobius soporator, showing upper free pectoral rays which a apparently used for supplementary respiration. 4. Dorsal view of right pectoral of Bathygobius soporad 5. Anterior quartering view of left pectoral of Ba*kygobius soporator BULLEIN AMER. Mus. NAT. HIST. VoL. 92, PIATE 22

I

S

K

2 1. Carcharhinus, showing the pattern formed by wave-focused light rays on its back. 2. The "tide-pool" aquarium, showing Abudefduf analogus (upper right) and Pomacrus leucostic4us BuuzrTm Am. MUs. NAT. HISr. VoL. 92, PLATE 23

4r !-WI.E. 11 I -1 .6. qlq. ,, i ,"# -k - ". -0- ki -,NNF4pmw....-o - lw4m

I

2 1. Young Tylosurus raphidoma, showing dark mandibular lappets and a dark dorsal, 150 mm. in standard length. This photograph is of a living fish as it rests near the surface of the water. 2. Young Eques pukher resting in the protection of shells. About 10 mm. in standard length 1948 BREDER: COLORATION IN MARINE SHORE FISHES 297 of the two additional eyes. The larger Gerres mesh-like network of brilliant lines which cinereus (Walbaum), also present on these usually are attended by spectral colors at beaches and too large for the little barracuda their edges. Under certain conditions this to eat, shows no such markings, while its effect can make an object peculiarly distinct, young of comparable size were not found in but in present connections it is evidently of these localities. Barracuda of a size ap- minor or no importance to the conspicuous- propriate to feed on these larger fish have ness of the small fishes under consideration. moved off into deeper water, evidently for This is because of their small size in reference natatorial reasons. to the usual sizes of the "mesh" of bright In any case it is thought that the pig- lines and the slight distance between the fish mentary conditions found over these light and the bottom. The meandering lines move sandy bottoms are associated with the im- over fishes and bottom alike without, in any possibility of having the dark pupil fade with evident way, making the fish more promin- the rest of the exposed surfaces for basic ent than they are under cloudy conditions visual reasons. Walls (1942) gives an espe- when no mesh of bright lines is present. cially pertinent discussion of the difficulties However, under different conditions and involved in an attempt to "hide" the pupil, with larger fishes swimming in deeper water particularly in the case of silvery fishes, and this effect can make the fish exceedingly Cott (1940) gives an extended discussion of prominent. For example, a shark such as the entire problem. Carcharhinus swimming in deep water re- In the strong light usual over these places mote from the bottom and in the focal plane a shadow is cast on the bottom which in of these bright, actually golden lines stands shallow water is nearly as distinct and dark out prominently as far off as it can be seen. as an equivalent one in air. As viewed from The all-over gray appearance produced by above, it is a very distinct mark and often the exquisite countershading of these sharls, serves to betray the presence of a fish before which otherwise renders them so inconspicu- it is actually seen itself. In deeper water the ous, is completely nullified. A shark made shadow lightens and becomes more diffuse prominent by these bright lines is shown in and is usually too remote from the fish to plate 22. This photograph was taken at the seem to have much significance. Viewed from Marine Studios near St. Augustine, Florida under water, laterally, such shadows are not (Breder, 1938). In a state of nature sharks nearly so conspicuous as might be thought are not usually seen under such conditions, from an above-the-surface view. When the and it would seem, on a basis of general field fish is close enough to be seen distinctly, the observation, that they avoid depths in which shadow is likewise distinct, but when the the phenomenon is possible. Most usually fish is at such a distance as to show only the large sharks are to be found in water too deep markings herein discussed the shadow is im- to be in the focal plane of these bright lines possible to detect. It would be difficult, if not or very close to the surface in water too shal- impossible, to imagine how a fish by some low for the effect to be seen. On the sandy chromatic adjustment could do anything shores under consideration sharks are seldom tending to obliterate such a shadow, in any seen in the daytime in any case, but as soon case, and there has been apparently no ad- as night falls they come in from deeper water justment in such a direction. and spend the night cruising over these areas Another detail of this environment is the for feeding purposes. On every moonlit night result of bright, direct sunlight playing their shadowy forms could be seen in num- sur- bers from the laboratory dock. through the often only lightly rippled to color face. Under these conditions, which may oc- The ability of barracuda change of time, the tiny is much more marked than is generally cur daily for long periods adults of 3 or more feet wavelets so focus the sun's rays as to produce realized. The large distance in length are usually seen in the daytime to a pattern of bright lines at a certain black dorsal sur- below the surface. When this focal distance have a very dark, greenish the face and a silvery under surface. However, approximates the depth of the water, this dark back is replaced by bottom is seen to be covered with a moving, after nightfall 298 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 92 one nearly as light as the ventrum. A con- fish usually swims or rests high above the siderable number of these fish kept in a large bottom. In this phase the green above passes enclosure 25 by 75 feet showed this diurnal gradually to silver on the side, and this more change very clearly. Large flood lights thrown abruptly to plain white upon the belly. In on them at night frequently caused anglers, this phase the fish is readily seen from above, quite familiar with the species, to suppose at but if viewed at even a short distance by a first sight that the fish were dead and floating diver working at its own or lower level it belly up. A few minutes of exposure to the seems gray, and is readily lost in the blue- floodlights brought back the dark back of gray haze of the water. In the banded phase A

B

I . I . I . I I. ,

FIG. 1. Diagram of pattern assumed by certain fishes over a clear sand bottom, free of litter. A. Sphyraena barracuda. B. Eucinostomus gula. daytime. Longley and Hildebrand (1941) the body is crossed by seven distinct bars. seem to have been the only students to have This pattern is displayed when the fish previously discussed at length the extent of rests over dark or variegated bottom. For the pattern-changing capacity of this species. example, of two 375-mm. fish seen at once at a Their remarks,-which are in essential agree- dock at Fort Jefferson, one lay in the shadow ment with present observations, are as fol- of a cluster of piles, and. the other was in lows: "The color is changeable in high bright sunlight over light-colored sand. The degree. The young fish from one to several first was dark and conspicuously banded; inches in length have a dark lateral stripe. *the second, pale and uniform in coloration The larger fish show a banded pattern that except for its countershading. Another ob- appears also in the young, as well as a plain servation was made upon a group of young one, countershaded and marked only by a barracudas, about 125 mm. long, at the Lab- number of black spots along the lateral line. .oratory dock. Some were over light-colored '"The plain pattern is that in which the sandy bottom, and were light and uniform in 1948 BREDER: COLORATION IN MARINE SHORE FISHES 299 color. Others were over bottom littered with basking near the surface of the water in the brown algae, black sea-urchin spines, and neighborhood of a coral head, a buoy or a dark-colored waste, and these were darker channel stake, appears much as a ghost fish, and plainly banded." a shadowy wraith. Another lying near the

FIG. 2. A group of three Eucinostomus as they frequently arrange themselves when resting quietly over a clear sand bottom. Scale same as in figure 1.

A / IF 0 44

B '40 A00

FIG. 3. Appearance of certain fishes over clear sand bottom as seen horizontally from diving equip- ment. A. Sphyraena barracuda. B. Three Eucinostomus grouped as in figure 2. This illustration is a tracing of figure 1 and figure 2, omitting the structural details not visible under the conditions in which the living fish are seen. Scale same as in figure 1.

Gudger (1918), after describing the pat- bottom over coral sand will so accommodate tern of various specimens in death or as pre- itself to its environment as to be almost un- served, made the following observations: "In noticeable so long as it remains quiet. Wood- iife the color of the fish readily accommodates Jones (1912) says that it is the hardest of all itself to its surroundings. A large barracuda sea fish to see." 300 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 92 The only other species noted to participate show the black pupil and a black edging to in this general sandy beach complex of show- the caudal. All those seen were clearly too ing a black-and-white effect was Mugil large for the little barracuda to bother. It trichodon, found in occasional small groups was further noted that there were two types of generally from three to five individuals. of black edging, as shown in figure 4, one These would have the sand color already de- about twice as wide as the other. It proved scribed for the former two forms and would impossible to catch any, but it is suspected that the differences may be sexual, as the A fish were all large enough to be sexually 400 mature, and groups were not infrequently

0 seen in which only one had a narrow band + while the rest, one to four, had the wide band. Along the outside beaches, actually in the light surf, not nmore than a few feet from 400'r shore the young of Trackinotus palometa Regan ran singly or in small groups, espe- cially towar(ds evening. Viewed from the sea- ward side, as from diving equipnment, they 0 show two strong diagonal black marks which indicate their long dlorsal and anal, as in- dicated in figure 4. Thus they evidently belong to this group of clean sand fishes which are mostly black and white. In slightly deeper water a few feet farther off shores species of 7'ylosurus aboundled, which when viewed laterally from (living equipment were similarly not at all easy to see distinctly. T'he young of some of these D have a dark dorsal and clark maandlibular lappets as described by B3recler (1932). Just 0 what connection this may have with the presently (liscussedI complex, if any, is not clear but coul(d be made the subject of further study. A photograph taken of a living fish on the west coast of Florida (pl. 23) shows E this con(lition in young T. raphidoma (Ran- zani) as seen dorsally. The verycdark young of Chaetodipterus faber (13roussonet) and their behavior over + 0 light sand bottoms have been discusse(d in considerable detail by Breder (1946). The phenomenon involved was that of black fishes scattered on a light ground together with other dark objects (infertile mangrove seed pods) which they so closely resembled FIG. 4. Comparison of variotis black marks on that it was extremely difficult to distinguish fishes as seen over a clear sand bottom, all reduced fish from inert items. As soon as these fish to a common scale. The light crosses indicate the a too to be easily confused total lengths where pigment is absent. A. Sphy- reached size large gula, with the inert objects, they moved off into raena barracuda, young. B. Eucinostomus or young. C. Mugil trichodon (mature male?). D. deeper water and took on aggregating Mugil trichodon (mature female?). E. Trachinotus schooling habits and developed the typical palometa, young. banded coloration of the adults. At this place 1948 BREDER: COLORATION IN MARINE SHORE FISHES 301 on the Florida west coast the water is beaches that they resembled in any great notably turbid, owing chiefly to the abun- detail, they nevertheless reclined and drifted dance of plankton organisms, so that the about so that they looked like some bit of phenomenon above noted could take place trash amid other more or less dark objects successfully only in very shallow water of a of greater or smaller size of a jetsam nature. few inches at most. They were not readily recognizable as fishes. On the other hand, at Bimini, where the Large adults were also seen in small num- water is remarkably clear and vision is ex- bers but always as solitary individuals. These cellent even at considerable depths, the be- presented themselves in one of two manners. havior of medium- and large-sized Ckaeto- They could be found lying as inert objects in dipterus over clean sand bottoms was found clear places on light sand and just as black to be notably different. Although none of the as the above-mentioned smaller sizes. Evi- sizes resembling small mangrove seed pods dently this type of behavior carries on into was encountered, specimens were secured the adult size in extremely clear water. Those that were considerably larger, and compa- seen behaving in this fashion were found in rable in size to those on the Florida coast that water from about 6 to 10 feet deep. Other schooled and showed the banded coloration individuals could be found under docks, and typical of adults. The smallest specimen these were in a normal vertical position and taken, 58 mm. in standard length, looks very showed the most contrasting black-and-white like the illustration of a 50-mm. specimen coloration of which they are capable, evi- in Hildebrand and Cable (1938, fig. 29) ex- dently to some extent matching the small cept that the pattern and blotching are piles that support the docks. One individual barely evident, even in the preserved speci- was first noted about 30 feet off the end of the men from Bimini, because of the much heavier laboratory dock, reclining on its side with over-all pigmentation. These at Bimini, virtually no perceptible motion. Little by moreover, were maintaining a solitary life, little it inched towards the dock, an opera- similar to that of the very small ones on the tion which took nearly an entire afternoon. Florida coast, were as black as they, and It then righted itself and rapidly took on the displayed similar hyaline pectorals and strongly banded pattern. Evidently this caudals. Furthermore, their behavior on the same fish lived under the dock for the rest of sandy shores where they were found can only our stay, a matter of 38 days. During this be described as similar to those smaller seed- time it always showed the contrasting black- pod resemblers from Florida. Although there and-white 'pattern, in daylight or at night. was nothing in particular on the Bimini FISHES OF THE CONCH FLATS EXTENSIVE SANDY FLATS form the eastern to count the number of conchs and fish to margin of the harbor- at Bimini, and parts obtain the ratio of fish to mollusk. These of them become dry at very low water. These counts were made on material taken on the flats are populated by large numbers of flats east of the laboratory dock and are mollusks and other invertebrates of various given in table 2. sorts, typical of such protected places in the In contrast to this ratio of one fish to Bahamas. The fishes associated with them fewer than 12 conchs, a collection of Strombus are also typical, but overlap in species those samba Clench made outside the harbor a already discussed under both Tide-Pool little east of Turtle Rock August 25, consist- Contents and Fishes of Open Beaches. One ing of 319 conchs, was found to contain not TABLE 2 COMPARISON OF THE NUMBER OF Strombus COLLECTED WITH THE NUMBER OF Astrapogon FOUND Date No. of Strombus No. of Astrapogon No. of Conchs per Fish From harbor flats July 26 20 1 20.00 Aug. 1 61 6 10.17- 6 41 3 13.66+ 7 21 2 10.50 9 211 7 30.43- 12 103 16 6.44+ 19 18 3 6.00 20 48 8 6.00 All 523 46 11.15+ From near Turtle Rock Aug. 25 319 0 species is largely, if not entirely, confined to a single fish. If the same ratio of conch to close association with these conch beds. fish existed here as in the harbor about 20 The inquiline association of this species, fish should have been recovered. Evidently Astrapogon stellatus (Cope), within the gas- either this association is chiefly, if not en- tropod Strombus gigas Linnaeus is well known tirely, characteristic of the inside banks at (Plate, 1908; Hildebrand and Ginsburg, 1926; this place, or it is a matter of specificity. Gudger, 1927, 1929), but there appears to be During this time of year, at least, one may no data on its frequency of occurrence, or expect from every sixth to every twentieth on its variation as associated with habitat. conch to contain an Astrapogon, with a mean Consequently when it became necessary to at a little over 11. An examination of the handle a considerable number of conchs, note tabulation of numbers indicates that there is was made of the number of individuals that a constant ratio of conchs to fish. This is contained such a fish. When the conchs were shown more clearly in figure 5. There was no taken and placed in a boat they naturally particular system to the collection of these withdrew within their shells, and any Astra- conchs, which were taken for use as human pogon that they were harboring were se- food and as bait. Various parts of the banks curely trapped therein. As they relaxed the were collected over, and this constant ratio fish dropped out and it was a simple matter in number of conchs per fish would seem to 302 1948 BREDER: COLORATION IN MARINE SHORE FISHES 303 indicate that the "infestation" of these in- quiines is not dependent on the concentra- tion of conchs. The collector simply picked up more of the mollusks when he found a denser concentration at whichever low tide he happened to work. The reason for the absence of the fish in outside waters is not clear, and it may be that these shelter-seeking fishes merely avoid the open ocean and confine themselves to sheltered areas, a possibility that could be made the subject of further study. Determination of sex was not made in the earlier collections, but all Ash were examined for the presence of eggs in the mouth, since the species is known to be an oral incubator. As time went on, more of the females were noted to be heavily gravid. On August 12 the collection consisted of nine gravid females, six males, and one immature fish. On August 20 two males were carrying eggs in their mouths. Evidently, then, the fish were just coming rinto their spawning season, towards the end of August. There appears to be a slight decrease in the ratio of fish to mollusk with the passage of time, as is indicated, but this shows no statistical significance. As the spawning sea- son approaches, it nevertheless suggests that there may be a tendency for the fish to stay more within, or closer to, their retreats in the conch. Thus more would be apt to be FIG. 5. Comparison of number of Strombus col- trapped when the conchs were taken, but lected with number of Astrapogon found, at the more collections should be made for purposes harbor flats and outside the harbor, at Turtle of statistical study. Rock. From data of table 2. DISCUSSION

ALTHOUGH VIRTUALLY nothing has been of peculiar interest. In addition to the three done, except in the most casual way, on the particular environments mentioned specific- subject of the mutual relationships of be- ally herein are others that bear a similar havior, environment, and pigmentation of relationship to the whole, although it was im- marine shore fishes, the author has long be- possible specifically to investigate them at lieved that a study of their evident interac- this time. They include beds of Thalassia, tions would prove exceptionally useful in any which have rather abrupt meetings with the attempt to evaluate the relative significance open sandy bottoms, and mangrove stands, of physical status and behavior in relation to which likewise meet the others with abrupt- survival and selection. In fact so little is ness, to mention only the conspicuous chief known about the general manner of life of ecological shore conditions. any considerable number of fishes which might Another abrupt and specialized environ- be so studied that much basic work must ment, which must be considered as "artifi- first be undertaken in order to provide a cial," is the shelter of docks and wharves satisfactory foundation for procedure. This which harbors a large variety of fishes not paper is to be considered a first move in this typical of any of the three environments direction. The very fact that so many items here discussed. Many of these are associated of pertinence to the subject could be un- with rocky places in deeper water. For ex- covered by the simple procedures indicated ample, the following species were noted to be in the body of this contribution is indicative habitually under the laboratory dock:

Lutianus griseus (Linnaeus) Mature Lutianus apodus (Walbaum) Mature Ocyurus chrysurus (Bloch) Half grown Haemulon sciurus (Shaw) Mature Haemulon plumieri (Lac6pede) Mature Upeneus martinicus Cuvier and Valencier Mature Chaetodipterus faber (Broussonet) Mature Chaetodon striatus Linnaeus Mature Pomacanthus arcuatus (Linnaeus) Half grown Pomacanthus aureus (Bloch) Half grown Holocanthus tricolor (Bloch) Mature Angelichthys ciliaris (Linnaeus) Mature Acanthurus caeruleus Bloch and Schneide: Mature Acanthurus hepatus (Linnaeus) Mature Eupomacentrus leucostictus (Muller and T Mature Abudefduf saxatilis (Linnaeus) Mature Iridjo bivittata (Bloch) Mature and half grown Thalassoma bifasciatum (Bloch) Mature and half grown of the truth of the above statement. These, Many of the species above listed could also and many more as yet undescribed details be found among the mangrove roots, which together with rigorous experimental pro- these docks, from a fish's point of view, must cedures, should go far to explaining the more or less resemble, and to which is added many superficially inexplicable variations of the attraction that human beings not infre- structure and behavior displayed by such quently drop or dump ovberboard things fishes. that are edible. Because these three specialized environ- On the Thalass'ia beds visible from the ments abut one another, yet can hardly be dock the prominent fishes identifiable mostly said to intergrade, such a region becomes one at sight included the following: 304 1948 BREDER: COLORATION IN MARINE SHORE FISHES 30S Pomacentrus leucostictus (Muller and Tr Mature Iridio bivittatus (Bloch) Half grown Sparisoma flavescens (Bloch and Schneid Mature Lactophrys trigueter (Linnaeus) Mature and half grown Lactophrys tricornis (Linnaeus) Mature Spheroides dorsalis Longley? Mature Chilomycterus schoepfi (Walbaum)? Mature

The truly schooling species, sensu stricto, Although the tide pools are cut off from found in these environments wander more or the sea only at certain tidal stages, the fish less freely in and out ofSall of them except the in them remain essentially the same for at tide pools, where only rarely are any of them least long periods of time and very definite trapped and then seemingly by the merest hierarchies are set up. These are dominated accident. All may be taken in much deeper by the species specialized in the direction to- water and apparently are only casuals close wards making them thoroughly accommo- to shore. The species observed here include the dated to this type of environment, while following: those not so obviously accommodated to it them- Jenkinsia lamprotaenia (Gosse) are in an inferior position but among Harengula macrophthalma (Ranzani) selves set up their own hierarchy, which Trachurops crumenophthalma (Bloch) is superior to the fishes that are not suited to such environments. An exception to this The second-listed species usually shows a is the final dominance of the young Acan- dark humeral spot and the last a dark opercu- thurus, which, while generally resembling lar spot, while the very small, first-listed the form of the other competing fishes, form shows no such distinctive mark recog- is provided with a particularly vigorous de- nizable at any distance. These fish always fensive mechanism in its peduncular lancet travel in groups of numerous, often very that it uses freely. Since these tide pools are large numbers of, individuals and because of to all intents and purposes naturally consti- their close packing in the tight schools can tuted "aquaria," it is not surprising that all show no such "recognition" marks as the such forms thrive in man-made aquaria, or resident sandy-bottom forms. This lack of that those accidentally trapped do not thrive distinct marking and their particular social especially well in small aquaria. As a case in behavior mark them as apart from the latter- point, at the old New York Aquarium all mentioned type, in addition to their wide the "typical" and "casual" species thrive ranging habits and only occasional appear- very well if not too crowded, but none of the ance on these sandy shoals. "accidental" species in the sizes here found A considerable amount of work has been were ever brought alive from the south, with done on the relationships of various, mostly the exception of Eques which proved to be fresh-water fishes, including learning, social expeptionally hardy. This form, although not behavior, pattern, and color recognition, regularly in tide pools, is typically found under captive conditions, but there has been amid shell piles and similar places. Plate 23 scant attempt to correlate the results of these shows an Eques resting in the shelter of some studies with the actual interplay of the vari- shells. Pomacanthus, Acanthurus, and Thalas- ous phenomena under feral conditions. Breder soma all do fairly well under such conditions and Halpern (1946) attempted to bring to- of captivity in their larger sizes which are gether the data on the various influences at never found in tide pools. Jenkinsia like most work on the aggregating behavior of the gold- herrings is almost impossible to handle alive, fish and other species, while Breder and Ras- and Eucinostomus never survived the trip quin (1947) indicated how a fine structural north. difference may greatly influence the behavior One of the more interesting points in this of closely related forms and discussed the connection was that in the tide-pool aquarium relationships between such features and a both size and darkness of coloration were very specialized type of environment. associated with peck-order, the darkest being 306 BULLETIN. AMERICAN. MUSEUM OF NATURAL HISTORY VO0L.. 92 at the top and the lightest, and non-suited cal. It is to be noted in thlis connection that to tide-pool environments, at the bottom. both are heavy fishes, sinking to the bottom The lightest, Eucinostomus, moreover fitted when not actively swimming, while the others very well into the environment of open are all "floating" fish. Furthermore, the goby beaches where its coloration evidently had has developed an excellent hold-fast formed of very special significance. its united ventral fins, while the blenny has If consideration be given to the systematic developed marked ability to hop about in position of the fishes found in these tide pools, water or on land. Of the "floating" fish only TABLE 3 SYSTEMATIC ARRANGEMENT OF TIDE-POOL REPRESENTATIVES Genera Order and Family Accidental Casual Typical Isospondyli Clupediae Jenkinsia - Percesoces Mugilidae Mugil - Percoidea Gerridae Eucinostomus _ - Sciaenidae Eques - Squamipinnes Chaetodontidae Pomacanthus Acanthuridae Acanthurrus Chromides Pomacentridae Eupomacentrus, 2 Abudefduf, 2 Pharyngognathi Labridae Thalassoma Gobioidea Gobiidae Bathygobiss Jugulares Blennidae Salarichthys

it is at once apparent that only the more ad- one family, the Pomacentridae, with four vanced acanthopterygians have representa- -species represented, has reached a point tives in the categories listed as casual and where it is possible to consider its representa- .typical. This point, which is probably more tives as other than accidental. The one genus, than accidental, is brought out clearly in Eupomacentrus, has taken to living about and table 3 in which the orders and families are within empty shells and evidently always listed at the left in the usual systematic order. nests in them (see, for example, Brinley, 1939.; Here it becomes at once apparent that six of Coonfield, 1940; and Longley and Hildebrand, the seven genera listed as accidental have 1941). The other, Abudefduf, nests in more 'been considered by authors as, at least, not exposed places. This is known for A. saxatilis, more advanced than those four listed as Longley and Hildebrand (1941), and in the .casual and typical. Although each may be absence of knowledge to the contrary may be specialized in its own particular direction, assumed -for A. canalogus. As a result of this -only the very advanced Gobioidea and Jugu- manner of activity, these fishes have consid- lares, in this region at least, have been able erable to do directly with the solids in their to establish themselves sufficiently in tide- environment. On the other hand, all the spe- -pool environments to be considered as.typi- cies listed as accidentals either are known to 1948 BREDER: COLORATION IN MARINE SHORE FISHES .30,7 shed pelagic eggs which receive no parental cident light (see, for example, Sumner, 1939,r attention, or there is good reason to suppose 1940, 1943; and Sumner and Wells, 1933)., that they do. Also it has been shown that there are both The markings of the fishes typical of the direct and optically mediated elements" light, clean, sandy bottorns present an item which are antagonistic in their action 'con.-f of considerable interest in the remarkably trolling the dispersal and production of me-: clear water of this area. The same species in lanin. Thus only blind individuals, of the less transparent water do not show the pe- various species studied, are able to show a culiar disappearing lightness marked only by truly dark phase, in the presence of light a few dark points as seen at this place. The and a light background, as is indicated by prominent forms, already discussed, are Hogben (1942), Odiorne (1937), and Rasquin shown comparatively in figure 4, all reduced (1946). However this may be for the fishes so to a common length. The precise significance studied, it is evident that in the case of Chae- of these various marks, which are specifically todipterus some different form of control must distinct and in one case evidently sexually have the ascendancy, for these individuals, distinctive, is not clear at this writing. This although clearly not blind, perform as though should be, however, amenable to an experi- they were sightless. Furthermore, they do mental approach. While it is evident that this only when over a relatively clean, light there is a general resemblance in all, it is bottom, reverting to what is evidently a more also clear that each has a definite character of usual type of chromatic reaction when-in a its own. Thus, so far as a human in goggles is different environment, such as described for concerned, these five patterns may be recog- the fish which took up its abode under the nized as soon as they emerge from the general laboratory dock and henceforth showed a blue gray haze and long before the fish itself strongly barred pattern. can be distinguished. It would probably be The darkening of the back of Sphyraena pointless to speculate further at this time may have some bearing on this, since it is on the presumed significance of these "hiero- evidently done in the presence of light and glyphics" to their owners or those who view quickly reversed on the coming of nightfall. them. It would seem, however, at the very This dark back clearly helps in its counter- least, that they do tend to increase the dis- shading, as an underwater view will attest, tance perception of these fishes, for, although but what utility the lightening of the back on this water is remarkably clear when viewed a dark night may have, if any, is far from from above, the horizontal view under water evident. in all such places is sharply restricted by a The whole question of the bearing of the bluish gray haze. To assume that these are pineal on this general question, as suggested all simple ruptive markings, a priori, would by Breder and Rasquin (1947) in their seem to be an unwarranted oversimplifica- study of cave fish, leads to a general specula- tion for, as in many such cases, it would not tion on these fishes. In the small fishes such as be surprising to find that each fish is solving Athericna and Eucinostomus, it might be im- some special problem by this means under the agined that some light penetrates to that area, restrictions of its particular environment. which in the terms of the cave fish studies Bearing on the above concepts is the rela- should tend to keep them in open water and tion between the behavior and pigmentation away from shelters. This they do, and, con- of Chaetodipterus, in which a fish instead of trariwise, the relatively heavily pigmented matching the background and thus "dis- pomacentrids of the genus Eupomacentrus appearing" does just exactly the reverse tend to hover in and about the shelter of and becomes "lost" amid other bits of shells, while Abudefduf, less heavily pig- truly inert trash which it resembles to a mented, would seem to be a stage away from greater or lesser degree of perfection. All such a habit. In both Sphyraena and Chaeto- background-matching fishes studied in suf- diPterus, the heavy pigment and thick tissues ficient detail have been shown to vary must protect the pineal to a considerable de- their pigmentation according to the albedo gree, but neither gerius is in the -least luci- and not according to the amount of in- fugous. The facts that Astrapogon is heavily 308 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 92 pigmented and hovers closely to its protect- mental and morphologic study of the con- ing mollusk, presumably prowling about at ditions in any or all of these should be il- night for food, are also suggestive, but that luminating and should form a firm step to- Eupomacentrus retreats completely hidden wards a better understanding of the vagaries within an empty shell at night suggests still of behavior displayed by these fishes. some other mechanism of control. An experi- SUMMARY 1. REGULAR INHABITANTS of Bahama tide while the latter spend the entire night hidden pools maintain their residence forlong periods, in such places. All the rest spend the night extending at least between drying on succes- resting in open places, with the exception of sive spring tides. Thalassoma which evidently spends the night 2. Peck-order hierarchies of some stability under cover. are established in these tide pools which are 8. Both Eucinostomus gula and young interspecif!c and evidently depend to a con- Sphyraena barracuda inhabit open, shallow, siderable extent on size and darkness of color- sandy beaches and show a bottom-matching ation. mottling over mottled bottom, but become 3. The contents of such tide pools may be plain and also bottom matching over clean divided into typical species which are domi- sand with the exception of certain fin tips nant and show specializations associated which become intensely black. with tide-pool life, casual species which are 9. The fact that the black pupil of the eye less abundant and show no obvious tide-pool cannot be faded suggests that the fin tip is specialization but are well accommodated to in some way connected with this fact in refer- them, and accidental species which are rare ence to recognition or confusion. and evidently not suited to such places in 10. Other fishes of these same clean beach which they are apparently occasionally areas that show similar black-and-white trapped. patterns, involv'ing only the fin tips, include, 4. Species in the area studied which are besides the above, adult Mugil trichodon and here considered as typical include Bathygo- young Trachinotus palometa. bius soporator and Salarichthys textilis; those 11. Both young and adult of Chaetodipterus considered as casual include the young of faber, in the very clear water of this region (in Eupomacentrus leucostictus, E. adustus, contrast to their chromatic behavior in places Abudefduf analogus, and A. saxatilis; while of greater turbidity), when on clear stretches those considered as accidental include Jen- of sand show their blackest phase and recline kinsia lamprotaenia, the young of Mugil on one side, resembling a piece of inert trash, trichodon, Eucinostomus gula, Eques pulcher, but when in the area of a dock right them- Pomacanthus aureus, Acanthurus hepatus, selves and hide among the piles with their and Thalassoma bifasciatum. boldest pattern of black-and-white vertical 5. In addition to respiratory specializa- bars. 12. Since the more usual background- tions and the ability to leave a given pool it and move overland, the typical species are matching behavior of other fishes makes only for blind fish to show their able to and do match the bottom on which possible back- they rest to a remarkable degree, while the darkest phase when in light on a light casuals in no case show such adjustments, ground, it follows that the visual-hormonal their bright colors making them conspicuous control of melanophores in Chaetodipterus must operate in a different manner. to a notable extent. in 6. The typical species are strictly carniv- 13. The fish AstraPogan, inquiline some slight extent on the Strombus gigas, shows an approximately equal orous and prey to of this mollusk irrespective of the casuals, but the latter as well as the acciden- infestation inside on such concentration of the latter, but only in tals are all non-predatory, subsisting and it is evidently absent vegetable and animal matter as they may sheltered waters, Strombus at on the sub- from the relatively unsheltered obtain by picking growths About one Strombus in 12 was found strate or grubbing in the sand. samba. in sheltered 7. The typical tide-pool species, and Eupo- to be inhabited by an Astrapogon macentrus among the casuals, spend much placesw as 14. The spawning season of Astrapogon time going in and out of cavities such in this region. empty shells, the former both day and night begins in August 309 LITERATURE CITED

BEEBE, W. 517, Papers Tortugas Lab., vol. 32, no. 1931. Notes on the gill-finned goby, Bathy- 12, pp. 169-178. gobius soporator (Cuvier and Valen- CoTT, H. B. ciennes). Zoologica, New York, vol. 12, 1940. Adaptive coloration in animals. London, no. 5, pp. 55-66. Methuen and Co., Ltd., xxxii+508 pp. BEE:BE, W., AND G. HOLLISTER FOWLER, H. W. 1933. New species of fish from the West In- 1947 Notes on Bahama fishes obtained by dies. Zoologica, New York, vol. 12, no. Mr. Charles G. Chaplin in 1947 with 9, pp. 83-88. descriptions of two new species. Notulae BEEBE, W., AND J. TEE-VAN Naturae, no. 199, pp. 1-14. 1928. The fishes of Port-au-Prince Bay, Haiti. GINSBURG, I. Zoologica, New York, vol. 10, no. 1, 1947. Amenrcan species and subspecies of pp. 1-279. Bathygobius, with a demonstration of a BREDER, C. M., JR. suggested modified system of nomen- 1925. Scientific results of the first oceano- clature. Jour. Washington Acad. Sci., graphic expedition of the "Pawnee" vol. 37, no. 8, pp. 275-284. 1925. Fishes. Bull. Bingham Oceanogr. GUDGER, E. W. Coll., vol. 1, no. 1, pp. 1-90. 1918. Sphyraena barracuda; its morphology, 1932. On the habits and development of cer- habits, and history. Carnegie Inst. tain Atlantic Synentognathi. Carnegie Washington Publ., no. 252, Papers Inst. Washington Publ., no. 435, Papers Dept. Marine Biol., vol. 12, pp. 53-108. Tortugas Lab., vol. 28, no. 1, pp. 1-35. 1927. the 1933. New locality records of Abudefduf ana- Inquilinism between cheilodipterid logus including one from the United fish, Apogonichthys puncticulatus, and States. Copeia, no. 4, pp. 191-193. the univalve mollusk, Strombus bituber- 1938. Observations at the Marine Studios. culatus. Zoologica, New York, vol. 9, Bull. New York Zool. Soc., vol. 41, no. no. 2, pp. 193-200. 4, pp. 123-129. 1929. On the morphology, coloration and be- 1943. The eggs of Bathygobius soporator (Cu- havior of seventy teleostean fishes of vier and Valenciennes) with a discussion Tortugas, Florida. Carnegie Inst. Wash- of other nonspherical teleost eggs. Bull. ington Pubi., no. 391, Papers Tortugas Bingham Oceanogr. Coll., vol. 8, no. 3, Lab., vol. 26, pp. 149-204. pp. 1-49. HILDEBRAND, S. F., AND L. E. CABLE 1946. An analysis of the deceptive resem- 1938. Further notes on the development and blances of fishes to plant parts, with crit- life history of some teleosts at Beaufort, ical remarks on protective coloration, N. C. Bull. U. S. Bur. Fisheries, vol. 48, mimicry and adaptation. Ibid., vol. 10, pp. 505-642. no. 2, pp. 1-49. HILDEBRAND, S. F., AND I. GINSBURG BREDER, C. M., JR., AND F. HALPERN 1926. Descriptions of two new species of fishes 1946. Innate and acquired behavior affecting from Key West, Fla., with notes on nine the aggregation of fishes. Physiol. Zool., other species collected in the same local- vol. 19, no. 2, pp. 154-190. ity. Bull. U. S. Bur. Fisheries, vol. 42, BREDER, C. M., JR., AND P. RASQUIN pp. 207-215. 1947. Comparative studies in the light sensi- HOGBEN, L. tivity of blind characins from a series of 1942. Chromatic behavior. Proc. Roy. Soc. Mexican caves. Bull. Amer. Mus. Nat. London, ser. B, vol. 131, pp. 111-136. Hist., vol. 89, no. 5, pp. 319-352. JORDAN, D. S., AND W. B. EVERMANN BRINLEY, F. J. 1896. The fishes of North and Middle Amer- 1939. Spawning habits and development of ica. Bull. U. S. Natl. Mus., no. 47, pp. beau-gregory (Pomacentrus leucostiaus). i-lx, 1-3313. Copeia, no. 4, pp. 185-188. LONGLEY, W. H., AND S. F. HILDEBRAND COONFIELD, B. R. 1941. Systematic catalogue of the fishes of 1940. Chromatophore reactions of embryos Tortugas, Florida, with observations on and larvae of Pomacentrus leucostictus. color, habits, and local distribution. Carnegie Inst. Washington Publ., no. Carnegie Inst. Washington Publ., no. 310 1948 BREDER: COLORATION IN MARINE SHORE FISHES 311

535, Papers Tortugas Lab., vol. 34, pp. 1940. Further experiments on the relations i-xiii, 1-331. between optic stimuli and the increase ODIORNE, J. M. or decrease of pigment in fishes. Jour. 1937. Morphological color changes in fishes. Exp. Zool., vol. 83, no. 2, pp. 327-343. Jour. Exp. Zool., vol. 76, no. 3, pp. 1943. A further report upon the effects of 441-465. visual environment on the melanin con- PLATE, L. tent of fishes. Biol. Bull., vol. 84, no. 2, 1908. Apogoniclthys strombi n. sp., ein sym- pp. 195-205. biotisch lebender Fisch von den Ba- SUMNER, F. B., AND N. A. WmELs hamas. Zool. Anz., vol. 33, no. 12, pp. 1933. The effects of optic stimuli upon the for- 393-399. mation and destruction of melanin pig- RASQUIN, P. ment in fishes. Jour. Exp. Zool., vol. 64, 1946. On the reappearance of melanophores no. 3, pp. 377-394. in blind goldfish. Copeia, no. 2, pp. 85- WALLS, G. 91. 1942. The vertebrate eye. Bull. Cranbrook SUMNER, F. B. Inst. Sci., no. 19, pp. i-xiv, 1-785. 1939. Quantitative effects of visual stimuli WOOD-JoNEs, F. upon pigmentation. Amer. Nat., vol. 73, 1912. Corals and atolls. London, Lovell Reeve no. 746, pp. 219-234. and Co., xxiii + 329 pp.