FAU Institutional Repository

http://purl.fcla.edu/fau/fauir

This paper was submitted by the faculty of FAU’s Harbor Branch Oceanographic Institute.

Notice: ©1980 The Ichthyological Society of Japan. This article may be cited as: Mok, H.-K. (1980). Notes on the classification of Actinopterygian intestinal patterns. Japanese Journal of Ichthyology, 27(1), 29-40.

Japanese Journal of Ichthyology Vol. V. No. I 1980

Notes on the Classification of Actinopterygian Intestinal Patterns

Hin-Kiu Mok (Received February 19, 1979)

Abstract A classification of actinopterygian intestinal patterns represented by 596 species is presented. Individual and developmental variations, distribution, and interrelationships of intestinal patterns were studied. The low conspecific variation, non-random pattern vari­ ation in groups with complex patterns, and resemblance of intestinal patterns among mem­ bers of monophyletic groups suggest that intestinal pattern provides valuable information regarding phylogenetic interrelationships.

Suyehiro's (1942) study of a variety of tele­ patterns which can be significant to phylo­ ostean digestive systems is the only system­ genetic studies. After the generalized acti­ atic survey of the digestive system within this nopterygian patterns were recognized, patterns group. His objective was to understand fish that differ from these generalized types may feeding habits by studying the digestive organs. be treated as apomorphic or derived char­ He discovered various types of intestinal pat­ acters at various taxonomic levels (Hennig, terns (within the peritoneal cavity the positions 1966). of the various gastrointestinal segments form the fish intestinal pattern; the elements of the Materials and methods pattern are the relative positions of the seg­ Specimens of 596 actinopterygian species ments). In contrast to this ecological-anatom­ examined were on loan from the American ical study, Harder (1960) used superficial gut Museum of Natural History, Academy of morphology (the stomach and pneumatic duct) Natural Sciences of Philadelphia, Australian as a basis for interpreting the phylogeny of Museum in Sidney, Bernice P. Bishop Museum, the clupeoids. Greenwood (1968) described California Academy of Sciences, South Austra­ the gut and other viscera in Denticeps clupeoides lian Museum, Scripps Institution of Oceanog­ in an attempt to resolve its systematic posi­ raphy, Institute of Animal Resource Ecology, tion among clupeiforms. Nelson (1972) found University of British Columbia, Western Aus­ evidence for monophyly of osteoglossomorphs tralian Museum, Zoology Museum, National (including hiodontids) from the intestinal Taiwan University, and Zoology Museum, pattern. A similar systematic survey of the University of Michigan. Data from examined digestive system, swimbladder, and related specimens are not listed herein, but are tissues (e.g., the gas-gland) of tetraodontiforms available in Mok ([978). was conducted by Mok (1975). Results of Dissections were made on one or more in­ that study suggested that the intestinal pattern dividuals of each species studied. The fish might be applicable to the reconstruction of were cut on the right side of the body cavity. phylogenetic relationships among other fishes. Intestinal pattern was recorded by tracing the Accordingly, the major objectives of the pres­ coiling pathway with a continuous line. ent study are 1) to classify the types of intestinal patterns that generally appear in Results actinopterygians, 2) to survey the distribution Loop f and loop a of these types, and 3) to understand individual In most actinopterygians, the anterior part and developmental variations of intestinal of the intestine proceeds anteriorly and then patterns. This study will provide baseline displaces to the right of the stomach, which information about actinopterygian intestinal is, in most cases, either U-, V-, or T-shaped

- 29 Japan. J. Ichthyo!. 27(1), 1980

A TYPE B B TYPE 01 c TYPE 02

'~ 1J~~~ ~c ('),-~\.~\'- '" '-----~:~.:.:~ ~:'~'."....;:.:. ~.,. ~.

D TYPE SP E TYPE S F TYPE Z

G TYPE HZ H TYPE LA TYPE T TYPE:~ J K

Fig. I. Left side-view of eleven basic types of actinopterygian intestinal patterns. A: Type B. B: Type Dl. C: Type D2. D: Type SP. E: Type S. F: Type Z. G: Type HZ. H: Type LA. I: Type T. J: Type LF. K: Type ST. i, intestine; r, rectum; s, stomach. Arrows, coiling (or folding) directions of loop a; fine stippled areas, loop f; heavy stippled areas, loop a; heavy lines, specific sections of the patterns (see text for explanation);-(-, junction of stomach and intestine.

(Suyehiro, 1942; Greenwood, 1968; Nelson, of the intestine between its beginning and 1972). The loop formed by the anterior the tip of the loop, and the length from the section of the intestine at the front of the tip of the loop to the anus, are about the peritoneal cavity I designate loop f (first loop; same. With this mid-intestinal position as a Fig. 1, fine stippling). By definition, this loop basic criterion, mistakes in recognizing this can be found in species in which the stomach loop will be few. Sometimes, however, the is not straight or l-shaped. The intestine in intestine is long and the pattern too complex the rear of the peritoneal cavity commonly to recognize loop a. In such cases, com­ forms another loop, which I designate loop a parisons may be limited to the overall winding (Fig. I, heavy stippling). Loop a may be tendencies of the intestine, not the exact recognized in most cases, because the length numbers and shapes of loops.

- 30- Mok : Intestinal Patterns

A Fig. 2. Left side-view of the intestinal pattern of Mene annocarolina (A) and Lumpenus lumpretaeformis (B). Arrows, the downward bending of the anterior section of the intestine (heavy line); stippling, loop a; -(-, junction of stomach and intestine.

A B

c o

E F Fig. 3. Left side-view of the intestinal patterns of six atherinoids. A: Allanetta harringtonensis. B: Atherinops affinis. C: Membras vagrans. D: Menidia menidia. E: Chilatherina sp. F: Melanotaenia nigrans. Fine stippling, loop f; heavy stippling, loop a; heavy line, esophagus.

Types of intestinal patterns Type B: Loop f is to the right of the esoph­ Actinopterygian patterns can be classified agus, and loop a extends horizontally to the into eleven basic types with reference to the posterior part of the peritoneal cavity (Fig. variations of loop f and loop a. They are IA). This is the basic and most widely dis­ types B, Dl, D2, SP, S, Z, HZ, LA, T, LF, tributed pattern in acanthopterygians, although and ST (Fig. I A~K). it occurs also in some pre-acanthopterygian

- 31 Japan. J. Ichthyol. 27(1), 1980 d7

?Pr- '------A B s . i:? ... c - Fig. 4.. Left side-view of the intestinal pattern of A: Bostockia porosa. B: Astrapogon stellatus. C: Cheilodipterus macrodon. D: Glossamia aprion. Stippling, loop a; -(-, junction of stomach and intestine. teleosts. Nelson (1972) reported the same mach. I have not yet examined the patterns pattern in primitive actinopterygians, Acipenser of Isonidae, Neostethidae, and Phallosteth­ oxyrhynehus (Acipenseridae) and Lepisosteus idae. A survey of these groups would be platyrhineus. Among the perciforms I have helpful in understanding the generalized (or studied, the vast majority (more than 80%,) primitive) pattern of this suborder. have this pattern (Mok, 1978). It is, there­ Loop a of Type B generally does not bend. fore, a generalized or primitive character Exceptions occur in loop a of Bostockia porosa state for acanthopterygians as a whole. (Serranidae), Glossamia aprion, Cheilodipterus There are some perciforms with patterns macrodon, Astrapogon stellatus (Apogonidae), that seem to be variants of Type B. In Mene and Doratonotus megalepis (Labridae) which annoearolina (Menidae) and Lumpenus lumpre­ distinctively tilt upward at a very steep angle taeformis (Stichaeidae) (Fig. 2A and B, re­ (Fig. 4). spectively) the patterns are similar in that Type Dl: Loop a winds dextrally to the the anterior section of the intestine bends right of the rectum (Fig. IB). Type ni has downward (arrows in these illustrations show a broad distribution among perciforms. Some the bending direction of that section of the examples are Scomberoides toloo, Vomer de­ intestine). clivifrons (Carangldae), Cirrhitichthys falco In osteoglossiforms and some atherinoids (Cirrhitidae), Kuhlia taeniura, K. malo (but loop f occurs left of the esophagus and stom­ not in K. marginata and Nantherina balstoni, ach (Suyehiro, 1942; Nelson, 1972). Nelson Kuhliidae), Gnathodentex aurolineatus, Mono­ considered the left position of loop f as an taxis grandoeulis (Lethrinidae), Monodaetylus apornorphic character of osteoglossomorphs [alciforrnis (Monodactylidae), Oplegnathus sp. (including hiodontids). Interestingly, loop (Oplegnathidae), Pseudopriaeanthus atlus f of atherinoids, Allanetta harringtonensis, (Priacanthidae), Dampieria eyclophthalma, Atherinops aJlinis, Membras vagrans, and Gramma sp. (Pseudochromidae), Stellifer ras­ Menidia menida (Atherinidae), Chilatherina sp., trifer (Sciaenidae), Amniataba percoides (Ter­ and Melanotaenia nigrans (Melanotaeniidae) aponidae), and Triehodon triehodon (Trich­ (Fig. 3) all occur left of the esophagus and sto- odontidae). This type does not characterize a

- 32- Mok : Intestinal Patterns c~ ~.~~ A B

c D Fig. 5. Left side-view of the intestinal pattern of A: Archosargus probatocephalus. B: Bryostemma nigator. C: Holocentrus rufus. D: Hexagrammos decagramma. Stippling, loop a; -(-, junction of stomach and intestine. genus or family as a whole, but might be Examples include most acanthurids (Mok, applied to the interpretation of species inter­ 1977), anabantoids (e.g., Colisa lalia, Macro­ relationships. podus opercularis, M. viridiauratus, Trichogas­ Type 01 is uncommon in non-perciform ter trichopterus, Helostoma temmincki; Osphro: actinopterygians. Among those examined, I nemus goramy), siganids (e.g., Lo vulpinus, have seen this type in Myoxocephalus quadri­ Siganus rostra/us), scatophagids (Scatophagus cornis (Cottidae), Limanda sp. (Pleuronectidae), argus), chaetodontids, and stromateids (e.g., and Achirus linea/us (Soleidae) (Mok, 1978). Pampus argenteus). Among non-perciform Ochiai (1966) found Type 01 in some Japanese actinopterygians, it has only been found in a soles Liachirus melanospilus and Pardachirus few ostariophysans (e.g., Hypostomus sp. and pavoninus, Loricaria cataphractay and cyprinodontoids Type D2: Loop a winds dextrally to the (e.g., Goodea sp.). left of the rectum and may extend anteriorly The exact patterns of the aforementioned to the left of the stomach (Fig. lC). Type group differ to various degrees. The intestinal 02 is limited to Chironemus marmoratus patterns of anabantoids, siganids, and scato­ (Chironemidae), Dentex [aponicus (Sparidae), phagids reveal a greater similarity between and Nomeus gronovii (Nomeidae) among the themselves than to the other groups with the perciforms examined. Type SP. Loop a of these three groups is In two Perciforms, Archosargus probato­ the only major loop and is coiled strictly in cephalus (Sparidae) and Bryostemma nigator a spiral. The patterns of aeanthurids (Mok, (Stichaeidae), and also in some non-perciform 1977), chaetodontids, and stromateids diverge teleosts such as Holocentrus rufus (Holocen­ from this basic pattern (Mok, 1978). tridae) and hexagrammines (Hexagrammos spp. Type S: The middle part of loop a is and Agrammus agrammus) loop a runs under­ slightly depressed, and loop a is to the right neath the rectum and folds upward to the of the rectum (Fig. IE). This type occurs left of the rectum (Fig. 5). These character only in adult Dipterygonotus gruveli (Emmel­ states of the position of the tip of loop a ichthyidae), Premnas biaculeatus (Pomacen­ are apomorphic. tridae), Psenopsis anomala (Centrolophidae) Type SP: The long loop a winds dextrally and juvenile Amphiprion ocellaris (Pomacen­ to the right of the rectum, forming a spiral tridae). intestinal mass (Fig. 10). Type SP has a Type Z: The winding of both sides (or limited distribution within the Perciformes. the anterior and posterior sections; heavy

- 33- Japan. J. Ichthyol. 27(1), 1980

A B

c

E F Fig. 6. Developmental change of the intestinal pattern of Acanthurus bahianus. A: 30 mm. B: 32 mm. C: 50 mm. D: 59 mm. E: 75 mm. F: 100 mm. Fine stippling, loop b; heavy stippling, loop a; -(-, junction of stomach and in­ testine. lines in Fig. IF) of loop a follows a Z-shaped (Leiognathidae), Abudefduf saxatilis, and pathway (Fig. IF; arrow indicates the winding young A. taurus (Pomacentridae). Other actino­ route) to the right of the rectum. It occurs pterygians with this pattern are monacan­ widely in perciforms such as Crinodus lophodon thids (Mok, 1975). (Aplodactylidae), kyphosids, most pomacen­ Type LA: Loop f occurs, and loop a is trids (also see Aoyagi, 1941), Agonostomus absent (Fig. IH). This type occurs in actino­ monticola (Mugilidae), and Psenes cyanophrys pterygians with a short intestine having only (Nomenidae). Loop a of Cichlasoma hetero­ one loop. It is not limited to specific gr,!ups, spilum, C. spilurum, Haplochromis eucinostomus, but occurs in various actinopterygian families. and Melanochromis johannii (Cichlidae) also Type T: The portion of the intestine cor­ winds in a Z-shaped pathway. However, the responding to the convex section of loop f homolog of loop f cannot be recognized in in Type B and Type LA (Fig. IA, H; heavy the examined cichlids [Mok, 1978). lines) is concave (Fig II; heavy line) and Type HZ: Loop a turns at the rear of the loop a is absent. Brachyistius frenatus (Em­ peritoneal cavity and extends anteriorly such biotocidae) is the only examined perciform that the tip is located between the anterior with this pattern. However, Type T con­ and posterior section of loop a (Fig. IG; sistently characterizes the patterns of some heavy lines represent these sections). This cyprinodontoids, such as Profundulus punctatus type can be distinguished from Type Z in (Cyprinodontidae), Oryzias latifer (Oryziat­ that the posterior section of loop a does not idae), Belonesox belizanus, Gambusia sp_ wind in a Z-shaped manner. (Poecilidae). This pattern has not been found There are only a few examples of Type HZ in other actinopterygians. found in perciforms: Leiognathus rivulatus Type LF: Loop f is absent (Fig. lJ).

- 34- Mok ; Intestinal Patterns

. ~~~' # (. ~ ~ ... :.:~ E Fig. 7. Developmental change of the intestinal pattern of Nomeus gronovii. A: 19mm. B: 42 mm. C: 54 mm. D: 60 mrn. E: 62 mm. F: 70 mm. Large arrows, dextral winding tendency of loop a; small arrows, leftward folding of the anterior intestinal section; stippled areas, loop a.

Stomach of the species with this type is Developmental variation straight. The anterior section of the intestine To determine possible ontogenetic change proceeds to the rear of the peritoneal cavity in intestinal patterns specimens of various where it turns and runs forward. Type LF sizes were studied in Acanthurus bahianus (10 occurs at least in some species of the follow­ specimens, 30~ 100 mm SL), Lepomis gibbosus ing perciform families: Carangidae, Cichlidae, (36 specimens, 24~70 mm SL), Nomeus gronovii Ephippidae, Serranidae, Gobioididae, Clin­ (6 specimens, 19~ 70 mm SL). The patterns idae, Chaenopsidae, Gobiidae, Pholididae, of A. bahianus and N. gronovii are shown in and Ammodytidae (Mok, 1978). In non­ Figs. 6 and 7, respectively. The patterns of perciform actinopterygians, Type LF occurs the individuals of L. gibbosus are similar. in some species of the families Cyprinidae, Little ontogenetic change was seen in the Gasterosteidae, Macrorhamphosidae, and Tri­ acanthurids examined. All specimens of A. glidae. Loop f is absent also from all tetra­ bahianus examined are similar in having two odontids, diodontids, molids (Mok, 1975) and loops-loop a and loop b; the latter is unique triodontids (Breder and Clark, 1947). Type to acanthurids and is located left of the LF is most widely distributed among gobioids stomach (Mok, 1977). The pattern develops and blennioids of the families Chanopsidae, early (prior to 30 mm SL), and no further Clinidae, and Eleotridae. change takes place beyond this stage. Type ST: The simplest pattern is the Unlike A. bahianus, N. gronovii manifests straight gut (Fig. lK). In actinopterygians, developmental change in specimens between it occurs in some species of the families 19~ 70 mm. Loop a can be recognized in all Gobiesocidae, Belonidae, Exocoetidae, Aulo­ specimens; a dextral winding tendency of stomidae, Syngnathidae, Symbranchidae, Ech­ loop a is found in these developmental stages eneidae, Odacidae, Kraemeridae, Microdesm­ (Fig. 7A~F; heavy stippling indicates loop a; idae, and Triacanthodidae (Mok, 1975, 1978). large arrows show the dextral developing tendency of loop a). A leftward folding

- 35- Japan. J. Ichthyol. 27(1), 1980 appears at a standard length of 54 mm (Fig. 60 mm or longer in standard length. 7C; small leftward pointing arrow). This On the basis of these observations, it seems folding tends to increase in depth during that patterns typical of adults develop early, development (Fig. 7D~F). The pattern typi­ and can be observed in early and late adult cal of this species is formed in specimens of stages as was shown also for pomacentrids

Table I. Distribution of the types of actinopterygian intestinal patterns. 1) also see Kafuku, 1958; 2) also see Matsubara, 1943; 3) Mok, 1975; * dominant type unknown due to the small sample size of the examined species; ** dominant type undistinguishable due to the equal occurring frequencies of the types; + presence of an intestinal pattern; * dominant type of the group concerned.

Intestinal type Taxon B Dl SP S Z LA HZ T LF ST D2 Others Chondrostei* + + Ginglymodi* + Halecomorphi Osteoglossomorpha *+ 1+ Elopomorpha + + Clupeomorpha + * + Ostariophysiu + *+ 1+ Protacanthopterygii + 1+ Paracanthopterygii** + + + Atherinomorpha** + + + + + Acanthopterygii Beryciformes* + + Zeiformes* -I- Lampridiformes* + Gasterosteiformes** --I- --I- Ghanniformes 1+ Synbranchiformes Scorpaeniformes" 1+ + + * Perciformes 1+ + + + + + + + + + + Percoidei 1+ + + + r- + + + Mugiloidei + Sphyraenoidei * Polynemoidei* + * Labroidei + + Trachinoidei 1+* + + + Notothenoidei 1+ Blennioidei** + + + +- --I- Icosteoidei 1+ Ammodytoidei 1+ Callionymoidei 1+ Gobioidei + 1+ -l- Kurtoidei** + + Acanthuroidei 1+ -f- Scornbroidei + 1+ Stromateoidei --I- + --I- Anabantoidei** + + * Luciocephaloidei 1+ Mastacernbeloidei 1+ Pleuronectiformes 1+ + 1+ + Tetraodonriformesw>' + + + + +

- 36- Mok : Intestinal Patterns ~ SP~~ ~r

C /~

. ~----- ~,. . B ...... S~~

Fig. 8. A hypothetical scheme showing the possible interrelationships among the eleven types of intestinal patterns generally found in actinopterygians.

(Fukusho, 1969). specimens that were examined. However, the Distribution and interrelationships of the ac­ developmental changes of the intestinal pat­ tinopterygian intestinal patterns terns of pomacentrids Abudefduf taurus and The distribution of the types of patterns in Amphiprion ocellaris suggest that Type Z can actinopterygians is summarized in Table 1. be develped from Type HZ or Type S. Smaller Type LA dominates the patterns of lower individuals of A. taurus (55 moo) and A. ocel­ actinopterygians, such as elopomorphs, clupeo­ laris (17 moo) have types HZ and S, respective­ morphs, protacanthopterygians. Type B be­ ly; whereas larger individuals of these species comes the dominant one in some acanthopter­ (135 and 42 moo, respectively) develop the ygians including channiforms, scorpaeniforms, Type Z pattern. A small individual of Micro­ perciforms, and pleuronectiforms (also see canthus strigatus (87 mm; Kyphosidae) has Type Ochiai, 1966). Several of the patterns are Dl, whereas a larger individual (122mm) has unusual in perciforms: types S, Z, D2, HZ, Type SP (Mok, 1978). This developmental and SP. Type HZ, which is limited to a change suggests the close relationships of few perciforms, becomes the dominant type Types DI and SP. The aforementioned ob­ in monacanthids and balistids, Its synapo­ servations provide direct indication of the morphic nature in Tetraodontiforrnes is un­ close relationship between these intestinal questionable. patterns (Fig. 8). Ontogenetic change in intestinal patterns In 48 cases of association between two types and frequency of particular intestinal pattern of patterns found in 31 monophyletic groups associated with members of monophyletic or (Acipenseridae, Clupeidae, Cobitidae, Cyprin­ hypothetical monophyletic groups should help odontidae, Gasterosteiforrnes, Scorpaeniforrnes, elucidate intestinal pattern interrelationships. Cottidae, Carangidae, Centrarchidae, Cirrhit­ Information on ontogenetic change is limited idae, Echeneidae Embiotocidae, Kuhliidae, in this study because specimens of different Leiognathidae, Nemipteridae, Percidae, Poma­ sizes were not usually available and no de­ centridae, Pseudochromidae, Sciaenidae, Spar­ velopmental change was observed in those idae, Teraponidae, Labridae, Odacidae, Pho-

- 37- Japan. J. Ichthyol. 27(I), 1980

lididae, Gobiidae, Kurtidae, Scombridae, little information to be of any phylogenetic Belontiidae, Pleuronectidae, and Soleidae), significance, (4) the long intestine of herbi­ the frequencies of 15 kinds of association (or vores may associate with high variation in frequencies of co-occurrence) of the types the pattern to such a degree that phylogenetic were obtained and they are listed in decreasing significance ofthe intestinal pattern diminishes. order as follows: B~LA, 17; B-DI, 9; B-ST, However, the results of this study indicate 3; DI-LA, 3; B-S, 2; LA-ST, 2; B-D2, I; B­ that these drawbacks have minimal "nega­ HZ, I; B-Z, I; B-SP, I; B-T, 1; S-Z, 1; DI­ tive" effect on the phylogenetic significance S, I; LA-T, I; LF-ST, I (Mok, 1978). Chi­ of intestinal pattern (also see below). square test of these data (x 2= 27.8796, df= Intraspecific variation in intestinal pattern 10, P:s;0.05) leads to the conclusion that the was considered to be insignificant (Fukusho, appearance of the types of intestinal patterns 1969: Mok, 1977, 1978). Observations made in a monophyletic group is not an independent on a limited number of individuals will, event. This conclusion suggests that certain therefore, give a reliable description of the types tend to associate. Only 15 kinds of species-specific pattern. the 55 possible types of association between Ecological factors such as feeding habits two of the eleven types appear in this sample. do not directly and significantly determine the The extraordinarily high frequency of associ­ intestinal pattern. To some extent, feeding ation between types B-LA and B-DI suggests habits may correlate with intestine length, a that Type B is closely related to types LA determinant factor of intestinal pattern. and DI (Fig. 8). Grounded on the occur­ Fishes with a similar food habit may be rence of their association within monophyletic different in their intestine length, but unique groups and also on ontogenetic evidence, it characteristic patterns may still be retained. is hypothesized that a close relationship exists Similarly, related species with different feeding between types B, HZ, S, and Z (Fig. 8). habits were also found to share a unique .Type B is also related to types D2, T, and intestinal pattern (Mok, 1977). ST. The multiple association of Type B to The short intestine of carnivores and pis­ the other types suggests that it is the most gen­ civores shows a simple pattern and generally eralized type, at least for acanthopterygians, contains less information but does not neces­ which changes into other intestinal patterns sarily reduce the phylogenetic significance of (Fig. 8). the intestinal pattern. The unique pattern of The problem with other simple patterns, piscivorous Hexagrammos and Agrammus types T, LF, and ST, are more complicated (Hexagramminae; see above) is an adequate because they may be independently evolved example. from various types by significant or mild Intestinal pattern is determined by intestine reduction of intestinal length. But on the length, volume and shape of the peritoneal basis of the present data, Type T is related cavity and developmental mechanism. The to types B and LA; Type ST is related to pattern complexity is dual in nature, both types B, LA, and LF (Fig. 8). No associa­ quantitative and directional. As such, it can tion between types Band LF was found. be measured by the numbers of loops and/or by the directional changes that a loop displays. Discussion For instance, the more loops a pattern is At the beginning of this study, I was be­ composed of, the more complex it is. Al­ wared of the possible drawbacks of intestinal ternatively, if loop a is the only loop, a pat­ pattern as an indicator of phylogeny: (I) tern with a dextral-winding loop a is more high intraspecific variation, (2) as a conse­ complex than a pattern with a straight loop quence of ecological adaptation, similarity in a. With equal volume of the peritoneal intestinal patterns may only indicate conver­ cavity, a species with a long intestine is gence or parallelism rather than possible likely to have a pattern more complex than monophyly, (3) the simple patterns of carni­ a species with a short intestine. For example, vores (due to their short intestine) carry too chaetodontids have an oval, small peritoneal

- 38- Mok : Intestinal Patterns cavity and a long intestine which generally acanthids (Mok, 1978). Their sharing of the leads to complex patterns. High complexity specialized character indicate that intestinal does not normally correlate with high in-group pattern reveals congruent evidence. variation in intestinal pattern. Conversely, In summary, the low conspecific variation, intestinal patterns are generally consistent as non-random variation in groups with complex examplified by cyprinoids, mugilids, chaet­ patterns, and resemblance of intestinal pat­ odontids, kyphosids, blenniids, and acan­ terns among members of phylogenetically thurids (Mok, 1977, 1978). It is not true that related groups suggest that intestinal patterns a long intestine is always associated with a com­ provide valuable information for the under­ plex pattern, because the size and shape of the standing of phylogenetic relationships. The peritoneal cavity also correlate with complexity. present classification of the intestinal patterns In most pleuronectiforms, for example, the is an attempt to set up an information system peritoneal cavity ends in front of the first for detailed comparisons. The criteria that I haemal spine, and a dextral-winding loop a adopt for this classification are a few among occurs. But in some soleid species (e.g., Hetero­ others which are applicable to the study of mycteris japonicus and Trinectes maculatus) intestinal patterns. the intestine is long, and the right secondary peritoneal cavity (the posterior extension of AcknowIedgments the cavity on the right side of the haemal This report is a part of the dissertation spines) makes room for the intestine (also see submitted in partial fulfillment of the require­ Ochiai, 1966). Loop a of these species, is ments for my Ph. D. at the City University straight and without the dextral winding of New York. I am grateful to my major tendency. It seems generally true that fishes advisor, Dr. Gareth J. Nelson of the Ameri­ with a long peritoneal cavity tend to have can Museum of Natural History for his guid­ a simple pattern, e.g., Ammodytes personatus ance. The final draft of this paper was (Ammodytidae), Atherina bleekeri(Atherinidae) benefited by the criticisms of Dr. Robert Jones and Trichiurus haumela (Trichiuridae) (Suye­ and Mr. Grant Gilmore of the Harbor Branch hiro, 1942). Foundation. Publication of this paper is Differences in the patterns may sometimes financially supported by Harbor Branch be found in two groups with similar determi­ Foundation. To this institution, I extend nant factors. For instance, soleids have either my thankfulness. sinistral or dextral coiling loop a within the I am grateful to the following curators for right secondary body cavity which is the shared granting specimen loans: Drs. J. E. Bohlke, determinant factor. Conversely, specialized W. F. Smith-Vaniz, Academy of Natural Sci­ patterns have been observed in groups that ences of Philadelphia; J. R. Paxton, Australian differ in certain determinant factors (e.g., shape Museum in Sydney; J. Randall, Bernice Bi­ and volume of the acanthurine and zancline shop Museum; R. Johnson, Field Museum peritoneal cavity). These phenomena suggest of Natural History; J. Glove, South Australian that intestinal pattern is governed by an addi­ Museum; the late C. L. Hubbs, R. Rosenblatt, tional determinant factor, the developmental Scripps Institute of Oceanography, University mechanism, which may enhance the phylo­ of California; N. J. Wilimovsky, Institute of genetic significance of intestinal pattern. Animal Resource Ecology, University of Brit­ Members of numerous groups traditionally ish Columbia; J. B. Hutchins, Western Aust­ considered as monophyletic share specialized ralian Museum; S. C. Shen, Department of intestinal patterns: osteoglossomorphs, Zoology, National Taiwan University; R. M. homalopterids, atherinoids, hexagrammines, Bailey, Museum of Zoology, University of chaetodontids (not including pomacanthids), Michigan; V. Springer, United States National kyphosids (including girellids and scorpids), Museum. Drs. V. Springer (USNM) and mugilids, ophioblennines, salariines (accord­ J. E. Bohlke (ANSP) kindly provided labora­ ing to the taxonomic scheme of Norman, 1957), tory space during my short visits to these acanthurids (including Zane/us) and mon- institutions. Dr. P. Moller, Mr. M. Oliver,

- 39- Japan. J. Ichthyo!. 27(I), 1980 and Me. G. Dingerkus generously lent me formes (Pisces: Acanthopterygii) based on gut specimens in their personal collection. This observations. Unpublished M. A. thesis. The is contribution No. 168 from the Harbor City Univ. of New York, 104 pp. Branch Foundation, Inc. Mok, H. K. 1977. Gut patterns of the Acan­ thuridae and Zanclidae. Japan. J. Ichthyol., 23 Literature cited (4): 215-219, figs. 1-5. Mok, H. K. 1978. Coelomic organs of perciform Aoyagi, H. 1941. The damsel fishes found in the fishes (Teleostei). Ph. D. dissertation. The waters of Japan. Trans. Biogeogr. Soc. Japan, City Univ. of New York, 414 pp. 4(1): 157-279, figs. I-52, pis. 11-23. Nelson, G. J. 1972. Observations on the gut of Breder, C. M., Jr. and E. Clark. 1947. A con­ the Osteoglossomorpha. Copeia, 1972(2): 325­ tribution to the visceral anatomy, development, 329, figs. 1-5. and relationship of the plectognathi. Bull. Norman, J. R. 1957. A draft synopsis of the Amer. Mus. Nat. Hist., 88(8): 287-320, figs. orders, families and genera of recent fishes and 1-8, pis. ·1l-14. fish-like vertebrates. Unpublished photo offset Fukusho, K. 1969. Notes on the intestinal con­ copies distribued by British Museum of Natural volution of the fishes of the Pomacentridae History, 649 pp. (Perciformes). Japan. J. Ichthyol., 16(4): 135­ Ochiai, A. 1966. Studies on the comparative 142, figs. 1-4. morphology and ecology of the Japanese soles. Greenwood, P. H. 1968. Notes on the visceral Spec. Rep. Misaki Mar. Biol. Inst., Kyoto Univ., anatomy of Denticep s clupeoides Clausen, 1959, (3): 1-97, figs. 1-39, pis. 1-2. a west African c1upeomorph fish. Rev. Zool. Suyehiro, Y. 1942. A study of the digestive Bot. Afr., 127(1-2): 1-10, figs. 1-5. system and feeding habits of fish. Japan. J. Harder, W. 1960. Vergleichende Untersuchungen Zoo!., 10: 1-303, figs. 1-190, pis. 1-15. zur Morphologie des Dames bei Clupeoidea. Z. Wiss. Zoo!., 163: 65-167, figs. 1-31. (Harbor Branch Foundation, Inc., RR I, Box Hennig, W. 1966. Phylogenetic systematics. 196A, Fort Pierce, Florida 33450, U.S.A.) Urbana: University of Illinois Press, 263 pp. Kafuku, T. 1958. Speciation in cyprinid fishes on the basis of intestinal differentiation, with some references to that among catostomids, Hin-Kiu Mok Bull. Freshwater Fish. Res. Lab., 8(1): 45-78, figs. 1-20, pis. 1-8. ~!Ui\(7) 596 fiU.::: -o v--c ~ ~ HllH!t, :SHiH.. , 11M 1* Matsubara, K. 1943. Studies on the scorpaenoid .., d.KW5 ••, *~(7)W~. mK~~i." fishes of Japan. Anatomy, phylogeny and tax­ Gk. ~~~~~••tt~g( •••~~it~~~ onomy. Trans. Sigenkagaku Kenkyusho, (I -~~~••K~~tt~~g, **(7)~OO~~MR~ and 2), 1-486, figs. 1-156, pis. 1-4. -cv- 7.>. ;:.tt..t:>(7);:' C ti~M(7)~~iI;*tiR:~~{.:::~v· Mok, H. K. 1975. Phylogeny of Tetraodonti- '( (7)lfAt ts. mm i tlJ'o#'<'t 7.> .: C 2:if- t. r v- 7.>.

-40-