Nb-
-ay) -
«4 EIS Prv„ FISHPRIES RESEARCH BOARD OF CANADA /1/24 , 1
. Translation Series No. 726 c r CliN,104
.',
The distribution.of Cottus laoonicus Okada (Cottidae) and of Tukugobius flumineus 1:Azuno (Gobiidae), with - special reference to peculiarities related to both the land-locking and their Speciation • from amphidromous ancestors
By Nobuhiko,,f•izuno
- • . From: Osaka Gakugei University (Pf-e- S>, pp. l29-161,,,,.19ô5,
Translated by: Kenji Makin°,
7 7-7- - BUreau -for Translations Foreign-Languages- Division, - . Department of the Secretary of State Of Canada
Fisheries Research Board of - Canada Biological -Station, - Nanaimoi- B. - .C. -----
•
I ' . r
fgt.; 44. sPEPARTMENT OF THE SECRETARY OF STATE r SECRÉTARIAT D'ÉTAT BUREAU FOR TRANSLATIONS BUREAU DES TRADUCTIONS
FOiZE-IGN LANGUAGES ; DIVISION DE-S LANGUES DIVISION CANADA ÉT.'2AN&.-"!`2'ES
TRANSLATED rRom TRADUCTION DE IN •ro - Japp nos ri sh
SUI3JECT - SUJET
Fish (lard-lock irg. and spooi.?,t:ion)
AUTHOR - AUTEUR Yobuhiko Mizuno
TITLE IN ENGLISH - TITRE ANGLAIS • I) The ..bl istri.buti on of Cottus Jp..nonicus 'Tuk-up,-obitims 1 71unineu s 7',.eference to thirdr /FrE- tr[1. -1,' ri-ties in both LI itand-Locking ar.d Seciatien erorn.
YITLEIN roREIGN LANGUAGE - TITRE EN LANGUE ETRANGEklE d . ,. • ``. • „ ■ “.17 \% 7 11 ( C 1 (-2 .7.2l > - / ) a./ 17.1" . ...
(0 • tt , 1.7 1 •
•
RErERENCE - ReFdRENCE (NAmE OF Book: OR PUBLICATION - NOM DU LIVRE OU PuBLIcATtoN) Osakaukbge71 University Per -l.ndfical .
PUBLISHER - dDITEUR OsakaE;akuFèi University Press
à CITY - VILLE DATE PAGES
Osaka, Jan. n . 1963- 129-161
• Derartrent Of Fisb.c.rrIps , Iknc ino . 3[297 • - REQUEST RECEIVED FRoM OUP. NUER T)-1-0 1-)1, REQUIS PAR * No -CRIZ DOR
• • Ken f, a no - DEPARTMENT TRANSLATOR • - MINISTRE TRADUCTEUR • - . - . - ')7 YOUR NUMBER DATE' CoMPLEV-:D - ( VOTRE DosSIER NO REMPLIE LE
DATE RECEIVED • Jart2r3r,1966. REÇU LE
-
• .
80r-200-10- 47- ;1
(129) - The Distribution of Cottus Janonicus and TukuJ...-obious Flur_lineus with Specinl Reference to t:12eir Pnculinrities in both Land-Docking and Sneciation - from Arnhidromous Ancestors
Nobuhiko MIZUNO
Biological Laboratory • Osaka Gakugei University
10. Foreword
The migration of fish, es vieil as that of birds, has been known as a peculiar phenomenon and has attracted the attention frP of scientist for a long time* Various classifications of these fish have been uu ,7gested (Myers, 1949). It is well known that some migratory fish cone to inhabit fresh-water regions permanently. This -phenomenon is Called land-locking. In my reportson the Tukugobious filleineus and the large-egg type Cottus japonicus of 1960 and 1961, I suggested that they were the fish which were land-locked and speciated fron the migratory Rhinogobius brunneus + (or i!elated-species) and the small-egg type CottUs japonicus (Lama°, 1960; Mizuno Tanba, 1961). In the following chapter, i would like to study the relationship. between the migration and land-locking of such fish mainly in our country and present the peculiar aspect of the land-locking. of the above-mentioned Tukugobious flumineus and the large-egg type Cottus . japonicus._ Further- more, I would like to investigate their way of spe(eiation in reference to their geographical distribution.
RO similis Gill was formerly'used as the name of this species until ft was corrected by Tagaki (1962) and R. brunneus
(Temminàk & Schlegel) . waà cdoptd. -
,- .• 2
I waeJ given a great deal of help by many people with the investigaticn and arrangement of the data. Professor D. Miyagi of Kyoto University and Professor H. Sugino of Osaka Gakugei University encouraged me and cornented on my study. Dr. Niwa of the Gifu prefecture and Dr. Kawanabe and i his group of Kyoto. University also gave me a great deal of 'assistance. Professor K. Dotsu of Nagasaki University, Profess- or I. Toyama of Tokyo University, and Professor K. Matsubara of Kyoto University gave me many specimens with helpful comment. Professor M. Tsuda of the Nara Womens University, Dr. H. Shiraish of the Fresh-water Products Laboratory, the Saito Requital Museum of Sendai-city, Dr. S. Kimura of Urama-city, Messrs. M. Urata 9 K. lirai of Osaka Gakugei University, and many others helped me a great deal with the collection of specimens. To all of the above-mentioned I would like to extend my deep-felt thanks. *
Th,- Peculiarities of the Land-lockinz-of the'Lpree-eeg Tyne Cottus japonicus t,md Tul_rueobious flumineus - The classification which was published by Myers in 1949 (130) seerned to be Quite relevant to land-locking in reference to the typé of migration. He named the migration between .. sea-water and fresh-water Diadromous migration, and divided this category into, three sub-categories; namely, Anadromous migration, Catadromoils migration, and Amphidromous migration. The Anadromous migration and Catadromous migration-are seen, typically'in the salmon-trout family and the,eel family. The fore' goes from the sea to fresh-water regions in its •n ,.tu-re period, and the latter reversely from fresh-water .regions to the .sea. Both imigrations - have the same purpose of spawning. Amphidromous Migratory fish, however, travel :,back and forth between the sea 'and.fresh-jfiater in their infancy„This is not - a spawning migration. According to Myersp• the : Sicydium ofgoby. ' genus and its relatives are examples. He says that these gobies lny eLigsin rivers and the hatched infants soon go down to the sea, live'there during the first half of their infancy, return to rivers, grow up there, and lay eggs.
, As for:the Japanese-grown Anadromaus migratory fish, there \!- 1 \ salmon and Salmopluvius•genera of the salmon-trout family; • are sJ <\ and Gastcrosteus aauleatus of the stickleback family. Besides, the 1 ariprsjy family migrates in the same manner, although it •••■ .\ does not belong to fish. These fish all go un rivers for spawning in their mature period and in re lation to land-locking it is interesting in what stage of their • growth they go down to the sea. They do not go down to the sea immediately after hatching. For example, the fish which. the salmon and Salno pluvius genus, and Tribolodon belong to • hakuensis have about 10 cm of body-length before going down, and the Gestarosteus aculeatus, whose full body-length is about 10 cm, start going down at about 3 cm. The lamprey family also stays in fresh-water until the completion of metamorphosis. As described above, they spend the first half of their infancy in fresh-water regions.
It is not irrelevant to the above-mentioned fact that land- locking often occurs among such fish. For example, each of the Oncorhynchus masou and Oncorhynchus nerka of the' salmon genus; Salvelinus malma and Salvelinus leucomaenis of the Salmo pluvius genus; Tribolodon haluensis and Gasterostus aculeatus of the stickleback family produces land-locked • fish such as Salmo nocrostoma, 0.n.f. adonis; and Tribolodon hakuensis, Gastercstus aculeatus (the last two have not been named yet). Furthermore those, except 0.n0f. adonis, are all land-locked in rivers (fluvial land-locked). Those which live in fresh- water regions and that in rivers for the first.half of their infancy should not have nuch difficulty renaining,there to live for their lifetime. In fat, land-locking occurs among them and there are also many examples that such land-locked fish live with migratory ones, such as Salvelinus nalma and
its land-locked fish (-U\i- x.c..7,,:.÷,te. in Saghalien, S. leucomaenis and Oncorhynchusmnsou andtheifc land-locked fich and Salmo mocrostoma in Hokkaido and Tohoku district, and G. aculeutus.and itsland-locked fish • (Kobayashi, 1957) in the Ishikari River. .
The term "land-lockine gives an_impression that migratory fish are passively locked-infreshwater regions, but as (131) 0no v(1933) . about Oncorhynchus masou and Ricker (1938) about Oncorhynchus nerka, the 'behaviour of lan d-locking seems to be explained mainly by the relationship between growth nrd naturity. It is not an abnormal phenomenon that Anadromous migratory fish are land-locked, since they do not have to change their early life histories, and in tact they never do. It also happens very easily that the descendants of land-lockec fish become migratory fish again. This phenomenon is proved in Oncorhynchus masou (Ono, 1933) 9 and in Oncorhynchus nerka (Ricker, 1938). It is known in Tohoku district that the migra- ,
tory female -Oncorhynchus masou inter-breeds with the land- locked tale and that rot of their male descendants remain in fresh-water regions, but that most of the females go . down to the sea (Tanaka, 1956). In short, in Anadromous migratory fish, migration and land-locking are, to a great extent, reversible.
;As: for Japanese-grown Catadromous migratory fish, there are Anguilla japonica, A. marmorata, and Trachidermus fasciatus and Cottus hazika of the Cottus japonicus family. There are, however, no land-locked fish reported. The Catadromous migra- tory fish live in the sea during the first half of their infancy, cone up the rivers, and go back to the sea again in their mature period to lay eggs there. The life history is completely reversed in this case and it seems to be difficult for them to become land-locked fish.
Japanese-grown Amphidrous migratory fish are Plecoglossus artivelis, fish of the goby family such as Eleotris oxycephala, Rhinogobious brunneus, R. ziUrinus, Chaenogobius urotaenia, Ch. annularis, Tridentiger obscurus, iuciogobius guttatus, and Sicydilim- japonicum, and the small-egg type Cottus japonicus. They go doWn to . the seaa immediately after hatching, spend the first half of their infancy:there. and.then they goup . and spawn. In this:regard they are rivers where they grow . Similar to 1,iyers' (1949) examples of Sicydium and its relativ-. es. From the morphological point of view, their bodies are frail-looking and their fins are incomplete immediately after hatching, and in theÉe respects they are distinctively differ- ent from Anadromous migratory fish in the . same situation. . One of the reasons that they go down -bothe sea directly _ . after hatching seems tobe that they.have difficulty staying in the stream of.their hatching-place morphologically. Corresponding- ly, there. is no example of their fluvial land-locking• . . that has been reported. . . .
The Amphidrous migratory fish are, however, often land-locked in lakes (the- lacustrine land-locking). The examples are • Plecoglossus altivelis in Biwa Lake and Ikeda Lake and the small-egg type Cottus japonicus in Biwa Lake and Nojiri Lake. Of - the aboVe-mentioned goby family, all species except Eleotris oxycephalà and Sicydiub japonicut is reported to haveA.ts land-locking fish (Uizuho l- i960; Dotsü,'in press)0
fl.though it is mentioned that:A±phidromdüs migratory fish • stay in the se“uring the first half of their infancy, they live only near the shore, where the salt density changes to a considerable extent. Besides, they go back and forth between fresh-water and salt-water during their infancy. • ,Such a life history is possible only if the fish have a 'strong resistance to the sudden change of salt-density. This ;resistance seems to i)e one of the factors that creates the possibility of the occurrance of land-locking in lakes and Iswnmps. It does not seem to natter very much to them h32) whether it is.fresh-water or salt-water as long as it is stagnant, which allows them to live their floating life of infancy. In Such a case also there is no difference between the life history of the miratory fish - and that of the land-locked fish (Mizuno, 1960). Even if they are land-locked once, they seem to be able to go back to the state of the • - Migratory fish easily. There is no possible proof for this at present, but there is also no reason to deny the hypothesis. In short, it can be said that the relationship between Amphidromous migration and lacustrine land-locking, as well afin the case of Anadronous migration, is reversible.
There is, however, one more type of land-locking of the AmphidromouS migratory fish. It is mentioned that there is no fluvial land-looking of Amphidromous migratory fish, but this is only about ones which do not change their life history. Thei-e are, however, ones, although a few, that change their life history to a great extent and that coma to live in rivers. Japanese-grown examples are Tukugobious flumineus (Mizuno,196o) and the lai-ge-egg type Cottus japonicuo (Mizuno, Niwa, 1961). Their bodies become already close to those of the full-grown fish during hatchin,z or deuto ,plasm absorption and they immediately start their lifu at the bottom as the grown fish do. To state this in extreme terms, they practically go through, before hatching, the growth steps which are eouivalent to those of the sea-living period an .Arlphidromous fish. Amphidromous migratory fish seem to be able to live in rivers only if they change the growth process:in such a way as mentioned above. Corresponding to the changes of the life history, there are also morphological transformations. Tukugobious flumineus is distinctiVely different from Rhinogobious brunneus which seems to be similar to the ancestral form of the former in the number of vertebral columns, pectoral fin rays and sDecks. They are clearly different species. The large-egg type Cottus japonicusliS also different from.the SmallHegg -:type in the number of pectoral fi n rays , e.-cg-ize, and life history. The differences seem to be sufficient enough to say that they are different species (Uizuno, 1960; Mizuno & N1wa 9 1961)0 Thus the changes in life history and egE-siza accompanied by norpholor'.cal transformations are necessary for Amphidro-• mous migratory fish to be land-locked. This process does not seem to occur very often. In fact, we have found the. lacustrine land-locking offspring for almost all of the Amphidromous migratory fish mentioned above, but only two fluvial ones. Mogurunda obscura is also a fluvial land-locking fish and its life history is quite similar to that of the two mentioned above. We l however, are not certain that this one is- also an offspring of the AnDhidromous migratory fish. Even if we.recognize this Q. an Amphidromous fish, we,only have • three of them. The land-locking which needs such a speciation or similar differentiation is naturally thought to be non- reversible. From this point of view, it is understood that the fluvial land-locking of Âmphidromous migratory fish is. a peCuliar case which is:different from the lacustrine land-
. , locking of them, and which is also different from the fluvial and lacustrine ones of Àndromous migratory fish, Norman (1931), Svardson (1957), or Wyne- :Edwards (1952) offer many facts, mainly about the life histories and land-locking of Anadromous migratory fish, although these articles, are fragmentary. Comparing. with these foreign-grown •examples, we still consider that the land-locking of Tuku(-:obious, flumineus and the large- egg'type Cottus japonicus are peculiar occurances. Or rather we should say that such land-locking is not yet known. Do such differances also affect s .ûeciations and fornations of (133; distribution ranges.which occur in the pro;:.:es of land-locking? • First, let us take a look at the formation of distribution ranges.:The distribution range of Salmo mocrostoma which is. the land-locked offspring of Oncorhynchus nasou covers the iwhole of Hokkaido and north-eastern Ja-oan and, moreover 9 it extends to the south-western and Japan Sea side of Japan as far as a part of Formosa. Most of the Salmo mocrostoma in these rivers have been produod indtrpendently from Oncorhynchus masou which ha:ye come up the rivers. This phenomenon seems to have occurred ubiquitously; so to speak. This iG not the result of dispersion of. Salmo mocrostoma through fresh-water from w'fiere the original speciation occurred. For exam'ele, as Ohchima (1936) explained, Salmo mocrostona which live in Daikokey in Formosa did not come from ,japan :but diffe .rentia- ted from Oncorhynchus masou which went up to Formosa during the Ice Age. Salvelinus malna und S. leucomaenis, which are the land-locked offspring of , are found in Sagl?halein and Hokkaido and they also seem to have been produced in th a same way 'as Salmo macrostoma. The case of Gasterosteus aculeatus and its land-locking sumecics, Gaterosteus aculeatus microcephalus • is also - the same (Ikeda, 1935). The sanie thing can also be said about the lacustrine land-locking of Amphidrous migratory fish. For example, Rhinogobious brunneus is land-locked in many lakes and ponds throughout the country from Hokkaido to Kyushu (Shiro, Tomito et al., 1960). Sone of them may be transported with young fresh-water trout and so on, as the ones in Towada Iake.which are from Biwa Lake, but certainly not all of them. It is certain that the lacustrine land-locking fish of Tridentiger obscurus in the reservoirs in Tsu City and the . ones in the irrigation ponds in Matsuyama City are independent- ly land-locked.
The rIubject of migration is often discussed with reference .to Anadromous micratory fish, but there seems to be no fact which definitely negates the hypothesis that the land-lockings have independently occurred.in different districts. Ikeda (1935) rather agrees with the hypothesis and Oshima's (1936) idea is also close to it. Hence the word land-locking seems to include such an idea of independent occurrence.
Now the second topic is that of speciation. There are land- locked fish which have differentiated from their migraory ancestors and have made new species or subspecies. For exanple 9 , there are 0.n.f0 adonis in Atan Lake and 0. kawanurai in . Tazawa Lake which are close relatives of Oncorhynchus .nerka,' 0 0 rhodurus and 0.r.±. macrostomus in. the. west of Chubu district , which are close relatives of Oncorhynchus masoU, Salvelinus pluvius and S. japonicus which have differehtiated from S. leucomaenis, Gasterosteus aculeatus microcephalus which is a subspecies of G. aculeatus, Chaenogobius isaza in Biwa Lake which is a close relative of Chaenogobius annularis and Ch. vrotaenia,. and so on. It is thought that such differentia- tions have taken place gradually during living in fresh water after land-locking. That is to say, the differentiation and . land-locking are not simultaneous processes, but there is a considerable period of time between thon. It is interesting 134) that Ricker's study (1939, 1940) of - fish in the Fraser River and its water system in Canada clearly proves this point. Many lakes belong to this water system and some of them have Kokanee and 0.n. kennerlyi which are the land-locki ,ig subspecies of red trout. However, Anadromous migratory red trout also annually go up to the lakes and lay eggs and part of the hatched fish do not go down to the sea but stay their fcr their life-time. That is to say, the land-locked fish and the land-locked subspecies .coexist in the same lakes. They are different not only in form and in contagious resistance to outside parasites, but also in the spawning time and place. Hence there is no interbreeding between the two groups. It • also happens that sone of the young of the land-locked'fish go down to the sea. This case, however, does not apply to Kokanee. Paying attention to this difference, Ricker thoroughly investigates both subsmecies'morphologically, sexually, and ecologically. He divides the differentiative process of Kokanee into two stages: (a) the occurrence of the land-lbcked off- . spring, for which the complex effects of sex and of rate of growth plays an important rôle. At this stage there is no morphological difference between the migratory fish and the land-locked fish and (b)the codification of such land-locked fibh into the typical Kokanee, which is said to be a long process of continuous small gene mutations and natural selee- tfon. He considers 0.n0f0 adonis in Akan Lake as Kokanee.
However, as I mentioned above, the changes in egg-size and life history are indispensable for the fluvial land-locking of Amphidromous migratory fish, and the morphological modif- ication accompanies these in the case of Tukugobious fluminaus (Mizuno, 1960). This also applies to Cottus japonicus (mizuno, Niwa, 1961). In short, the differentiation at the species or subspecies levei- in these cases occur concomitantly with the land-locking process and hence such periodical diviàions as in the case of Kokanee cannot ba applied here. It is possible to introduce different ideas about the differentiation of Amphidromous migratory fish to the fluvial land-locked fish and to the distribution of the latter than to the cases of land-locking of. Anadronuus migratory fish- andlthe lacustrine land-locking of Amphidromous migratory fish. That is to say, it is plossible to think 'that the fluvial land-locking fish which differentiated from Amphidromous migratory fish in a certain location have spread out as far as to the present distribution range. This is a rather orthodox idea because our discûssion on the origin of the present distribution range of primary fresh-water fish has been predominantly based upon the idea of the dispersion from where the original specia- , tiOn.was effected. Furthermore, such fish as Tukugobious flumineus and the large-egg type Cottus japonicus which have completed their differentiations with the fluvial land-locking process seem to be more suitably dealt with as primary fresh-water fish such as cyprinids or catfish. In short, it is rather odd to apply the idea that the differentiation at • the species level or the level worthy of the species occurred independently in each river. However, when we compare the distribution range of the large-egg type Cottus japonicus with that of other Japanese fresh-water fish, we cannot help assuming that they differentiated ubiquitously an so did Tukugobious flumineus. In the next chapter, surveying their distribution ranges and living conditions, I would like to investigate the basis for the ubiquitous differentiations and the concerning problems.
III. The Distribution Range of CottUs jaronicus
First, it is necessary to talk about the scientific term Cottus japonicus. Cottus japonicus in Honshu (the mainland of Japan) used to be identified as Cottus pollux. The name Cottus pollux.was first used by Günther in 1873 for a specimen collected in *the Ishikari'River in Hokkaido. However, the studies of Aoyanagi (1957), Watanabe (1958), and Okada - (1960) indicated thé fish in Honshu WaS different frem the Cottus pollux. Finally Okada (1960) classified it es a' different speciéS_and named it Cottus japonicus. He alSo understood that:Cottus nozawae was a synonym for Cottns pollux. Since I do not have any objection on this point, I have nsed the ter.re Cottus:japonicus. ..ccording to Watanabe (1958), besides Cottus.japonicus, C..kazika, and there is another species called C. reinii whose distribution range is Honshu, Shikoku,r and Kyushu. Howe»ver, as far as I know, there is no such fish that can be classified as C. reinii in Honshu. As I mentioned in my report of 1961, both the large-egg type fish and the small-egg type fish belong to Cottus japonictLs.
The large-egg typ&fish and the small-egg type fish in this species are different not only in egg-size and aire history but also in the nimber ?)f pectoral fin rays. We can distin- guish collected specimens by this - point toc. The distribution ranges of these two types distinguiàhed by three differences were mentioned in the previous report (Mizuno, Niwa, 1961). We, however, did not have enough space for the names of the locations of collections. Vie decided to include them in this report. For the production of the list of locations, I was helped to a great extent by Tokyo University Zoological Laboratory, Saito Requital Museum, and others. I would like to extend thanks for their kind cooperation. I found old names , of the locations in the old samples, but I changed them into new ones in this report with reference to Japan Prefectual Maps and Place-names (1961 edition), edited by the Internat- ional Geographical Society.
Locations of Collection of Large-egg Tyne Cottus Jàponidus
Iwaki River /. Hamano Town, Hiromal• City, Aomori Prefecture Yonashiro River / Numadate, Odate City, Akita Prefecture Kitakami River / Nakatsugawa, Morioka City, lwate Prefecture Towa Town, Wagagum., Iwate Prefecture Tozawa Town, Iwate Prefecture Iwabuchi, Onikubi Village, Tamazukurigun, .Miagi Prefecture Naruko Town, Tamazukurigun, Miagi Prefecture Kisen River /• Oriheki Village, Bansengun, Iwate Prefecture Prefecture Huga River / Hokaigun, Yamagata (ii) Mogami River / ishida Town, Kitanuraiamagun, Yamagata Pref. gfbranches; -upper:-reaChes of -thé; ank-ako -RiVer Y: • middle reaches of the Run River / Sazawa Town, Yamagata Prefecture ' Akagawashin River / Asahi Town, Takawagun, Yamagata Prefecture Ara River / Funawatashi, K. kuni Town, Nishiokiigun, Yamagata P. Abukuma River. / Shirakawa City, Fukushina Prefeeture Koseki, Nishishirakawagun,Fukushima Prefecture Kaji River! Yonekura Village, Kitakanaharagun, Nigata Pref. Kuji River / Kurosawa Villaze, Kujigun, Ibaragi Pref. Kadonoi, DaiShi Town, Kujigun, Ibaragi Pref. Oksa, Daishi Town, Kujigun, Ibaragi Pref. Guna River / Yutcugami Village, Nasugun, Tochigi Pref. Tone River / Niko Town, Tochigi Pref. Muromachi Village, Asogun, Tochigi Pref. Sarugakio, Shinji Village, Tonegun, Guma Pref. Atsuta, Azuna Town, Azumagun, Guma Prof. Ara River / Ashigekubo Village, Chichibugun, Saitama Pref. Tama River / Ochiaihachioji City, Tokyo Sagami River / Taninura Town, Tsuru City, Yananashi Pref. Yanagawa, Otsuki City, Yamanashi Pref.
I Sakanioi River / Gotemba City, Shizuoka Pref. Fuji River / Tenjinpio, Kofu City, Yamanashi Pref. Kaminoyama, Hizaki City, Ya=nashi Prcf. Umekubo, Minamikiomagun, Yamanashi Pref. Shinano River - Chikuna River / Fuse Village, Kitasakugun, Nagano Pref. Buceki Village, Kogatagun, Nagano Pref. Komaki, Ueda City, Nagano Pref. Kawatapura, Kamitakaigun, Nagano Pref. Shinano River - Sai River / Kamishirigawa, Kamiminauchigun, Nagano Pref. Ooka Village, Sarashinagun, Nagano Pref. • Noguchi, Daimachi City, Nagano Pref. - Sotochimada Village, 1.atsumoto City,Nagano,P. I. Village, Linamiandongun, Nagano Pref. Nagawa Village, Minamiandongun, Nagano Pref. Tenriu River / Tatsuno Town, ramiinagun, Nagano Pref.
ma City, Nagano Pref. • Toyo River / Shimotsugu Village, Kitasetsuratugun, Aichi P. Kiso River (main stream) / Tamataki Village, Nishitsukumagun, Nagano Pref. Pukushima Village, Nishitsukumagun, Nagano Pref. Natsugawa, Natougawa City, Gifu P. Fulmoka,:Tukuoka Village, Enagun, Gifu P. Kiso River - Hida River / Karo Town, Ekitagun, Gifu P. Jinbuchigawa, Shchiso Village, Kamogun, Gifu Pref. Kiso River - Nagara River / Hirugano, Takawashi Village, Gunjogun, Gifu P. Yawata Town, Gunjogun,. Gifu Pref. Takasa3o, Takatamura, Gunjogun, Gifu Pref. Itadori Village, Bugigun, Gifu Pref. Gifu City, Gifu Pref. Kutoriu River / Ishitetsuhaku Village, Onogun, Pukuyi Pref. Yu-Pa River / Sasazato„ Elyama Town, Kitakuwatagun, Kyoto. U River / D, kenogur Kyoto Mia River / Nomatabachi, Miakawa Village, Takigun, Mie Pref. Biwa Lake - Amano River / Tanjo, Yonehara Town, Sakatagun e, Shiga Pref. Yodo River - Kitsu River / Nogi, Otsue Village, Udagun, Nara P Nahari City, Mie Pref. Yodo River - Katsura River / Ido, Kiokita Town, Kitakuwatagun, Kyoto Akuya, Kiokita Town, Kitakuwatagun, Kyoto Kameoka City, Kyoto Uenobashi, Ukio-ku, Umetsu, Kyoto Sakio-ku, Yase, Kyoto Yodo River - Angi River / Tobetsuin, Kameoka City, Kyoto ma River / Nose Town, Toyonogun, Osaka' Ichi River / Kanzaki Town, Kanzakigun, Hiogo Pref. Seho River / Tobisawa Village, Shishikurigun, Hiogo Pref. Asahi River / Yuhara Village, Babagun, Okayama Pref. "Cori() River / Tetsta Town, Atetsugun, Okayama Pref. Eno River / Mirazaka Town, Ftutamigun, Hiroshima Pref. Doki River / Nakatoshi, Miai Village, Ayautagun, Kagawa Pref. Hiji River / TOnohijikawa Town, Kitagun, Ehime Pref. Muromi River / Hitotcuya, Sagara Town, Sagaragun, Fukuoka P. Chikugo River / Sakae Village, Hitagun, Oita Pref.
• ••• ' Loctions. of Collection of Sm!:-.11-eg Type Cottus Japonicus
Towada Lake / Kamitizagun, Aomori Pref. Natori River / near Aoba Castle, Sendai City . Abukuma River / Ejiri, Tsunoda City, Miagi Pref. Shirakawa City, Fukushima Pref'.
Kuji River / Tatsunokuchi, Ibarai Pref. «; 3 Fuji River / Minoshiro Town, Minamikiomagun, Yamanashi P. Nojiri Lake / Shinano Town, Kamiminaochigun, Nagano Pref. Toyo River / Shiragi, Minamisetsurakugun, Aichi Pref. Ishimaki, Toyohasi City, Aichi Pref. Kiso River (Main Stream) / inuyama City, Gifu Pref. Kiso River Nagara River / Takasago, Takata Village, Gum_jogun, Gifu Pref. Hodojima, Seki City, Gifu Pref. Shirogane, Seki &ity„ Gifu Pref. . Kose, Seki City, Gifu Pref. Chiusetsubachi, Gifu City, '3-ifu Pref. Kiso River - Kai River / Gifu Pref. Sai River / Kanazawa City, Ishikawa Pref. U River / Tango Town, Takenogun, Kyoto Biwa Lake / Iba Village, Kanzakigun, Shiga Pref. Asai Village, Ihoreun, Shiga Pref. Nagahama City, Shiga Pref. Sakatagun, Shiga Pref. 7atada Town, Shigagun, Shiga Prefecture Shiga Town, Shigagun, Shiga Pref. Kino River / Hashimoto City, Wakayama Pref. Wakayatia City, Wakayama Pref. Kako River / Kako City, Hiogo Pr. Yada River / Totsuka Village, Mikatagun, Hiogo Prefw Ohio River / Kahara Town, Yatogun, Katuri Pref l Saba River / Bofu City, Yamaguchi Prof. Nyodo River / Ino Town, Awagun, Kochi Pref. Muromi River / Nishishin Town, Fukuoka City, Fukuoka Pref. Unknown river - Muramatsu Village, Sonogigun, Nagasakie Pref. •
Illustration 1 shows the above7mentioned places of (138) collection. As you can cee in the illustration, Cottus japonicus, either large-egg type or small-egg type, is distributed in Honshu, Shikoku, and Kushu. Hokkaido used to •be includedln this distribution range because Cottus pollux was identified as Cottus japonicus. Cottus pollux, as 'mentioned before, is a different species. Okada and Nakamura (1958) already mentioned that the distribution range of Cottus
- Saponicus covered almost all of Japan except Hokkaido. Although .,_oyanagi (1957) mentioned that none of this species lived in Kochi prefecture and the Pacific coast of Kyushu, recently, •Ito collected the small-egg type. in the Niyodo River in the • Kochi prefecture and Kamahara (1961) also reported that this species had lived for a long time already in the rivers east of Ahi City in this prefecture.
'As you can see in Illustration 1, -Cottus japonicus has a wide .range of distribution and moreover, it is distributed (139) in the rivers in each district. In'Okada's repôrt (1960), the Ohata River in Aomori prefecture, the Haguro River in Yamagata prefecture, the - Nureishi River, Sarugaishi River, Sekii River e •Koshi River in Iwate prefecture, Kitakata -district of the Fukushima Prefecture, the Sumida River in Tokyo -, and the Takatsu River in Niigata and Shimane prefedture are also men- tioned for this species. (The types. they belong to are not mentioned). As these fish are apt to hdde themselves among the stones at the bottom of rivers, they are rather difficult to be collected and observed. Hence we are apt to miss them - in rivers during a short period of collection and observation. If we consider this point, we can hardly be wrong to judge that this species, especially the large-egg type fish, inhabits almost all of the rivers in Honshu, Shikoku, and Kyushu except those with strong acid content or polluted ones. • In my report of 1961, I pointed out that while the_laree-egg type Cottus janonicus which spends its whole life in rivers was collected mostly in the upper streams, the sma11-ei2g type which were Amphidroous migratory fish, as well as Cottus Kagika which were Catadromous migratory fish, were mostly found in the middle or lower rivers. This tendency is clearly seen in Illustration 1. The large-egg type Cottus japonicus definitely belongs to stream fish.
jOkra
cà 1. à E; irjU311(0) • iigi%M(0)0)5111: -1M Fig. 1. Distributions of the large-egg type (0) and the small-egg type (0) of Coitus japonicus •
■.■
The Distribution Pane Cf Tukugobious flumineus
It is known that the distribution range of Rhinogobious brunneus in our country covers all of Jaman including Hokkaido and Ryuku (Tomiyama, 1936; Okada, 1960). My invest- igation also confirms it and furthermore it Droves that Tukugobious flumineus lives only in the south-west of Japan (Mizuno, 1960). As the locations of tha collection also are no -t; shown in the report as in the case of CottUs janonicus, I would like to present them here with additional ores which • . have been recorded since the last report. As in the case of Cottus janonicus, I was helped very nuch by Tokyo University Zoological Laboratory, Saito Requital Museum, and others. (140 1 also changed the old names into the new ones. Locations of Coljectir‘n of Tulcurobius flumineus
Fuji River / Uizaki City, Yananl.shi Prefecture • Tenriu Rive / Tatsuno Tom, Inn Citlr, Eg ,fano Prefecture
Inazato Villaue, In City, .:*.ag -lo P.
Ina City, Magano P. •
Toyo River / Toyohashi Cit ,•Ichi. P.
Jintsu River / Takayama City, Gifu P.
Sho River / Tode Town, Toyama P.,
Kiso River / (main strear;.) ..Enegun,Gifu P.
Natsnp-awa C4 ty, Gifu P.
Kiso River --Hida River / resudagyn, Gifu P.
Og-unra Masudagun, Gifu P.
Shirakawn Toyn, Kamogun, Gifu P.
,Kiso Aver--Nagara River! Gu -,, jor.un. Gifu P.
- Seki City, Gif*,:' P.
Gifu City, Gifm P.
Kiso River / Yorc«un, Gifu P.
. . Makite Village, Gifn P.
Biwa Lake-- .rA)su .Y.-ernre Field, Toynni Tor.Kovegn ,-, Shigsa P.
Ocra Field; Toyema Tonn, Kogagun, Shiga
*The iroper reaches Posu Rive— Dan, Toyana To•n,
Kog;, 1711r Shiga P.
Yod° River--Kitsu River,/ Kagagun, Mie p"
teno City, Mie P.,' •
*Nahari City. Mie P.. •
• :Udarrun; Nara P. •
• SaFercgur, Koto
Uatsnka Town. Sagarawun, Kyoto
Yodo River--Ugi River / Otsu City, Shign P.
Yodo River--K-tsura River/Kitakuwatn7un,
. . . . ' Kameoke City, Kyoto . .
• • Salçyffl•cn. Yyoto. ,
Yodo River ( rain. st,ream ) / Debai , T itsu1i CYr, Oka (2,"' Oen Kemooka Cit.. , Osaka
Y-rato River Katakand '1,32 Tra n Tc -14 n 1.' r ikoc.c n r. Osaka h:i no, ro City, ,iÇreh 1 ro Cuir 057i; ka
Kin() River Yorl-'1.ro . 'Pe rp (1).1e1-1 shilroFhi ro Vi.-1 Yo phinot7-un 'Ara P.
Kaman° River--Totsu River • An ,111 AmP YoshiroFur 1g^ -.5•;:, P. Takio, Ara P.?ra. Vi11re, Yoshirorun„ Nar? P.
z Inc. River nomEr3 , rond. Osaka • crt Os.:9 1/e
Y.., mrhe Os.-.1.1ca Karanihi Cjtir, Hioczo P.
Ura River / rrn ta. , TOT-Tr ,K.i•talcuwe.tp.pun, K7,reto • rrs Kitak.-..)ratar;un, Kyoto Tye be City, Kioto
Soho irer Shishikuri Ffun. 7.-lio‘:."0 P. , • Asana. .L1-71.ver \Ur e- e. Okavara. P. • • Uhs.ra rr,..rfwa Villa-e, Okayama P.
Korio River / Yorio City, P.
Eno River itsut..'a City, rosbina P. • ° Ota ?liver I Koh-i.re Ton. 11".1 rtarfl.)n, irohirr P. 7.1rci reem-n, P. (->ki pun, I•ii roshi ra P. C.-31116,cm Eiroshi ,-a P. Town ,Yari.bp-.n.tapur., Hi.roshi ma P.
Yoshino River P.,:er.9 a Torr , shi shiya P. ipu n , Tc-.)1<-us1.-ti P.
yoci o P. 0 P nd V i:e , CUM. _
Shi rsh à River Su2.3.k-_i_ City, Kochi. P. . _ . .
Soshi a River TP rai VfLl ple, Eciu r, -ire P. - . Jushi n RiVer--Tobe River / Tobe Ivorrun, hire P. ..
River - -1:(.alf.7.,-Prr.e. River/ Tom, -Iv° , Eh.ireP.
Oitp iiiver--Seri River Tour, Ncor1.7r -n, Oit. a .
Eki c'a.te . v er -Po ••= River/ • ..;•..ns.hini.n Torn., Us a n, 01.ta. P.
P--uver Yanakuri Town, Shl.roteg...un , 05 ta., P. . :
Sc.leda 911, !U1 1 P. -
Tsukugo River •• / Kj.takoo.ini Vilie, Asogun, P. er,\ The above-mentioned places of collection are shown in (-14.1-) Illustration 2. But the places which are in the carne river • and are close to each other are abridged. Although they are . not shown in this illustration, Xitatane•Village, Tanigawa- , guchi in Tanegashina,. Hisha River in Okinawa, and Amami- oshima have this species.
)
• 2- /3 3 .1*, 05-)le Fig. 2. Distribution of Tukugobius flurnineus
There is no collection in the south of Kushu. I, however, •(142) Would not negate the possibility of their living because our investigation was rather rough. Since they inhabit Tanegashima or the Ryuku Islands, they quite likely do in the south of Kyushu. In short, this species inhabits.the south-west of Honshu, Kyushu, Shikoku, and Ryuku Islands.
I would like to discuss the east boundary of this distribu- tion. From Sept. 6, 1961 to Sept. 10, we surveyed thè - - Touryu River, the Sai River which is a branch of the Sinano River, the Puji River, the Sagami River, '3hs. Sakenioi•River, and the Oce River, a branch of the Kano River, laying stress on the upper middle basins. Of these rivers, the Tenryu River and the Fuji River had this species. They are Ine 'act • lisited above. However, despite/that, we collected quite a few fish.ineluding the fish that lives at the botton of the water, we neither - collected nor observed Tukugobious flumineus in the Sagami ' River passing through Otsuki City. and Yamura Town in Tsuru City in the Sakenoi River in Gotemba City , and in the Ose River at Iwanami in Gotemba City. And Norio Pukushima of Gunma University, who undertook a wide range of collection of Cottus japonicus along t1-1 Tone River, did not see this species J. Considering the above results, I am quite certain that the distribution range of this specied on the Pzicific Coast side is limited to the west of Fuji River. As for the Japan Sea side, we surveyed th Sai River in i:atsumoto City and in Meika,Town. We collected Rhinogobious brunneus with rainbow trout and Tribolodon hakuensis, but it seemed that . the lacustrine type land-locked fish had fled from a breeding pond. And Tukugobious flumineus was not found in both places. Although these two places are not enough to decide, if the Sinano River water syStem (Sinnno Basin) also is not inhabited . the species, the boundary is drnwn between the Gintsu River by and the.Sinano River, as you can see in Illustration 2. The line from the Itoi River on the Japan Sea side to the Fuji River on the Pacific coast side almost accord with the central trough belt which is the border line of the west according to the geological structure (from the Itoi River to the . Shizuoka line). It is an interesting fact that the distribution of Tukugobious flumineus is limited to the west of this line. (143) I discussed the distribution range of Tukugobious flumineus in the previous chapter, but I would like to mention its living condition in this chapter0 I already reported that Tukugobious flumineus which were the fluvial land-locked fish were collected mostly in the upper-middle and upper reaches of rivers, while Amphidromous migratory Rhinogobious brunneus were found in the middle-lower reaches of rivers (Mizuno, 1960) Mizuoka (1962) also confirmed this in the Ota River water system and Illustration 2 clearly proves it. Many fish of this species were collected in the rivers in the south-west of Japan. In Illustration 2 we can find almost all of the main rivers in the area, which amount to 82. It is confirmed that in 29, that is, more than 1/3 of 82 rivers, the fish of this species live. If we undertake a survey over the rest of the rivers to know whether or not they live there, we are quite sure to find that they do. Because we find them almost always whenever We examine the upper-middle reaches of rivers in the area. 'o' we investigated last year for the first time such rivers in the listas Sinraku River, Ani River, Akuta River, Makio River, Inona Riverand the branch, Dairoji River, Yono River, Yamate River, Yamat6 River and the branch, Ishi River, and Tojo River, and we collected them without fnil. As far as I know, the only exceptions are the U River in Kyoto-fu and the Muromi River in the Fukuoka Prefecture. They are, however, little rivers whose lengths are about 20 km0 end the very fact that the fish do not live there is rather interesting, as I explain later. • • ,
As' Mizuoka (1962) mentioned with -srecial reference to the Ota River water system, the fish of this species live everywhere in the middle and upper reaches of the branch or main stre-ee, of a river. This tendency is shown to a certain extent in the Kiso River water system in Illustration 2 It was also confirmed in the Kitsu River and the Inona River in 1962. I plan to report this in detail later. mus they inhabit quite widely in the main stream and branches Of a water system. And the population density is quite high as compared with that of other fish. 3-5 per 1 m2 is a «rather 2 low figure for them. The density often reaches 20-60 per 1 m in sone places. However, the situation is a little changed near the border of.the distribution range. Sone branches do not have then. For example, in the Tenryu River water system, they inhabit in great quantities the lower reaches of Sulea Lake, but none of them do in the Kami River and Kokotu River, which flow into the lake, except the lacustrine land-locked type Rhinogobious brunneus that have come up from the lake. In the Fuji River water system also, they inhabit the En River but not the upper middle reach of the Puefuki River°
Tuitugobius flumineus does not seem to inhabit lakes and • - ponds. They were all Rhinogobieee .brunneus, but not this species that we collected from more than 20 lakes and ponds (I will report later). Herre (1927) 2 who made a record of the sub- sp ec i es of IthiS species, also :Lentione.d that the fish lived in the streams. In short, it is understood that the fish of. the Tukug obius flunineus genus are no t only mountain fish (144) but also river fish.
Study
I would like to examine Cottus japonicus first. As nentionod above, Cottus japonicus, both the large-egg type and the small-egg type inhabit Honshu; Shikoku, and Kyushu. Especially the large-egg type.inhabit almost . all of the upper reaches of rivers within the distribution ranee. One of the problems here is whether the distribution range has been gradually fo.ened from a long time ago or whether it has been forred recently. If it is rather recent, the differentiation of • the l e-e - '- e" type has ouite likely taken place independently in eech river. But if it is old, the differentiation seems to have occurred in a certain location. But the recency end oldness cannot be decided by such a standard as a million • e yea,s ago . They have to be discussed in a comparative sense with other fish. To prepare for this, we should discuss the formation period of fresh-water fish.
concerning the distribution ranges of Japanese frash-water j. • • fish, Jordan (1901), Kawamura (1918), Tanaka (1931), Berg (1933) . and Mori (1936) have done research. Recently, Aoyanagi (1957) • and Okada (1961) have published the results of their research. . All of the research 9 however, belongs to divisional biogeog- raphy and mainly concerns dividing Japan into some geoeraphical divisions. As far as the topography is concerned, they discuss • only to the extent whether there is any bridge. There are apparently two groups of fresh-water fish which differ,from each other to a great extent in the spread of distribution , . that is, the migratory group whose members spread their \ distribution by going through the sea, and the primary fresh- water group whose members do by fresh-water only. The above- mentioned research does not seem to concern this difference. Aoyanagi (1957) - seems to be interested in this point but also ignores the difference when he arranges the biological divisions.
Geherally speaking, fresh-water fish in Honshu, Shikoku, and • Kyushu consist of primary fresh-water fish which came from thesouth and migratory fish which came from the north. They ./mee-- together in these islands so to speak. When we study , the distribution condition in -,:hese areas 2 we cannot ignore the difference in migration and dispersion of the two groups. Referring to Aoyanagi's data (1957), he chose fresh-water fish which were distributed almost all over the country and
divided into th'e; two groups 9 that is 2 primary fresh-water • fish and migratory ones which included their land-locked fish. The result is shown in Table
A point which attracts our attention in this table is that almost all of the primary freshwater fish tend to inhabit the lower reaches of rivers,,that is to say, they are ground fish. The only exception to this in the table is Moroco steindachneri which ie a mountain fish and-is seldom found7.. 14-5) in the lower waters. Although there are SOM.': 'other Mountain
fish such as Pungtungia herzi, Zacco temmincki, Hemibarbus 1st,. longiroseris, Cobitis delicata, and Liobagrus Reini, they= only inhabit south-western Janan. Tukugobius flumineus which can be treated as a primary fresh-water fish after being land-locked is . also a mountain fish, and, as mentioned in Chapter 5, inhabits south-western Japan. The difference between the ground fish and mountain fish can be found in comparison with their eubspecies. Of the above-mentioned mountain fish, Zacco temmincki, Hemibarbus longirostris, and Cobitis delicata
have subspecies of the same genera such as Z. platypus, • H. barbus, and C. biwae. The eubspeci.es inhabit the ground,.
as well as the mountains, but their distributions spread more towards north-eastern Japan than the distributions of the former three.
When temperature plays an important rôle as.a limiting cause of the distribution, one with a higher position in the vertical distribution will generally inhabit north in the horizontal distribution. This is well-known as Hunbolt's Law. It is well-known that the distribution of salmon and Salmo pluvius genera which are migratory follow Hunbolt i s Law. . However, the fact that these mountain primary fresh-watèr fish only inhabit south-western Japan directly opposes th'e law. The difference in the speed of their migration and in their topographical relationships seem to be very relevant. to the conflict.
e 1. *en • nEl • 12J:e.-.o.3