66 NATIVE AND INTRODUCED SPECIES IN NEW ZEALAND

native species on various occasions; and thefood of both native morepork and intro-duced little owl includes small birds, eithernative or introduced.

REFERENCES

BULLER, W., 1905. Supplement to Birds of Ncw Zealand.,

The Author, London.

CUMBERLAND,K. B., 1949. This is New Zealand. Whit~combe & Tombs, ChristChurch.

Gum, J. A., 1961. Ecology of the birds of KaingaroaForest. Proc. N.Z. Eco/. Soc. 8: 29-38.

GUTHRIE-SMITH, H., 1921 and later editions. Tlaira.Blackwood, London.

LA<:K, D., 1933. Habitat selection in birds. with specialreference to the effects of afforestation on the Rreck-land avifauna. J. Anim. Ecol. 2: 239-262.

LACK,D., 1939. Further changes in the Breckland avi~fauna caused by afforestation. J. Anim. Ecof. 8:277-285.

RELATIONS BETWEEN SALMONIDAE THE

FRESHWATER FAUNA

K. RADWAY

MARPLES, B. L 1942. A study of the little owl, Alhenenoctua, in New Zealand. Tran.~. Rov. Soc. N.Z. 72:

-237-252.

MASTERS, S. E., HOLLOWAY,J. T. and McKELVEY, P. 1.,1957. The national forest survey of New Zealand,1955. VoL I. Govt. Printer, Wel1ington.

MOON, G. J. H., 1956. White-faced heron nesting inNorth Auckland. Notornis 6: 244.

MYERS,J. G., 1923. The present position of the endemicbirds of New Zealand. NZ. Journ. Sci. Tech. 6: 65-99.

THOM'SON,G. M., 1922. The lIafllralisation of animahand plants in New Zealand. Cambridge UniversityPress.

TURBon, E. G., 1953. Notes on the occurrence of thebellbird in North Auckland. Notornis 5: 175-17&.

TURBOIT,E. G., 1957. Native and introduced birds. InScience in New Zealand. Reed, Wellington.

WODZICKI, K. A., 1950. Introduced mammals of NewZealand: an ecological and economic survey. Bull.Dept. ScL Ind. Re,~. 9&.

AND NA TIVE

IN NEW ZEALAND

ALLEN

Fisheries Laboratory, N.Z. Marine

Department

The freshwaters of New Zealand providean environment into which relatively fewexotic species have penetra ted, partly be-cause the problems of survival duringtransportation are so difficult for a fresh-water animal as to make accidental immi-gration very unlikely, and partly because themotives which may stimulate deliberateimportation by man are so few in this case.So much remains to be done in the studvof our freshwater fauna that a number ofexotic species of the smaller invertebratesmay exist unrecorded in this country, butthe known freshwater fauna of non-nativeorigin is limited to one snail and some dozenspecies of fish all deliberately introduced.

Most of the fish were introduced to providesport, and it is not surprising that a largeproportion of them belong to the Salmoni-

dae since in its native Northern Hemispherethis family contains the majority of the mosthighly regarded game species. Of the speciesthat were introduced, two have establishedthemselves successfully over a large part ofthe country, and a third in a more restrictedarea. These are the brown and rainbowtrouts (Salina trutta and S. gairdnerii) andthe quinn at salmon (Oncorhynchus tscha-wytcha) respectively. A few other speciesexist in small numbers in limited areas. Forsimplicity, I propose to limit considerationof the introduced freshwater fauna largelyto these few highly successful species ofSalmonidae; both because of their wide-spread abundance and because, on accountof their popular esteem, they have beenrelatively intensively studied.

It should be realised that fresh waters

ALLEN: SAI"lvl0NIDAE AND FRESH\VATER FAUNA

do not constitute an environment entirelyisolated from those which surround it. Manvanimals of a variety of groups divide theli-time between freshwater and either landor air, either intermittently, or at set stagesof their life history. It is therefore necessarvto refer occasionally to events occurring out-side the limits of the watery environment.

Since it is almost axiomatic that gamefish should be carnivorous, it follows thatone of the principal impacts of the Salmoni-dae on the native freshwater fauna is theirpredation on small native animals, bothvertebrate and invertebrate. The converseis, of course, also true; that the nativefauna may, by variations in its compositionand abundance, play an important part in.determining the density and structure of theintroduced fish populations.

The principal food of the Salmonidae inalmost all waters is derived from the inverte-brate bottom fauna and from small fish.The degree to which the various species areeaten depends upon their abundance andupon the extent to which their habits andstructure make them available to and eat-able by the fish (Allen 1941). The latter mav,in turn, be considerably affected by the sizeof the fish themselves; the larger the fish,the larger the animals which it eats, and themore easily are hard-shelled forms con-sumed. Among the invertebrates the princi-pal groups eaten by stream-living Salmoni-dae in almost all countries are Chironomidae,Ephemeroptera, Trichoptera and Mollusca,the fish tending to advance through a sequencein this order as they increase in size. In lalresconsiderably greater diversity occurs. Smallfish are mainly eaten by the larger salmonids.the particular species consumed varyingaccording to the composition of the localfauna; in New Zealand the forms mainlyeaten are bullies (Gobiomorphus and Philyp-nodon spp.) and smelts (Retropinna spp.).

The native invertebrate fauna of most ofour rivers and streams is generally of aboutthe same level of abundance as that of thcnorth temperate region; for instance, therewould commonly be between 100 and 1000animals over about 3mm. in length for eachsquare foot of stream bed. This fauna isclearly sufficiently abundant to support goodtrout populations in waters where condi-tions are otherwise suitable. Similarlv man"., .,

67

of our lakes contain numerous small nativefish which provide good trout food; thelarge populations of fast-growing trout inmany of the lakes of the Rotorua districtare, for instance, enabled to develop by the

abundant smelt stocks in these lakes.

While it is thus a simple matter to declarethat New Zealand freshwaters now containa native fauna of small animals which isgenerally suflicient to support strong popu-lations of introduced salmonids, it is muchless simple to determine what effect, if anv,the salmonids have had upon the originalfauna before the present position wasreached. The difficulty arises partly from thealmost complete lack of observations com-paring the fauna of waters before and aftcrthe introduction of salmonids, partly fromthe scarcity at the present time of reallycomparable waters, some of which containtrout and some of which do not, and partlvfrom the other environmental changes whichhave taken place simultaneously with theestablishment of salmon ids in New Zealand.These environmental changes include bothphysical changes like the increased instabil-ity and destruction of pools which oftenfollow the clearing of catchment areas, andbiological changes such as the introductionof insectivorous birds which may prey onthe winged stages of insects with aquaticlarvae.

Workers on various native aquatic animalshave realised that the species they wercinterested in were subject to predation bytrout and have suggested that this mightthreaten their survival. Generally howeverlittle evidence has been adduced as to theactual extent of the threat and one cannotavoid the feelino; that sentiment has playedan undue part in the formation of some ofthe opinions expressed. In 1915 a paperdealing with the freshwater crayfish (Paran-ephrops spP.) declared "It is evident that jncommon with other members of our nativefreshwater fauna, the crayfish are beingdestroyed by eels and the introduced fishes.Trout

-

have been caught with partiallydigested crayfish in their stomachs, and it isprobable that the restriction of crayfish tosuch places as cannot be inhabited bv thesefish is only a matter of time". Now, 45 yearslater. there are still lakes and rivers in whichcrayfish are abundant despite the presence

68 NATIVE ANn INTRODUCED SPECIES IN NEW ZEALAND

,

of numerous trout, and there is little doubtthat here, as in the northern hemisphere,trout and crayfish can exist side by side in abalanced relationship.

Among the mayflies there are some forms,particularly the various species of Deleati-dium, which can maintain themselves iuvery large numbers despite the fact thatthey are freely eaten by trout. In the Horo-kiwi Stream, for instance, in 1940 theygenerally occurred at a density of 50 to 100per sq. ft. despite the fact that they madeup 60-90 per cent. of the food of trout intheir first nine months. It has been suggestedhowever that other mayflies, and particu-larly Oniscigaster, are more vulnerable totrout and that this may ultimately causetheir extinction (Till yard 1926). Somedirect evidence regarding reduction in thenumbers of O. wakefieldi is given by Hudson(1904), who quotes Hutton as saying "In1874 the insect was common in the neigh-bourhood of Christchurch. I have lived thercduring the last] 9 years without seeing asingle specimen. Whether they have beenkilled off by the trout or by the sparrowsI cannot say." Speaking of the related O.dislans, Hudson himself seems more surethat the insect is destined for extinctionthan of the cause which will bring it about.He says of the nymph, "It is not very activc,and owing to its large size, is easily seenand captured by means of the hand alone.Hence its speedy extinction by trout appearsinevitable". However, when speaking of theimago, he writes, "Whilst thus resting it isno doubt often devoured by birds and theinsect's early extinction through this causealone would appear probable". Whatevertheir numbers may have been originally themayflies of this genus are now far fromcommon in New Zealand, although stillwidely dispersed. If the numbers have indeeddiminished severely, introduced fish andbirds may have been at least part of thecause, but physical changes may have playeda part. The nymphs of this genus are activeswimmers and also rest in exposed positionson the tops of stones. They are thus ill-adapted to withstand strong water-currentsand normally live in sheltered pools andbackwaters. Where, as has occurred in manyplaces, man's activities have reduced thestability of river beds and diminished the

amount of pools and other shelter, the cxtentof possible habitat for this genus has prob-ably been drastically reduced. It is interest-ing that one of the few localities where itoccurs regularly near Wellington is at themouth of a tributarv stream where in timesof flood quiet conditions due to backing upfrom the parent river occur instead of theusual scouring.

An opportunity has arisen for studyingthe effects of the introduction of rainbowtrout in L. Waingata in Northland. Thereis some evidence, for instance, that the fresh-water crab, Hymenosoma lacustris, whichwas previously fairly common in the lake,may have diminished considerably in num-bers since the introduction of trout.

Work has recently started on certain tribu-taries of the Wangaehu River which trouthave been unable to reach because of thepermanently toxic conditions in the mainriver. There is considerable variation in thebottom faunas among the troutless streamsthemselves, apparently due to environmentalconditions, but a recent comparison of theirfaunas with that of another tributary with anabundant stock of rainbow trout supportsthe following general conclusions.

1. The stream containing trout has gen-erally a greater number of individualanimals to the square foot than thetroutless streams.

2. In the stream with trout the propor-tion of the fauna made up of animalsfreely eaten by trout is at least asgreat as in the troutless streams.

3. The troutless streams apparently con-tain no species which do not com-monly occur in streams containingtrout.

Thus, here, at any rate, the introductionof trout seems to have made no substantialdifference to the composition of the bottomfauna.

The general conclusion is therefore thatthe introduction of Salmonidae has probablynot led to any great change in the fauna

ALLEN: SALMONIIJAlj AND

of New Zealand freshwaters, particularlywhere the bottom fauna is concerned. It ishowever possible that a few species havebeen adversely affected; these are likely tobe forms which combine habits and appear-ance making them attractive to trout, witha low biological potential giving them littlecapacity to withstand increased predation.This conclusion might have been unexpectedif New Zealand fresh waters had been origin-ally without carnivorous fish, but in mostof our waters the bottom fauna was in factalready subject to considerable predationfrom a fish population including, to varyingextents in different localities, the smallereels, the larger bullies and galaxiids, and thegrayling (Prototroctes oxyrhynchus). Thevirtually complete disappearance of graylingbetween 1870 and 1925, from previous greatabundance, is, incidentally, by far the moststriking biological change which is knownto have occurred in New Zealand waters.Since it has happened in some places wheretrout are still scarce or absent, and in othershad begun before trout arrived, it does notseem that, whatever its cause was, it canbe laid at the door of the salmonids.

The reverse relationship, predation onSalmonidae by native animals, may be morebriefly considered. The two principal groupsof predators are shags, particularly the blackshag (Phalacrocorax carbo), and eels, espe-cially the long-finned eel (Anguilla dieffen-bachi). Although both species are undoubt-edly predatory on trout to some extent, theexisting trout populations were establishedin the face of this predation, and it seemsunlikely that the native predators will seri-ously menace their continued survival.

Work in progress on the effect on troutpopulations of the removal of eels is stillin an early stage, but results suggest that insome circumstances it may lead to a sub-stantial increase in the numbers of trout.Areas in New Zealand which are perman-entlv eel-less, as a result of natural barriersto ihe upstream migration, tend to haveabundant trout stocks. The structure of thetrout population in these circumstances maydepend on the balance between food supplyand spawning grounds. Where the feedinggrounds are extensive in relation to thespawning areas, trout may be large as well

FRESflWATER FAUNA 69

as numerous, as in Lakes Taupo and Roto-rua; but where there is abundant spawningground in relation to the food supply thetrout may only reach a small size; thisappears to occur, for instance, in the upperWaihou River system. Whether the intro-duction of salmonids has affected the eelpopulation of New Zealand is hard to deter-mine. Although they have provided an addi-tional source of food for the larger eels,over about 26in. in length, they may becompeting to some extent with the smallereels which feed mainly on the bottom fauna.Burnet (1959), however, has found thatfluctuations in numbers of trout sufficient toaffect their growth rate through intraspecificcompetition for food, have no measurableeffect on the growth rate of small eelsinhabiting the same waters. The introduc-tion of trout has also led to some deliberatedestruction of eels with the object of bene-fiting the trout stocks; it is doubtful, how-ever, whether this has had any significanteffect on eel populations except occasionallyin very restricted areas. In any case, destruc-tion of eels for this purpose may have merelyreplaced the taking of eels for food byMaoris in pre-European times. It is morelikelv that the numbers of eels have beenreduced by the destruction of cover, sincethis is a very important factor controllingeel population (Burnet 1952).

The relationship between the shags andtrout is complicated by the fact that, unlikeeels, shags are able to move freely andrapidly between one body of water and an-other, and indeed between inland watersand the sea. Shags are also to some extentgregarious, and quite large bands of themsometimes descend on a particular lengthof river, fish it industriously for a time, andthen depart elsewhere. Occurrences of thiskind may have a considerable effect on thetrout population of the water concerned,provided that other prey is not available.Where, however, other less active and pre-sumably more easily caught fish are present,as for instance the goldfish in Lake Rotorua,very few trout seem to be eaten (Dickinson1951). Shags probably have their greatesteffect on trout in areas, such as parts of theSouth Island, where other species of fish arerelatively scarce in the rivers. Destructionfor the intended benefit of the trout has

70 NATIVE AND INTRODUCED SPECIES IN NEW ZEALAND

probably been more effective against shagsthan against eels, and in some areas fewinland. nesting sites are allowed to remainoccupied for long. The coastal colonies,however, which are largely untouched, prob-ably provide a reservoir from which theinland populations can be quickly rebuiltwhen control measures are relaxed.

REFERENCES

ALLEN, K. R., 1941. Comparison of bottom faunas assources of available fish food. Tram. Amer. Fish.Soc. 7\: 275-283.

BURNET,A. M. R., 1952. Studies on the ecology of theNew Zealand long~finned eel, Anguilla diefJmbachiiGrey. Ailsf, J. Mar. Freshw. Res. 1: 32-63.

BURNET,A. M. R., 1959. Some observations on naturalfluctuations of trout population numbers. N.Z. J.Sci. 2: 4\0-421.

DICKINSON, p" 1951. Stomach contents of New Zealandinland shags. Alut. J, Mar. Freshw. Res. 2: 245-253.

HUDSON, G. Y., 1904. New Zealand Neuroptera. West,Newman & Co., London.

TILLYARO,R. 1., 1926. The insects of Australia and NewZealand. Angus & Robertson Ltd., Sydney.