Food of the Fishes of Otago Harbour and Adjacent Sea.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 68, 1938-39, Page 421

Food of the Fishes of Otago Harbour and Adjacent Sea. By David H. Graham, F.R.M.S., F.Z.S. [Read before the Royal Society of New Zealand, Canterbury Branch, April 6, 1938; received by the Editor, July 12, 1938; printed separately, March, 1939.] Contents. Introduction. Sources of Data. Classification of Fishes in Terms of Food. Food-group I: Fishes Preying on Fishes. Table I. Analysis of Food-group I. Food-group II: Mollusca. Table II. Analysis of Food-group II. Food-group III: Crabs. Table III. Analysis of Food-group III. Food-group IV: Other Crustacea. Table IV. Analysis of Food-group IV. Food-group V: Annelida. Table V. Analysis of Food-group V. Food-group VI: Echinoderms. Table VI. Analysis of Food-group VI. Food-group VII: Plants. Table VII. Analysis of Food-group VII. Food-group VIII: Miscellaneous. Analysis of Food-group VIII. Summary. Acknowledgments. References.

Introduction. Graham (1938, pp. 399–419). In this section the food of fishes is studied in detail, including the types of food, allowing classification of fishes in terms of food to be suggested. From the examination of eleven thousand stomachs of fishes, specific lists are given of the food organisms identified, with an analytical discussion in each case. The list includes fishes, molluscs, crabs, other crustaceans, annelids, echinoderms, plants, and miscellaneous. Sources of Data. Though unavoidably incomplete in many directions, the data collected during the period of study permit an attempt at analysis and deduction, and indeed the importance of this subject, and the central position of the food-problem in the complex of factors which conjointly form the marine environment, demand some such attempts at analysis. The main sources of data concerning diets are as follows: 1. Structural studies, viz. examination of the mouth, teeth, gillrakers, etc. This evidence may be indirect or deceptive, e.g., the discovery that the diet includes microplankton captured by a straining mechanism of gill-rakers does not eliminate the possibility of normal predatory methods by jaws and teeth. On the other hand, this method may detect feeding habits which might not otherwise be indicated, or may explain migratory or other peculiarities. 2. Aquarium observations. These indicate possibilities which should be verified in other ways rather than accepted uncritically as representing normal feeding habits. 3. Examination of stomach contents. This is the most obvious and direct means of investigation, and the one which is most reliable and fertile in its results. There are, however, difficulties, notably those arising from (a) the rapid maceration of food, especially in the case of fast-travelling pelagic species, (b) opportunism, (c) changes in diet with the seasonal changes, growth differences, etc., and (d) the tendency for some fish, especially deep-water fish such as groper, bass groper, and bream, to regurgitate the stomach contents when hooked. Classification of Fishes in Terms of Food. The following grouping appears as consistent as any hitherto devised, for although the categories are not mutually exclusive, in that certain species have a claim to inclusion in more than one group, yet this is not due to the choice of heterogeneous bases for the subdivisions, but to the inescapable fact that some species exhibit a much wider range of feeding habits and of foods than most. The basis of classification is in fact strictly homogeneous, in that it concerns exclusively the nature of the food eaten. Unfortunately, circumstances did not permit the preparation of food-lists of the microphages, an especially interesting group in that its members illustrate various stages in the forsaking of macrophagous nutrition (which is ancestral among fishes,

as shown by the dentition), and have abbreviated the usual food cycles by feeding on micro-organisms which normally represent early points in the cycle. To this end there is a tendency for the gill-rakers, originally a straining mechanism for the protection of the gills, to become elaborate sieves whereby micro-organisms are removed from the respiratory current to provide an accessory food supply, or in specialised cases perhaps even the essential food supply. Although the point is too wide to be decided in terms of fishes only, it appears likely that the present provisional classification, strictly in terms of food, may prove less satisfactory than one which takes into account the nature of the food-capturing mechanism. In particular, the group usually referred to as the planktivores is composite in that some members (John Dory, silver dory, boarfish) are microphages with a typical microphagous straining mechanism, others (pilchards, sprat, ahuru, opalfish) feed on small organisms such as copepoda by means of a macrophagous mechanism. Similarly the larval tarakihi is a planktivore, yet perhaps as truly a macrophage as the adult. On the other hand the sand flounder seems to be a microphage in its early stages and later a macrophage. Flounders, though listed as feeding on microbenthos, do so by means of the normal macrophagous mechanism. For these reasons the treatment of the microphages is provisional only. A. Macrophages; feeding on larger (macroscopic) organisms. I. Indiscriminate devourers of animals and plants (“omnivores”)—Red cod, warehou, kahawai, snapper, threepenny, cockabully, leather-jacket. II. Discriminating between animals and plants, not devouring both. (a) Carnivorous; feeding on macroscopic animals. (i) Feeding on the nekton—food-group i, in part food-groups ii, iii, iv: Most sharks; dogfishes; conger eel, hake, bass groper, horse mackerel, yellowtail, barracouta. (ii) Feeding on macro-benthos—food-groups ii, iii, iv, v, vi, in part I: Skate, elephant fish, silverside, witch, flat-fishes, tarakihi, moki, spotty, girdled parrot-fish, stargazer, Crapatalus, flathead, black cod, thornfish, rockfish, red rock cod, toadfish. (iii) Overlapping composite group; feeding on nekton and benthos: Eel, rock cod, black cod, green-back flounder, bream, groper, trumpeter, blue cod, hake, ling, sea perch, pigfish, red gurnard, sucker. (b) Phytophagous; feeding on the macrophyton—food-group vii: Garfish, marblefish, kelpfish, scarlet parrot-fish, blue cod, bully. B. Microphages; feeding on micro-organisms and detritus (no food-group prepared). (i) Indiscriminate devourers of plankton and micro-benthos: Bellows-fish, pipe-fishes, seahorse.

(ii) Discriminating, of limited vertical range. (a) Feeding on plankton: Pilchard, sprat, minnow, ahuru, roughey, John Dory, silver dory, boarfish, larval tarakihi, Notothenia, opalfish. (b) Feeding on micro-benthos: Moki, flounders (part). Food-group I: Fishes Preying on Fishes. The first table includes the species of fishes found to have been preying upon at least three other species, the predators being named in the vertical column. It so happens that, with the exception of H. medium, the prey includes all species of fish known to have been eaten by at least three predators. Table I.—Fishes preying on three or more species of fishes. Predators. Prey pilchard sprat S.S. pipefish L.S. pipefish seahorse garfish red cod ahuru sand flounder common sole lemon sole mullet warehou blue cod barracouta opalfish cockabully ling sea perch pigfish suckerfish horse mackerel Others Total tope X X X 3 porbeagle shark X X tope 3 spined dogfish X X X X X X 6 skate X X X X X X X X 8 conger eel note (a) 3 whiptail X X X 3 red cod X X X X X X X X X X X X X X X X X note (b) 28 witch X X X X 4 bream X X X X 4 bass groper note (c) 3 groper X X X X X X X X note (d) 12 kahawai X X X X X X X 7 trumpeter X X X X X 5 blue cod X X X X X X X X X X X X note (e) 16 flathead X X X X 4 Maori chief X X X X X X X X taumaka 9 barracouta X X X X X X X X red gurnard 9 kingfish X X X X X 5 ling X X X X X X X X X X X X X X X note (f) 19 sea perch X X X X X X 6 toadfish X X X X 4 red gurnard X X X X X X H. medium 7 (a) Conger eel eats marble fish, greenbone, threepenny. (b) Also smooth-hound, skate, elephant fish, silverside, smelt, witch, rack cod, green-back flounder, greenbone, H. medium, leather jacket. (c) Smooth-hound, javelin fish, whiptail. (d) Whiptail, bream, yellowtail, and its own young. (e) Rock cod, bastard red cod, spotty, rockfish. (f) Lamprey, skate, flathead, triggerfish.

In addition to the above, the following predators were found to have been feeding on only two species:— Predators. Prey. smooth-hound sprat flathead thresher shark pilchard sprat elephant fish red cod lemon sole hake pilchard ahuru rock cod cockabully suckerfish green-back flounder ahuru mullet warehou pilchard ahuru horse mackerel pilchard witch catfish common sole Helcogramma medium Finally, the following cases of predation on only one species were detected:— Predators. Prey. carpet shark cockabully sand eel sea perch silverside ahuru eel freshwater perch short-finned eel freshwater perch bastard red cod suckerfish brill red cod sand flounder ahuru common sole pilchard lemon sole red cod yellow tail pilchard snapper pilchard banded parrotfish seahorse stargazer ahuru rockfish cockabully Analysis of Food-group I.—One important qualification in interpreting these lists is the number of fishes examined. In general, the examination of a very few stomachs of a given species is almost nonsignificant, whereas a very large number would unduly increase the number of infrequent and unimportant entries. This qualifies the following comment concerning red cod, of which well over a thousand stomachs were examined; the only others in which the number exceeded a hundred were groper (800–1000 stomachs), ling (500–600), blue cod (400–500), barracouta (400–500), smooth-hound (400–450), kingfish (350–450), kahawai (350–400), spined dogfish (350–380), sea perch (300–350), red gurnard (250–350), and bream (200–250). In the other species the number of cases examined was approximately in proportion to the frequency, as elsewhere recorded (Graham, 1938), and as far as possible distributed evenly through the annual cycle.

With these qualifications, points of significance are that:— (1) Red cod easily heads the list for the number of species eaten. (2) The Elasmobranchs, of no commercial value at present, figure prominently among the predatory fish. (3) The flatfish, on the contrary, rely on other sources of food. (4) The larger abundant species, including several shoal fishes, make free use of this food supply, as, indeed, from their numbers they are compelled to do. The chief ones are spined dogfish, red cod, groper, kahawai, blue cod, barracouta, ling and red gurnard. (5) On the other hand, certain other fishes, some of them even more abundant, smaller in size, depend rather on plankton and other invertebrate sources of food. In this group come the pilchard, sprat, minnow, silverside, smelt, ahuru, threepenny and cockabully. The last series deserves special attention, because these superabundant fish, depending on numbers and exuberant fertility rather than size and individual self-protection, are the main sources of food for the fish-eaters. The most important is the pilchard, known from 19 of the 47 fish-eating species; of these, four are ground fish, three others frequent rocky localities, and the rest are rovers. Next in order is the ahuru; it is surprising that such an abundant fish, detected in numbers in red cod stomachs within three days of the commencement of the investigation, had not hitherto been known from Otago waters; it is now known to be eaten by at least 17 species, including six species of ground fish and three of rock fish (whence it is evidently not restricted to the bottom), and must be recognised as a major source of food for fish-eaters, second in importance only to the pilchard. Sprats were found in (if cannibalism can be included in the count) the stomachs of 16 species, and red cod may claim the same distinction; both species inhabit all depths. The cockabully, known from nine species, illustrates the point that the number of species is an imperfect guide to importance in the food cycles, for this fish does not bulk largely in the stomach contents. Other fishes eaten by a number of predators include sea perch (nine species, including cannibalism), mullet, opalfish, seahorse (each seven species), short-snouted pipefish, common sole (six), long-snouted pipefish, sand flounder, lemon sole, pigfish (five), blue cod, barracouta (four). Altogether the lists include 47 predatory species and 48 food-species. More prolonged observations would undoubtedly increase these figures, and meanwhile it would be premature to attempt too close an analysis of the data. Mustelus antarcticus (smooth-hound) needs special mention in that though strong, active, swift, and roving by nature, it does not deserve its bad reputation. The vernacular name, and association with shoals of small fish, have led to the belief that it devours fish in large quantities, but an examination of at least 400 stomachs gave evidence of fish-eating on three occasions only; two specimens from Blueskin Bay had 23 and 19 sprats respectively, and one from Quarry Point had eaten two specimens of Kathetostoma.

Food-Group II.—Molluscs. In table II, showing the fishes found to have been feeding on three or more species of molluscs, and in the supplementary lists, it is to be understood that octopus and squid, which are important foods of many fishes, are each treated as a single species. Table II.—Fishes preying on three or more species of molluscs. Predators Prey octopus squid Maorimactra ordinaria Zethalia zelandica Zenatia acinaces Myllitella vivens Nucula nitidula Turbonilla zealandiae Tawera spissa Micrelenchus tenebrosus huttoni Calliostoma punctulatum Chlamys celator Rhyssoplax aerea Inschnochiton maorianus Others Total smooth-hound X X X X 4 skate X X X 3 elephant fish X X X X X X Note (a) 15 red cod X X X X X X Note (b) 8 brill X X X 3 sand flounder X X X X X X X Chione stutchburyi, Gari lineolata 9 g. b. flounder X X X Amphidesma subtriangulatum 4 common sole X X Paphirus largillierti, Xymene plebeius 4 lemon sole X X X 3 warehou X X X X Note (c) 8 snapper X X Gari stangeri 3 tarakihi X X X 3 moki X X X Note (d) 12 trumpeter X X Mytilus maorianus, M. canaliculus 4 scarlet p. fish X X X X Acanthochiton zelandicus 5 spotty X Note (e) 5 girdled p. fish X X Note (f) 7 banded p. fish X X X Note (g) 6 blue cod X X X X X Note (h) 9 ling X X X 3 pigfish X Antisolarium egenum Cantharidella tessellata 3 red gurnard X X X X X X Maoricolpus roseus 7 (a) Also Antisolarium epenum, Notolepton sanguinea, Gari stangeri, Soletellina nitida, Macoma gaimardi, Solemya parkinsoni, Paphies australis, Notosetia sp., Zediloma corrosa. (b) Zediloma arida, Maoricolpus roseus. (c) Melagraphia aethiops, Xymene plebeius, Lepsithais lacunosa, Cominista glandiformis. (d) Notosetia sp., Paphirus largillierti, Dardanula limbata, Subonoba foveauxiana, Paricoplax crocina, Rhyssoplax canaliculata, Amaurochiton glacus, Haliotis iris, Atalacmea fragilis. (e) Zediloma arida, Dardanula limbata, D. olivacea, also the brachiopod Terebratula inconspicua. (f) Chlamya zelandiae, Chione stutchburyi, Amphidesma subtriangulatum, Tugali elegans, Paricoplax crocina. (g) Haliotis virginea, Buccinulum mutabile, Cellana ornata. (h) Mytilus maorianus, M. canaliculus, Paradione multistriata, Cantharidella tessellata.

In addition to the seven species in the above table which had fed on cephalopods and other molluscs, the following had fed on both octopus and squid but not on shelled molluscs: seven-gilled shark, porbeagle shark, spined dogfish, conger eel, whiptail bream, groper, bass groper, horse mackerel, yellowtail, Maori chief, barracouta, and (southern) kingfish. The only other species found to have eaten two species of molluscs were seahorse (Micrelenchus dilatatus, Melaraphe cincta), and greenbone (Micrelenchus tenebrosus huttoni, Melagraphiaaethiops). Finally, the following are known to have eaten one species only: Predator. Prey. turbot Micrelenchus tenebrosus huttoni mullet Tawera spissa stargazer Zethalia zelandica catfish Amphidesma subtriangulatum black cod Mytilus maorianus rockfish Sypharochiton pellisserpentis suckerfish Zediloma corrosa silver-warehou squid Analysis of Food-group II.—It is not implied that these lists are complete; confirmed mollusc-eaters such as elephant fish and moki undoubtedly consume many molluscs other than those recorded, and the stomachs of these and other species contained quantities of fragmentary material too broken and corroded for satisfactory identification. As it is, 45 species of fishes were found to have eaten molluscs; half of the species had devoured three or more food-species, a third (15) had eaten two species and the rest (8) had eaten one only. The elephant fish heads the list with 15 species. Moki, which comes next with 12 species, appears to swallow much grit and shelled organisms from rocky ledges, including for example remains of chitons, so that the list is incomplete for this species. Sand flounders and blue cod each with 9 species, and red cod with 8 species are closely followed by warehou, red gurnard and girdled parrot-fish. Reconsideration of the lists in terms of the food-organisms shows that squid and octopus are easily the most frequently mentioned; they are however groups larger than single species. No less than 20 species of fish feed on both, of which 7 feed on shelled molluscs also. The remaining 13 are mostly large roving fishes which prey on other fishes, and for ecological purposes it would clearly be appropriate to group the nektonic cephalopods with fishes instead of with molluscs as sources of food. Of the shelled molluscs, Zethalia zelandica is eaten by 12 species, including benthic and roving species in about equal ratio. Great quantities are commonly eaten by pigfish and others. Similarly, Maorimactra ordinaria, the favourite food of elephant fish, is eaten by 10 species, including rovers such as warehou, tarakihi, and red gurnard, which therefore feed at times on the sea floor.

Food-Group III.—Crabs. The term “Crab” is here used in an ecological rather than a taxonomic sense, to include not only the Brachyura but also the crab-like Porcellanids of the genera Petrocheles and Petrolisthes. The distinctness of Filhol's species Petrolisthes novae-zelandiae from the common P. elongatus was verified, partly from the Portobello collections, by Bennett (1930), but at too late a stage for separate records to be quoted here. The same applies to Paramithrax, where most or all of the records refer to the common P. minor. The main taxonomic difficulties occurred among the Hymenosomidae, for although several rare species, not recognised since Filhol's description (Filhol, 1885), were secured, such as Halicarcinus edwardsi from red cod, the boundaries of the commoner species are at present too ill-defined to allow close specific distinctions; hence “Halicarcinus spp.” is a group-name for species with tridentate rostra, and “Hymenicus spp.” for those with trilobulate lamellar rostra. The name Hombronia is revived for a swimming crab hitherto regarded as rare, but now recognised to be not only generically separable from other members of the family, but ecologically of major importance. Table III.—Fishes feeding on three or more species or groups of crabs. Predators Prey Paramithrax sp. Halicarcinus spp. Hymenicus spp. Hombronia depressa Cancer novae-zelandiae Nectocarcinus antarcticus Ommatocarcinus macgillivrayi Hemigrapsus sexdentatus Hemigrapsus crenulatus Cyclograpsus lavauxi Helice crassa Petrolisthes spp. Others Total smooth-hound X X X X X X X X X X 10 spined dogfish X X X X 4 skate X X X X 4 red cod X X X X X X X X X X X Hemiplax hirtipes, Petrocheles spinosus 13 rock cod X X X X X X 6 bastard red cod X X X 3 brill X X X 3 sand flounder X X X X X 5 green-b. flounder X X X X 4 lemon sole X X X X 4 tarakihi X X X X 4 moki X X X X X 5 spotty X X X X X 5 stargazer X X X 3 blue cod X X X Petrocheles spinosus 4 Maori chief X X X Eurynolambrus australis 4 barracouta X X X 3 ling X X X X X 5 sea perch X X X X X X X X X Leptomithrax longipes 10 pigfish X X X 3 toadfish X X X 3 red gurnard X X X X X X X X Hemiplax hirtipes 9

The following were found preying on two species or groups:— Predators. Prey. mullet Nectocarcinus antarcticus Halicarcinussp. gurnard " " " Crapatalus " " Hombronia depressa bream " " Paramithrax sp. black cod " " " hake " " Cancer novae-zelandiae witch " " " " warehou " " " " groper " " " " carpet shark Hemigrapsus crenulatus " " flathead " " " " elephant fish Halicarcinus sp. Hombronia depressa thorn fish " Petrocheles spinosus common sole Hymenicus sp. Hombronia depressa suckerfish Hemigrapsus crenulatus Petrolisthes sp. Finally, the following were found to prey on only one species or group:— On Nectocarcinus antarcticus: conger eel, horse mackerel, kahawai, snapper, trumpeter, scarlet parrot-fish, banded parrotfish, kingfish. On Petrolisthes sp.: rockfish, threepenny, cockabully, red rock cod. On Cyclograpsus lavauxi: silverside. On Hombronia depressa: turbot. On Hemigrapsus sexdentatus: girdled parrot-fish. Analysis of Food-group III.—It follows that 52 species of fishes were found to have been feeding on a total of 16 species and higher groups of crabs. The chief predator is red cod, with 14 species, followed by smooth-hound and sea perch, each with 10, and red gurnard with 9. No doubt further investigations would increase the number of predators and of prey, but the list suffices to show that a considerable number of important fish depend largely on crabs for their food. Lists compiled in this way are defective in not being quantitative, but it may be said that in general fishes which eat crabs do so frequently and in great numbers; stomachs of red cod, for example, are often found crammed with crabs, suggesting that other food is neglected if crabs are available. The importance of the planktonic zoeal stages can only be presumed, and a further item must be credited to the crabs in that a number of the predators in the present list appear as prey in food-group I; that is, fishes which eat few or no crabs may feed on fishes which depend largely on crabs. The crab most abundantly consumed was Nectocarcinus antarcticus, a prolific swimming species which frequents all depths and was seen in abundance after dark at the surface. If acclimatisation of supplementary food supplies for marine fishes is mooted in other areas of similar latitude, this species is undoubtedly New Zealand's first suggestion. It is known from 37 species, including bottom fish and rovers, and is commonly consumed in great quantities.

Next on the list comes another swimming species, Hombronia depressa, known from 22 species of fishes. Other Hymenosomids prove to be of greater importance than hitherto suspected, making up in numbers what they lack in size. Cancer novae-zelandiae is a benthic crab known from 19 species, and the severe toll taken by fishes suggests a reason for the failure to establish a European congener in New Zealand waters after several years' trial at the station. Petrolisthes was found in 9 species of fishes, but not frequently and not in large numbers, so that their importance is less than might be inferred from the number of entries. Both species are seclusive, hiding under stones and clinging tightly when disturbed; and their relative infrequence in fish stomachs, in spite of their extreme abundance on shallow bottoms, confirms the inference that these cryptozoic habits, in conjunction with their flattened shape and inconspicuous bluish-grey colours, are of protective value in an environment which proves precarious for less effectively concealed species. A case in point is that of Hemigrapsus crenulatus, a very abundant species in estuaries and on reefs, more venturesome in habits, and although active and armed with massive chelae, not provided with any noteworthy means of concealment; it is eaten in great quantities, notably by red cod from the harbour, and ranks very high as a source of food for fishes. The fact that it is known from only 8 species of fishes emphasises the contrast with Petrolisthes, and shows how a list of entries could be fallaciously interpreted if not qualified by quantitative data. The suggestion that Hemigrapsus crenulatus defends itself by its activity and strength and by its strong chelae is somewhat strengthened by the relative immunity of its congener P. sexdentatus from attack, save by a few of the larger fishes. This crab is known only from four species, and is characterised by its large size, unusually formidable chelae, and bright red colour; possibly the conspicuous pigmentation, in contrast to that of its drab mud-coloured relative, serves as a warning device. It would appear that swimming habits (Hombronia, Nectocarcinus) imply a heavy toll and a dependance for racial survival on prolific breeding, as in other relatively small organisms of the nekton (sprat, pilchard, whalefeed) and of the plankton (Copepods); that relatively defenceless benthic crabs (Halicarcinus, Hymenicus, Cancer) also require in a lesser degree a high fertility in order that the species may survive; that cryptozoism and concealment colours (Petrolisthes) may have genuine protective value; that bulk and massive weapons (Leptomithrax, Jacquinotia) may give partial or even complete immunity from attack; that these defences may be effectively supplemented by warning colours (Hemigrapsus sexdentatus); and that there is a refuge from many species of predators in estuaries (Helice crassa, Hemiplax hirtipes) and at similarly higher levels on the beaches (Cyclograpsus lavauxi). Without closer discrimination of species it is difficult to determine the functional value of the supposedly protective mask of weeds (Paramithrax minor, P. longipes) or hairs (P. ursus) or sessile animals (Leptomithrax spp.) characteristic of the Majoid crabs.

Food-Group IV.—Other Crustacea. The more conspicuous Crustacea were recorded under specific or group names as in Table IV, which includes fishes found to be feeding on at least three non-brachyurous groups as defined. Table IV.—Fishes feeding on at least three species or groups of crustaceans. Predators Prey Squilla armata Whalefeed (Munida gregaria) Crayfish (Jasus lalandii) Pontophilus australis Alope palpalis Brachycarpus audouini Nyctiphanes australis Amphipods Other Isopods Calanids Other Copepods Others Total smooth-hound X X X X X X Exosphaeroma gigas 7 spined dogfish X X X X X 5 red cod X X X X X X X X X Idotea elongata 10 warehou X X X Ostracods 4 snapper X X X Ostracods 4 tarakihi X X X 3 moki X X X Hippolyte bifidirostris 4 blue cod X X X X X Note (a) 8 barracouta X X X X Euthemisto thomsoni 5 cockabully X X X 3 ling X X X X X 5 sea perch X X Palaemon affinis 3 red gurnard X X X 3 silverside X X Exosphaeroma gigas 3 seahorse X X 2 sand flounder X X X 3 (a) Also Hippolyte bifidirostris, Idotea ungulata, Phronima novae-zelandiae. The following species fed on two of these groups (specific names as in preceding table):— Predators. Prey. elephant fish whalefeed Brachycarpus sand eel Amphipods Isopods rock cod Squilla Idotea elongata common sole whalefeed Brachycarpus groper whalefeed crayfish kahawai whalefeed Nyctiphanes greenbone whalefeed Caprellids banded parrot-fish Pagurids Cirripedes rockfish Amphipods Isopods Heleogramma medium Caprellids Copepods pigfish Caprellids other Isopods

Twenty-five species of fish were found to have fed on one only species or group of crustaceans, as follows:— On whalefeed: carpet shark, skate, hake, witch, lemon sole, half-banded sea perch, trumpeter, spotty, girdled parrot-fish, southern kingfish. On Brachycarpus audouini: blind numbfish, turbot, green-back flounder, Hemerocoetes waitei. On Copepoda: minnow, ahuru. boarfish. Notothenia purpuriceps. On Paranephrops: eel, short-finned eel. On Caprellids: bellows fish. On Calanoids: pilchard. On Nyctiphanes australis: silver dory. On Alope palpalis: thornfish. On Isopods: lump fish. Analysis of Food-group IV.—Red cod head the list with 10 groups or species, but as already noticed, more stomachs of this fish were examined than of any other. Other fishes prominent in this series are blue cod (8 groups or species), smooth-hound (7), spined dogfish, barracouta, ling (5). But by far the most important conclusion is the extremely valuable role of whalefeed, of which the bottom and swimming forms, in their extraordinary though seasonal abundance, provided enormous quantities of food for at least 26 species, including 10 species not known to feed on any other nonbrachyurous crustacea. Brachycarpus audouini, eaten by 13 species, and Nyctiphanes australis, eaten by only nine, but taken in great quantities, are also of major importance as food for fishes. Closer study would no doubt add to the list of those feeding on Copepoda and other minute forms. The above records do not attempt to evaluate the part played by the larval and other planktonic forms. The abundance of whalefeed eaten by red cod must have a considerable bearing on the food value of this much-despised fish; the body-fat of red cod has been found (Carter and Malcolm, 1926, p. 648) to show a close seasonal agreement with the available supplies of whalefeed. In the swimming stages whalefeed was found to feed principally on diatoms. Food-Group V.—Annelida. As in the case of other wholly soft animals, the rapid maceration implies an incompleteness of records; but there is already ample evidence that worms are by no means a negligible fraction of the diet of some important fishes.

Table V.—Species of fishes feeding on at least three species of worms. Predators Prey Glycera ovigera Physalidonotus squamosus Nereis australis Nereis ammblyodonta Nereis vallata Timarete anchylochaetus Eunice australis Lepidonotus polychroma Eulalia microphylla Terebellids Others Total smooth-hound X X Note (a) 6 spined dogfish X X X X X 5 red cod X X X X X 5 sand flounder X X X X X X Lumbriconereis sphaerocephala, Aphrodite talpa 8 common sole X X X Gonida grahami 4 moki X X X 3 blue cod X X X X X 5 sea perch X X X 3 red gurnard X X X X 4 (a) Also Lumbriconereis sphaerocephala, Aphrodite talpa, Phycosoma annulata, Hemipodus simplex. The following cases of predation on two species were noticed:— Predators. Prey. skate Glycera ovigera Physalidonotus squamosus witch " " " " brill " " " " warehou " " " " tarakihi " " Timarete anchylochaetus greenbone " " Lepidonotus polychroma spotty Nereis amblyodonta Eulalia microphylla pigfish Harmothea praeclarox " " lemon sole Lepidonotus polychroma " " Single species were identified in the following cases:— Glycera ovigera: silverside, rock cod, stargazer, barracouta. Terebellids: yellow-belly, black flounder. Nereis amblyodonta: mullet. Eulalia microphylla: girdled parrot-fish, thornfish, suckerfish. Timarete anchylochaetus: rockfish. Finally, Nemertines were recovered from the green-back flounder.

Analysis of Food-group V.—In addition to the Terebellids and Nemertines 15 species of worms were recognised among the stomach contents. Much the most frequent was Glycera ovigera; save for Physalidonotus squamosus and Eulalia microphylla, the rest occurred in only one or a few species. It would however give a false impression of the role of Annelids as sources of food if these results were treated as approaching completeness. The fishes most dependant on worms are the smooth-hound and the sand-flounder, from each of which seven species were recognised, in addition to Terebellids in the latter fish. In general, the benthic fishes, notably the smaller Elasmobranchs and the flat fishes, are the chief users of this source of food. Food-Group VI.—Echinoderms. Echinoderms were eaten as follows:— Table VI.—Species of fish feeding on echinoderms. Predators Prey Chiradota Cucumaria Evechinus chloroticus Ophiuroids smooth-hound X elephant fish X sand eel X red cod X witch X X sand flounder X black flounder X common sole X X lemon sole X ling X Analysis of Food-group VI.—Although the echinoderm fauna is limited in species, the entries are few and the absence of asteroids especially noteworthy—the more so as they could scarcely be overlooked if present. The main conclusion is that several flatfishes and a few other bottom-feeding species ingest ophiuroids and occasionally other echinoderms. The role of echinoderm larvae in the food cycle of the plankton remains uninvestigated. Food-Group VII.—Plants. Table VII. Predators Prey Zostera marina Ulva lactuca Macrocystis pyrifera Algae various garfish X X X sand flounder X X mullet X X marblefish X X trumpeter X X scarlet parrotfish X X greenbone X blue cod X X

Analysis of Food-group VII.—Obviously no serious attempt has been made to distinguish the species of algae, and the phytoplankton remains uninvestigated. In some cases there are grounds for doubting whether the weeds are eaten for their own food-values or for that of the molluscs and crustaceans swallowed with them. Phytophagous fish are likely to have a high iodine content. Food-Group VIII.—Miscellaneous. Analysis of Food-group VIII.—Scyphozoans were recovered from elephant fish, actinozoans from blue and black cod, and hydrozoans from moki, red cod and marble fish. These coelenterates are of minimal nutritional value, and little more could be claimed on behalf of polyzoans (marble fish), salps (elephant fish, warehou, trevally) or simple ascidians (red cod) or compound ascidians (bream, trigger fish). Flotsam in the form of petrels was found in the stomachs of groper, ling, and red cod, and to the scavenging habits of the last were further demonstrated by the discovery of various mammals. Summary. The smooth-hound is exonerated from the current lay accusation of scavenging propensities. From the above tables it will be seen that, with–few exceptions, the food of larval fishes has been omitted; this is due to the almost impossible task of identifying the microscopic forms found therein. Only macro-organisms have been identified, analysed, and included in the foregoing classification, but this has been regarded as a necessary preliminary to the fuller investigation which should eventually include a study of the micro-organic food supply as well. Acknowledgments. From the nature and scope of the work, it is clear that such a paper would not have been possible were it not for valuable help from many sources. Specialists who have named material belonging to various groups include Dr. W. B. Benham (Annelida), Dr. E. W. Bennett (Echinoderma, Brachyura), Dr. H. J. Finlay (Mollusca), Dr. C. E. Laws (Mollusca), Mr. A. W. B. Powell (Mollusca), and Mr. R. M. Laing (Algae). For the correct usage of the data supplied the author must accept full responsibility. Finally, I wish to thank Dr. E. W. Bennett for helpful assistance in the preparation of this paper. References. Bennett, E. W., 1930. Notes on New Zealand Brachyura and related Crustaceans, Rec. Cant. Mus., vol. 3, no. 4, pp. 255–261. Carter, C. L., and Malcolm, J., 1926. Food values of New Zealand Fish, Trans. N.Z. Inst., vol. 56, pp. 647–650. Filhot, H., 1885. Mission de l'Ile Campbell. Graham, D. H., 1938. Fishes of Otago Harbour and Adjacent Sea with additions to previous records, Trans. Roy. Soc. of N.Z., vol. 68, pp. 399–419.