In the Glentui Iver, Canterburlrr New Zealand

Fisheries Research Bulletin No. 17

Ag., Growth, and Condition of the Common River Galaxias, Galaxias vulgafis Stokell, in the Glentui iver, Canterburlrr New Zealand

by P. L. Cadwallader

Fisheries Research Division New Zealand Ministry of Agriculture and Fisheries Ag., Growth, and Conclition of the Common River Galaxias, Galaxias vulgaris Stokell, in the Glentui River, Canterbury, New Zealand [4. M. R. Burnet photograph

Frontispiece: The common river galaxias, Galaxias vulgaris, a fish which occurs in the rivers and streams of the South Island of New Zealand, mainly to the east of the southern Alps. Fisheries Research Bulletin No. r7

Ag", Growth, and Condition of the Common River Galaxias, Galaxias vulgoris Stokell, in the Glentui River, Canterbu ry, New Zealand

b;, P. L. Cadwallader,

Department of Zoology, University of CanterburY*

I Present address: Fisheries and Wildlite Division, Ministry for Conservat.ion, Snobs Creek Freshwater Fisheries Research Station and Hatchery, Private Bag zo, Alexandra, Victoria 37r4, Australia

Fisheries Research Division New Zealand Ministry of Agriculture and Fisheries t9i8 Published by the New Zealand Ministry of Agriculture and Fisheries Wellington 1978

rssN 0110_1749 FOREWORD

ONE of the aims of Fisheries Research Division has been to build up a background of information on the indigenous freshwater fishes of New Zealan¡J.

It is therefore a pleasure to acknowledge the work done by Phillip Cadwallader, which represents another worth-while advance in our under- standing of one of these little-known species.

G. DUNCAN WAUGH, Director, Fisheries Research Division. CONTENTS

Page

1l 11 ll Biological Characteristics 72

GENERAT METHODS 13

Sampling Programme . . 13 Electric Fishing l3 Measurement of Length and Weight t4 Sex Determination 14

AGE DETERMINATION. 15 l5 Length-frequency Analysis 20

GROWTH 2l Annual Growth in Length Seasonal Growth in Length Seasonal Growth in Weight

LENGTH_WEIGHT RELATIONSHIP AND CONDITION . 26

FACTORS AFFECTING THE GROWTH OF G. VULGARIS 28

ACKNOWLEDGMENTS . . 30

REFERENCES .. 31 FIGURES

Page

1. The study area 10

2. Temperature and rainfall recorded in the study area from June 1970 to May 1971 l1

3. Otoliths of G. vulgaris t6

4. The regression of otolith radius on total length ancl correspondence of annuli in male and female G. vulgaris 17

5 Time of annulus formation in the otoliths of G. vulgaris l8

6. Length-frequency distributions of G. vulgaris from June 1970 to May 1971 .. 19

7. Length-frequency distribution of G. vulgaris in the Glentui River in July 1970, with the length-frequency distributions ofthe component age groups as determined from otoliths . . 20

8. T=ength-frequency distributions of samples of G. vulgaris taken in different parts of the Glentui River in 1971 . . 22

9. Seasonal growth in weight of male and female G. vulgarisfrom June 1970 to May 1971 24 and 25

10. Two-monthly variation in the length-weight relationship of G. vulgaris . . . . 27 ll. Two-monthly variation in the length-weight relationship and the length-somatic weight relationship for hypothetical male and female G. vulgarii of 100 mm TL 27

TÄ.BLES

Page

1. Regular samples (and subsamples) of G. vulgaris taken from the Glentui River . . . . 13

2. Mean, annual, back-calculated lengths of G. vulgaris derived from otoliths, with their 9ílconfrdencelimits .. .. 2l

3 Mean lengths of G. vulgaris in the Glentui River from June 1970 to May l97l . . 2l

4 Mean_lengths-and 951 confidence limits of age 0* G. vulgaris caught during the same periods in different paris of the Glentui River -

5 Length-weight relationships oî G. vulgaris from June 1970 to May 1971 26 INTRODUCTIOI\

The family Galaxiidae is widespread in the southern This bulletin presents inlbrmation on some basic temperate region, with species occurring in New population parameters (age, growth, and condition) Zealand, Australia, South America, and South Africa of the common river galaxias, Galaxias vulgaris and also on some islands near these land masses Stokell, a fish which is restricted to the South Island the (Stokell 1953, Darlingtonl95T, McDowall 1970). The of New Zealand and which is found in most of major river basins to the east of the Southern Alps taxonomy of the group has been the subject of much and also in the Upper Buller River System, to the confusion. However, the work of Stokell (1938, 1945, west of the Alps. It occurs beneath and between (1967, 1949, 1959) and McDowall 1969, 1910, 1972) boulders in fast or broken water of rivers and streams, has greatly clarified the systematics of the New but it is not usually found in streams entering lakes Zealand fauna. Thirteen species are now recognised (Stokell 1949, McDowall 1970). in New Zealand (McDowall 1970, l9l2) and are It was first described by Stokell (1949), who also divided into two genera-Galax¡as with l0 species (Stokell 1959) described Galaxias anomalus, which has and Neochannø wilh 3 species. Apart from Galqxias since been synonymised with G. vulgaris by McDowall maculatus (Jenyns), which forms the basis of a com- (1970). Benzie (1968) compared its life history with mercial and sport fishery (Hopkins and McDowall that of G. maculatus, with particular emphasis on 1970), little attention has been given to the ecology embryological development, growth rates, and of the New Zealand galaxiids. breeding biology. lDepartment of Lands and Survey photograph.

Fig. I : Aerial view of the study osition of the waterfall furthest down stream (height is 2.5 m). C_D: Sectiolof the river, atout 1. were taken to obtain data on age, growttr, aÀd óndition ä C.rit- garis. E-F: Section of the rive were taken in March, April, anã liay 197'1, H: Bald Hills Stream, the main tributary of the Gle

10 THE STUDY AREA

The investigation was carried out in the Glentui boulders up to 0.7 m in diameter in rifles. There is River in the provincial district of Canterbury, New fìne mud in quiet backwaters and deep pools. Water Zealand. The Glentui arises at an altitude of about velocity (measured with a Gurley No. 625 Pygmy 840 m and flows into the Ashley River at a point velocity gauge) varied from 0 m/s at the bottom of (43" 14' S, 172' 18' E) 180 m above sea level. It the deep pools to a mean of 0.4 m/s in riffies. Surl'ace originates in a forest of mountain beech, Nothofagus water was in the main channel throughout the inves- solandri (Hooker), and in its lower reaches flows tigation, though in the autumn of l97l there was no through cultivated farm land. Its largest tributary is surface water in several other local rivers, including the Bald Hills Stream, which enters the river at an parts of the Ashley River. Although the Glentui is altitude of 300 m. A number of smaller tributaries more stable than many other similar rivers in Canter- carry run-off from the surrounding hills (Fig. 1). bury, severe floods after heavy rainfall often cause the water velocity in the narrowest rifles to exceed 3 m/s. Such flooding precludes the establishment of macro- LOCAL CLIMATE phytes in the main channel and has a pronounced effect on the character of the river bed. Apart from The prevailing winds are from the north-west and isolated stands of Myriophyllum it quieter back- are fairly warm. Air temperatures jn 1970 ranged from a summer maximum of 34.4'c to a winter minimum of -3.9oc, and similarly in 1971 the range A v/as between 33.5" and -4'c. Mean, maximum, and minimum air temperatures from June 1970 to May l97l are presented (A, Fig. 2).

Mean annual rainfall (1947-70) for the area is 1080 mm and the period of maximum rainfall is rJ tl usually in spring and early summer, from October to o January (data from New Zealand Meteorological Ð rl o Service). Rainfall from June 1970 to May l91l (¡, o- totalled 880 mm (C, Fig. 2). E É I

I THE GLENTUI RIVER tll The distance between the origin of the Glentui ïr River and its point of entry into the Ashley tlil I River is about 9.5 km. Near its origin there is a series of îlll waterfalls ranging in height from 2.5 ro 24m. Sampling was conûned to the stretch of the river below the waterfalls, a section which consisted of riffies interspersed with quiet stretches.

There was great variation in both depth and width E of the river within the sampling area. Approximations E based on measurements taken throughout the sam- o .s pling period showed that water depth ranged from o É. 20-30 mm in riffies to 0.7 m in pools formed behind obstacles such as fallen trees, and width varied from 1to9m. DIJ The geological deposits over which the river flows 1970 Time (month:) 1971 are of two main types. Above an altitude of 305 m they are strongly indurated and consist mostly of Fig. 2: Temperature and rainfall recorded in the study alea graded-bedded greywacke and argillite. Below 305 m from June 1970 to May 1971. A: Mean monthly air the deposits consist mainly of glacial outwash gravels temperatures. B: Mean monthly water temperatures. ln (Gregg both A and B the temperature range is indicated by vertical 1964). bars. C: Monthly rainfall (graph), compared with mean monthly rainfall for the 25-year period from 1946 fo 1970 The river bed is composed of fine gravel, with large (histogram).

Inset 2 waters, the only plant cover in the river is that pro- The invertebrate fauna was dominated by larval vided by fallen trees and other debris of terrestrial insects, which included species of Ephemeroptera, origin carried down by floods. Trichoptera, Diptera, Plecoptera, Megaloptera, Col- eoptera, and Hemiptera (Cadwallader 1975a, 1975c). study pH water regular During the tile of in the Other invertebrates were Gordius (Nematomorpha) sampling area varied 6.8 7.3. The water was from to and the gastropod Potamopyrgus antipodarum (Gray). fairly soft, alkalinity ranging from 4l .9 to 62.0 mg of Terrestrial arthropods and plant debris were fre- CaCO to Conductivity at 25"c varied from 128 quently found in the drift, and terrestrial lumbricids 150 "/1. moderate the Canterbury ¡rS lcm, a value for (Oligochaeta) were sometimes found in shallow, quiet region. rvater at the river margins. Water temperatures were taken with a maximum- Apart from Galaxias vulgaris, other fish in the from minimum thermometer at monthly intervals sampling area were the upland bully (Gobiontorphus June 1970 to 1971 (8, Fig. 2). All readings were May breviceps (Stokell)), long-finned eel (Anguilla dieffen- made the same riffie after the thermometer had in bachii Gray), short-finned eel (Anguilla australis position the 24 hours. been in on river bed for Ice schmidtii Phillipps), and brown fiout (Salmo trutta formed on standing at the sides of the river for water Linnaeus). short periods in the winters of 1970 and 1971, but did not occur in the main channel. Black shags, Phaløcrocorax carbo (Linnaeus), were occasionally seen near the river. These birds take galaxiids, eleotrids, salmonids, and anguillids in New BIOLOGICAL CHARÄCTERISTICS Zealand inland v/aters (Falla and Stokell 1945, Diatoms, notably Gomphonema, were on most of Dickinson 1951, Boud and Eldon 1960, Duncan 1968), the boulders on the river bed. Myriophyllum was the but they were not considered to be important predators only macrophyte found in the study area. of fish in the Glentui River.

12 GENERAT METHODS

SAMPLING PROGRAMME E300 generator via a l2-Y battery. Positive pulses were passed to the water through an electrode held The sampling programme was part of a more by the operator. An earth return system consisting extensive plan aimed at studying not only age, growth, mainly of a long, flexible metal cord completed the and condition of G. vulgaris, but also breeding circuit. The current loading cottld be altered to four biology, home range and movement, food and feeding settings with maxima of 0.22, 0.4, and 0'5 A and a habits, and interrelations with other frsh in the river. setting which could not be overloaded. The output Only that part of the programme concerned with the was a positive square wave pulse at about 100 c/s elucidation of basic population parameters is outlined with peak voltages of300, 150, 100, or 50 V as required. here. A second pulse with an 8O/" on duty cycle at about 3 c/s was imposed on this (Woods 1967). Regular samples of G.'vulgar¡s were taken from a 1.5-km stretch of the river (C-D in Fig. 1) every The use of electric fishing equipment, briefly 28 days from June 1970 to May 1971. A1l fish in these reviewed by Hynes (1970), is one of the least selective samples were measured to provide monthly informa- of fishing methods (Lagler 1968, Johnson, Rinne, tion on growth in length. A stratified subsample of and Minckley 1970). However, several workers, for at least two fish of each sex in each 5-mm length class example, Saunders and Smith (1954) and Woods was then taken and preserved in l0l formalin, the (1964), have shown that there is usually some bias remainder being returned to the river. The sub- towards the larger length classes. sampled fish were later weighed and aged and their fat deposits estimated. The equipment was highly effective in the Glentui River, where the water was generally shallow and of Table I shows the number of fish in each sample moderate conductivity and the substrate of gravel ancl subsample taken in the study; it includes only and boulders was of high resistance. Its effectiveness those flsh caught in the normal adult habitat, that is, decreased in deep pools. The radius of the effective riffies, and does not include age 0f fish less than field rvith the positive electrode at its centre was about 30 mm long which occurred in quieter parts of the 300 mm for fish 100 mm long. Under these conditions river. Information on growth of recently hatched all length classes of fish were sampled adequately fish was obtained from further samples taken during except recently hatched G. vulgaris, wliich inhabited and after the breeding season in 1910-71. the still, deep parts of the river. However, these fish were taken easily with dip nets.

ELECTRIC FISHING Mortality caused by electric fishing was negligible. That which did occur was because of overexposure Most sanples were obtained with portable electric to the electric field as in fish caught between boulders fishing equipment. Power was supplied by a Honda and not immecliately seen by the operator. Towards the end of the sampling period some G. vulgaris (fewer than 20) showed redclening of the caudal TABLE 1: Regular samples (and subsamples) of C. vulgaris taken from the Glentui River peduncle. This condition was caused by the rupture of blood vessels and is related to the type of pulse No. of fish Sample produced by the electric fishing equipment. Hauck 1970 (1949) reported a similar condition in rainbow trout 1 27,28 June 89 (30) 2 25,26 July 139 (13e) (Salmo gairdneri Richardson) that were shocked August (5s) 3 22 5s 110 alternating current. Spencer (1961), 4 19,20 SePtember 100 (39) with V 5* 17 October 24 (24) working with bluegill (Lepomis macrochirus P.afr- (51) 6 14, 15 November 1'7 | nesque), chanrrel catfish (Ictalurus pr'tnctatus (Rafr- 7 12, 13 December t'7s (3r) nesque)), and large-mouth bass (Miuopterus salmoides 191 | (Lacépède)) exposed to 230V alternating current' (24) 8 16. 17 January 146 and dislocations ofthe vertebral 9 6, 7 February 128 (50) showed that fractures 10 6, 7 March 176 (21\ column occurred in the caudal region. Lalge fish 3, 4 April 262 (lil) 11 susceptible this type of damage than 12 l,2 May 180 (2e) are more to l3 29, 30 May 163 (2s) fish of the size of the G. vulgaris taken in this study, long' *Unfavourable fishing conditions after heavy rainfall the largest of which was only 125 mm

fnset 2+ MEASUREMENT OF'LENGTH AND \ryEIGHT a dissecting microscope at low magnification ( x 6.3). During this period the anatomy the genital After capture and before further handling alt fish of region differs in males and females. Males have papilla, were anaesthetised with benzocaine. Total length a at the tip of which is the genital (TL), that is, length from tip of snout to the distal opening. Females lack a papilla, and the genital region is much more bulbous end of the central rays of the caudal fin, was read to than in males. The male papilla 0.1 nm with a measuring board fitted with a vernier can be made to protrude by flexing the body. scale (Woods 1968). Wet weight of both fresh and formalin-preserved fish was measured to the nearest This method of sex determination unreliable milligram. Correction factors for the effect of formalin was during the resting and early-ripening on length and weight of G. vulgar¡s were obtained stages of the reproduciive cycle, from December from Cadwallader (1974). Throughout this bulletin to April. At this time the female genital region is less bulbous length refers to total length unless otherwise indicated_ and the male papilla is less distended and cannot always be made to protrude. Sex determination of age 0* SEX DETERMINATION fish by this method was unreliable, because, though mature males were easily identified, immature males From May to November, during the ripening, could not be distinguished from females. Fish which spawning, and post-spawning stages of the reproduc- could not be sexed externally were sexed by examina- tive cycle, adult G. vulgaris were sexed externally with tion of the gonads after dissection.

l4 AGE DETERMINATION

Methods for the determination of age in fish are result from non-periodic variation in growth, that is, numerous, and the extensive literature has been re- secondary rings. Rings were considered to be annuli viewed by Chugunova (1959) and Tesch (1968). on the criteria that they \vere more distinct, were Benzie (1961) describecl the otoliths of Gahxias mqcu' uniformly spaced, and extended right around the latus.McDowall (1968) found that otoliths and length- otolith (Staples 1971). The consistency ofrecognition frequency analysis could not be used to age G. of annuli was tested in the July 1970 sample of 139 maculatus because of its prolonged breeding season fish. In this sample the number of annuli in each from September until June. Pollard (1971), working otolith was counted twice, the second reading being with a landlocked population of G. maculalus, used 6 months after the trst. Out of 139 otoliths, all but 3 length-frequency analysis with some success, but found were read the same on both occasions-a consistency otoliths of little use as an aid to ageing the flsh. of 97.8)(. The criteria for distinguishing annuli were Hopkins (1971) used both length-frequency analysis therefore considered to be adequate. Only annuli were and otoliths to age Galaxias divergens Stokell, but did used to determine age, and the number in each otolith not validate the use of otoliths for this species. In this was counted. study sagittal otoliths and length-frequency analysis rvere used in combination to age G. vulgaris. In some otoliths two translucent zones occurred close together. On the criterion of spacing they were consiclered as representing one annulus. Secondary rings were more readily seen in the thin otoliths of OTOLITHS young fish. Typically, there were two of these narrow The use of otoliths to age fish depends on changes translucent rings. One, surrounding the central nucleus, in the rate of growth or metabolism, which are reflect- was formed at total lengths ranging from 12 to 17 mm ed in the otoliths as alternating bands of visibly and coincided with the end of larval life. The other clifferent material. The sagittal otoliths of G. 'tulgaris ring formed at total lengths of from 30 to 50 mm and are plano-convex and basically discoidal, with a coincided with the transition from the juvenile to the forward projection at the antero-ventral corner. This adult way of life. These rings were considered not to projection is not well developed in small otoliths, but be annuli on the fi.rst criterion mentioned above. becomes more prominent when the otoliths increase Similarly, Gambell and Messtorff (1964), working in size. with whiting, Merlangius merlangus Linnaeus, considered the nuclear edge to reflect a change from Both sagittal otoliths were removed from frsh sub- pelagic to benthic habits rather than to be an sampled from regular monthly samples and from indicator of age. recently hatched fish sampled after the breeding season from September 1970 to January 1971 and these The sagittal otolith of one side of the head was the were stored dry in tube vials. mirror image of that on the other side in all fish examined, except two, in both of which one otolith was normal and the other was underdeveloped. In one Interpretation of Otoliths instance the nucleus was normally developed, but the transparent. Otoliths were read under a microscope at x 80 mag- rest of the otolith was extremely thin and nification; they were immersed in xylol and viewed In the other instance the abnormal otolith, though with reflected light against a black background. With showing annuli, was of a more hyaline nature than is this method a number of opaque (light) and trans- normal. It has been shown that inorganic material, lucent (dark) zones were visible in the otoliths (Fig. 3). chiefly in the form of aragonite, is laid down through- The outer border of each translucent zone forms a out the otolith, and organic rnaterial is restricted to distinct ring where it abuts the next outer opaque the more opaque parts (Irie 1955, Dannevig 1956, zone. Since the edges of the otolith are thinner than Mina 1968, Degens, Deuser, and Haedrich 1969). It there was a the central nucleus, they often appear translucent and appears that in the two abnormal otoliths this must be borne in mind in the interpretation of breakdown in the processes governing the deposition otolith zones. of organic material.

It is important to distinguish between rings laid Although both otoliths were generally similar, to down on a regular temporal basis, that is, primary standardise procedure all measurements were made rings or annuli (4. C. Jensen 1965), and those which on the flat side of the left otolith.

l5 Validation of the Otolith Method a year, as described by Matsuura (1961), Mio (1961), For annuli to be used to age G. vulgaris it had to be and Yunokawa (1961). The otolith radius (,R) and the shown that they were formed at regular time intervals. radius of each annulus (r") were measured from the centre of the nucleus along the dorso-ventral axis (see The pattern of annulus formation was examined by Fig. 3) with a micrometer eyepiece. Each ro was considering the relationship between the radius ofeach measured to the point where the annulus abutted the annulus and the total length ofthe fish over a periocl of opaquc zone. The r,? measurements were then stan-

IF. McGregor photographs.

Fig. 3: Otoliths of G. vulgaris. A: Fronr nrale of TL 56 mm taken in May 1971. Radius is 0.34 mm. B: From male of TL 64 mnr taken in November 1970. Radius is 0.43 mm. C: From female of TL 93 mnr taken in November 1970. Radius is 0.59 nrm. The arrow indicates the axis along which measurements were made. Nunrbers indicate the positions of annuli.

t6 I rl, ¡= 3;

E E ¡ ö-rl, n= 152 f I c) c o o .2 ! o É.

å-i-i-i-+-i- 11, n=2o2

30 40 50 ó0 70 80 90 100 ll0 120 130 Totol length (mm) Fi annuli (rl-r4) in male and female G' of each ànnulus radius in each 1 0-mm nd 52.18 (100- to 110-mm grouP) and

out dardised to adjust for individual variation in growth Therefore all subsequent analyses were carried the two sexes.) by multiplying by the factor .R/-R, where R Ìvas the separately for mean otolith radius for any observed total length. ,R The mean values of standardised annulus radii for males (252 fish) and relationshiP between total his relationshiP is indicated lines in Fig. 4. of otolith radius on total length (The regressions annuli occur in the same position in the otolith irres- were compared by analysis of foi males ãnd females pective of both the total length of the fish and the (Snedecor Cochran 1967) after covariance and number of annuli in the otolith. For example, the of variances (Sokal and Bartlett's test of homogeneity position of annulus I in a 2-anntrlus otolith from the l0- to 90--* length class is similar to that of annulus I in a 4-annulus otolith from the 110- to 120-mm length class. Such correspondence of annuli- indicates thJ regular temporal pattern in which they were formed. Although it has been assumed by some workers that annuli arè formcd onçe a year, this is not always so

17 (for example, see Yunokawa (1961)). The time of annulus formation in G. vulgans was estimated from monthly changes in the marginal growth index Moles of the n=71 otolith. The growth index, as given by Matsuura rlr (1961), is represented by: TIi Gr:o-'o - R _ rn_t I ilrtllrlii c where G¡ is the marginal growth index, -R is the total otolith radius, and rn and rn_, are the radii of the ¿0. E Femoles ultimate and penultimate annuli respectively. All I q n=lO2 measurements were made to the outer edge of each annulus where it abutted the next outer opaque zone. The mean marginal growth index was determined lirll monthly for fish of each sex. The indices from June 1970 to May 1971 are presented (Fig. îiiir¡{liîTrll 5) wirh their 95f i r confidence intervals. As shown by the decrease in the I growth index, the annulus formed in the otolith from J August to 0. I October. Within this period the exact time JJ A SON of formation varied between fish, DIJ FMAM and I October was 1970 1971 chosen arbitrarily as the date of annulus formation. It Time (months) was formed during the Fig. 5: Time of annulus formation in the otoliths of G. vul- g, and one opaque zone garis as indicated-by changes in the marginal growth index. summer, I ne mean value t-or each index is given wjth its 95\ con_ and autumn. fidence intervals (vertical bars). Fish in their flrst year of life with no annulus formed were designated by international convention water temperatì.lre were increasing ancl when growth (Tesch 1968) as age 0f . The first annulus was formed rate was at a minimum. This coincidecl with the bY the after hatching breeding season. However, annuli formed at the same (which to Novembei in the time in young-of-the-year fish irrespective of whether 1976a)). Afrer n9t they spawned. Thus, all young-of-the-year format entered their 9r females formed annuli, though .rone oi them shed second growing season and were designated age l f, eggs or were even young-of_the_year and so on. ripe, and all males formed annuli, though only s&re of t-hem spawned (Cadwallader 1976a). Therefore, spawning as such did not effect annulus formation, Factors Affecting at least iñ the Formation of Annuli young-of-the-year fish. Several workers have indicated a correlation between the time of annulus formation and certain external parameters, (K. W. Jensen 1957, Ouchi 1931a, l93lb, 1932). Hartle li ro be S ( versal of the diel activity pattern (Cadwallader 1975b) S and a change in the normal sedentary habits e of thó fish (Cadwallader 1976b). was indicate It theréfore a time of that certain external factors play an important physiological readjustment. role in some species, it is difficult to isoúte the directly responsible for . As Nikolsky (1963) pointed out, ir would be 1967). Recently Pannella rncorrect to assume that annulus formation occurred urrence of daily growth perely as a response to changes in growth or meta_ species he studied. He bolism brought about by altered external conditions. suggested that there is a relation between the rate of calcification and reproduction. Annulus formation in G. vulgar¿s occurred during late winter and early spring, when day length anã

l8 LENGTH.FREQUENCY ANALYSIS

Length-frequency distribution analysis may be used Totol to age fish which have a fairly limited spawning n=13ó season. At any one time a population of such tsh consists of a series of discrete age groups. The size range of each age group tends to be distinct from that adjacent groups and may L of be indicated by a mode in 0 I a length-frequency distribution. This method of ageing is applicable to G. ,trlgaris, which has one well- l0 defined spawning season in late winter-early spring. .9-c. Age group 0+ n= 57 The number of fish in each 2-mm length class was o recorded for each regular monthly sample and for o samples of recently hatched fish. The resulting poly- l) E modal length-frequency distributions from June 1970 f to May l9ll are presented in Fig. 6 (1970 August z 'l+ and October samples are omitted because of small l0 Age group sample sizes). In most months two, and in some n=ó5 months three, distinct length groups are indicated, but for most of the year only the flrst two groups can be separated by this method. Since the whole population was represented in the samples referred to above, Agegroup 2+ n=9 the first group represents fish in their first year I I J of life, Aqe qroup J+ t1=2 r designated age 0f, the next group represents fish in 0 Aqeqroup 4+ n=3 r r their second year of life, designated age 1f , and so 0 on. 40 50 ó0 70 80 90 r00 ll0 120 Totol length (mm) For a comparison of the otolith and length- frequency distribution methocls the age of each fish Fi quency distribution of G. vulgaris in the in the July 1970 sample Ìvas determined from otoliths. n July 1970, with the length-frequency the component age groups as determined The length-frequency distribution of fish in each age group was then plotted and is given (Fig. 7) with the total length-frequency distribution for that month. bimodal nature of the second (age I and third Fish designated 0* and 1f by both methods coincide. f) (age 2|_) length classes in most months and is shown This further validates the use of otoliths for ageing particularly well the December 1910 the G. vulgarís. in and January and February l97l length-frequency distri- The method of length-frequency distribution butions (Fig. 6). analysis is inadequate for ageing older fish, mainly In practice smaller fish were aged by length- because of the small numbers of fish surviving for frequency distribution analysis, and larger fish were more than 3 years. There is also an increasing overlap aged by reading of otoliths. in the length-frequency distributions of older ûsh produced by individual variation and reduced growth Most G. vulgaris taken in the present study belonged rate. The situation is further complicated by different to the age groups 0+, l+, and 2l; a few were in rates of growth in males and fenales after the first the 3* and 4-F age groups, none \ryere in the 5f age year of life (page 2l). This is indicated by the group, and one was in the 6* age group.

20 29 39 4.0 50 ó0 70 80 90 loo tlo 12o

29,30 Moy 1971

1,2 Moy 1971

't0 3,4 Apr 1971

6,7 Mo¡ 1971 --.E - LL-.----- n=176 Feb 1971 - 6,7 n=128 rJ¡- b- ¿--- 16,17 )on 1971 I¡r-_-- "=I -{b- -rr - I 13,14 Dec 1970 14,15 Nov 1970 n= 222

b00 Ê) 100 250 4 Nov l97O n =126 0 20o 13 Oct l9Z0 n= 256 0 800

2Q,21 Sep 1970

Totol length (mm)

Fig. 6 : Length-frequency distributions of G. vulgaris ftom J une 1970 to May l97l . From September to December samples of recently hatched fish were obtained from quiet stretches of the Glentui River immediately below the riffies in which the adults were found. From January onwards all fish were sampled in riffies.

l9 GRO\ryTII

ANNUÀL GROWTH IN LENGTH For fish aged from I to 4 years mean back-calculated lengths with their 95)( confrdence limits are presented Data on annual growth in length of G. vulgaris (Table 2). were derived by back calculation from otoliths with the formula given by Nikolsky (1963): l.ength reached at the end of the first year of life was not significantly different (P > 0.05) between t--;" (L--c)Ls males and females (¡ : 0.6351, d.f. : 363). However, after the first year, growth in length of males lagged where L,, is the length of the fish at age n (that is, behind that of females. Student's I tests, used to when the nth ring was formed), ,L is the length of the compare the mean lengths achieved at the end of the fish and ,R is the otolith radius when the sample was second, third, ancl fourth years showed taken, R,, is the radius of ring n, and c is the hypo- of life, significant differences (P 0.01) between males and thetical length of the fish at the moment the otolith < females (t:2.9068, d.f. : 205; t:4.7828, d.f. : 75; began to form. The value of c is derived by extra- : 4.7512, d.f. : 20 for conparison lengths polation from the regression of otolith radius on total t of achieved at the end of the second, third, and fou,rth length (Fig. a) and is the intercept on the total length years of life respectively). axis; for males c : 0. 56 and for females c : - 8 .08. The von Bertalanffy equation (von Bertalanffy 1938) adequately described annual growth in length TABLE 2: Mean. annual. back-calculated lengths (TL) of G. vulgaris derivetl from otoliths, with their 95/, of G. wtlgaris. The equation is represented by: (CL) confidence limits It: L-(l - e-KG-to)) Age Mean TL 95)( No. of where /¿ is the length at age t, L* ìs the average (years) (mm) fish CL "maximum" or asymptotic length, K is a constant Males determining the rate of change in the length incre- I 6l .33 +0.7156 15'l ment, / is age in years, and to is the hypothetical age 2 79.06 +t.2872 85 when length is zero. The equation parameters were 3 9t.12 +1 32 .9425 computed (1966) 4 97.81 +4.3493 8 by the Allen method from data from individual fish up to age 4f . For males ( n: 157): Females (* It : 112.6 (l e - 0'4271 + o'saoa) ¡ | 6t.66 +0.'7307 208 - 2 8L52 +1.0791 122 and for females (n : 208): 97 45 3 .4s +t.8622 0'22s5 (r + r'rrsz¡1. 4 110.07 +3.8940 14 lb - 158.9 (l -- e -

TABLE 3: Mean lengths (in millimetres, with9Sl confidence limits (t) in parentheses) of G. vulgaris in the Glentui River from June 1970 to May 1971. (Age 0l fish have been included as males; August and October sarnples are omitted because of small sample sizes.)

1970 Age group Jun Jul Sep Nov Dec Jan Feb Mar Apr May May 27 8 2Ç27 20--21 l4-15 t3-14 t6-17 6-7 6-7 34 1-2 29-30 Males and 11.3* 13.3f 19 .2 29.9 43.2 51 .4 59 .7 6t .6 63.5 females 0 (0. 1) (0.4) (0.8) (1.e) (2.0) (0.e) (l .0) (1.1) (1 .1) t2.11 16.3$ (0. l) (0. 5) Males 0*/l-l- 57.8 56.3 58 .3 61 .4 62.8 67.7 67.5 10.6 7s.4 't6.0 76.8 (1 .5) (1 .1) (1.3) (0.8) (1.4) (0.e) (0.6) (1.0) (0. 5) (r .2) (1 .2) Females 0+/l+ 68.6 73.4 73.8 7'7 .9 82.6 85.4 85.3 (l .6) (0.e) (0.8) (0.6) (1.0) (0.e) (1.7) Males l*/2-l 76.6 '77 .6 75.6 76.3 77 .4 81 .0 84.7 84.7 90. 3 90 .7 94 .6 (1.1) (1.1) (l .5) (1.5) (0.3) (0.8) (0.8) (0. 5) (0.8) (0.6) (1.5) Females l+/2+ 82.8 81.6 83.1 82.2 84.0 85.4 91.1 90.9 95.8 98.2 100.5 (0.8) (1.6) (0. e) (0.5) (l.4) (0.4) (0.e) (0. 8) (0.7) (1.6) (1.2)

* Sample taken on 30 September. f Sample taken on 13 October. I Sample taken on 4 November. $ Sample taken on 20 November. 2l SEASONAL GRO\ryTH IN LENGTH 0+ C. vulgaris in another Canterbury river (the Cass River), where fish reached a mean length of 58.6 mm Seasonal growth length vyas estimated from in in their first year of life, with no significant growth monthly changes the mean length each age in of occurring between June and September*. group. Mean lengths vr'ere calculated directly from groups of recently hatched fish where there was no Differences in growth in length between year overlap in length distribution with older age groups. classes were apparent; for example, in the May For other groups in the regular monthly samples (29-30) l97l sample, age 0f fish had reachecl a mean lengths were estimated from length-frequency length of 63.5 mm, whereas in the June 1970 sample distributions by the probability paper method (Cassie age 0* frsh had attained a length of only 57.8 mm. 1950). This method was usecl to measure seasonal Seasonal growth length also varied in different growth in fish up to age 2-F; the small numbers in in parts River. This was indicated by the population did not permit the use of this method of the Glentui comparing length-frequency distributions of flsh to follow seasonal growth in older ûsh. Mean lengths the caught at the same time in different parts of the river. of G. vulgør¿s from June 1970 to May l97I are shows the length-frequency distributions of presented (Table 3). Figure 8 the regular samples taken in March, April, and May The period of increase in length for both males and l97l with those of samples taken during the same females extended from November to the following periods from further down stream. The greatest May, \,ith maximum growth occurring from December to April. Growth ceased from June to *Benzie's data were converted from standard lengths to total Benzie (1968) lengths with a conversion factor of 1.149 (derived from the October, in winter and early spring. ratio between these two measurements in G. vulgaris given by found a similar pattern of growth in length in age McDowall (1970)).

s .2

o o -o E ztr

3,4,5 Moy n= 2O2

1,2 Moy B n=180 -r-Ü.-E 40 50 ó0 r00 110 120 r30 Totol length (mm) --

Fig. 8 : Length-frequency distributions of samples o1 G. vulgaris taken in different parts of the Glentui River in 1971. A : Samples taken from section E-F (see Fìg. 1). B: Samples taken from section C-D (see Fig. l) during the regular sampling programme.

22 TABLE 4: Mean lengths and 95 f confidence limits (l) of age 0l G. vulgaris caught during the sa_me periods in different parts of the Glentui River. (The diherence bctween paired means is significant at the 0.001 probability level.)

Regular samples (C-D, Fig. l) Other samples (E-F, Fig. 1) Difference Mean Mean between length length means Date (mm) 95%CL Date (mm) 95%CL (mm) f d.f. 6, 7 March 5t.4 0.9 8, 9 March 59.2 1 .l 7.8 10.579 107 3, 4 April 59.7 1.0 6, 7, 8 April 63 .5 0.6 3.8 6.433 254 1,2 Ìl;l.ay 6t.6 1.1 3,4,5 May 65.4 0.8 3 .8 5 .797 206

differences were those between age 0+- fish; mean Fish in the 0* and I + age groups of each sex lengths of this age group with the differences between generally gained v/eight throughout the year, though samples are presented (Table 4). Different rates of most weight was added during summer and autumll. growth in fish of the same species in different parts of Similarly, growth data for age 2l and 3 -F fish of each a river have been reported also by Went and Frost sex approximate parabolas \ryith the low point at (1942) and Purkett (1958). September-November (the post-breeding period), which indicates that most weight was gained in summer and autumn, from January to May. Growth in length ceased during the spawning period, but SEASONAL GROWTH IN WEIGHT growth in weight was most affected becanse of shedding of the sexual products. 'Wet weights were obtained from fish subsampled from regular monthly samples. l)ata on growth in Tlre largest G. 'tulgaris taken in the present study weight for fish in the younger age groups are shown was an age 4! female which weighed 2A.9 g and was (Fig. 9); older fish were too few to provide adequate 125 mm long; the largest male was also age 4f , data on growth in weight. weighed 13.5 g, and was 107 mm long.

23 Moles

't0

Time (months)

o+ I r* I z* I ¡* Age group

Fig. 9- (a in lveight of male and female G. vulgaris from Pojnts 10 ñsh'ín the sample. trttreiãlãie io'òr more ñ fish is (ndicated Uv ttrdveriicai Uárj.-tiã num¡er of above

f .|ì r¡ t, Femoles

.37

. tl

J' 'M' 'M' 'J' 's J' 'M' 'M Time (months)

3+ Age group

25 LENGTH-\ryEIGHT RELATIONSHIP AND CONDITION

Data on length and weight for each 2-monthly For G. vulgaris immediately after hatching, in period were combined for analysis of the length- October and November, the å coefficient was not weiglit relationship in G. vulgaris. Males and females significantly different from 3.0. However, during the were treated separately and, for each sex, fish older juvenile phase, from December to March, extremely than 0f were lumped together in one group and high ó coefficients were recorded for both males and treated separately from age 0f fish. Regression lines females, which indicated a rapid increase in weight for for the two age groups of each sex were used to a small increase in length. From the end of the juvenile describe the length-weight relationship for each phase to the end of the first year of life the å coeffi- 2-n-ronthly period; regression coefficients (with their cients were again not significantly different from 3.0. 95\ conñdence limits) for these relationships are Similarly, Macphee (1960), working with the large- presented (Table 5). scale sucker, Catostomus macrocheilus Girard, reported å coefficients of 4.9125 and 3.3496 for fish below and above 20 mm in length respectively. He concluded that the point of intersection of the two regression lines indicated the upper limit of the postlarval stage of development. The ö coefficient of female G. wlgaris older than age 0* was significantly range at certain stages of growth (Macphee 1960). greater than 3.0 in 8 out of 12 months; with males ö Values of å which differed significantly from 3.0 tended to be lower than for females and was signi- were detected by use of the formula given by Snedecor ficantly greater than 3.0 in only 4 months. For both and Cochran (1967): sexes ó was significantly greater than 3.0 just before the breeding season and was reduced during spawning 4- _b-B from August to September. S¿ Unless the å coefficient is the same for all groups of where / is a critical value of the I distribution and has flsh whose condition is being compared, coefficient a n degrees freeclom, -2 of ó is the observed gradient, cannot be used as a measure of their relative condition iq ttrg expected gradient { (in this case B : 5.0¡, anA (Le Cren 1951). This applies particularly when changes S¿ is the standard error of ó. in the two coefficients are correlated, which is usual

TABLE 5: Length-weight relationships of G. vulgar¡s from June 1970 to May 1971., Relationships are of the form: : a b log /. (Length is in millimetres, -clÏ log w I weighi in grams; ãó"nä"ti"Ë'í¡*it, (+_) ,nã-RH-6.;ì;"uy hatchert fish.) Males Females No. No. in, Coeff. a 95%CL Coeff. á 95%CL in Coeff. sample . a 95%CL Coeff å 95%CL Age 0f sample Jun, Jul 38 0.015 3.303 24 -5.695 0.542 33 -5.063 0.013 2.944 o.5l I Arg,-Q.p -4.s2s 0.0ié 2.880 ó'.4i4 ,u. 0 031 3.508 0.67s 9ct. Nov (RH) 7 _s.j76 0 06ô 3.051 1-.210 -u Dec, Jan :1, 13 -7 .452 0.018 4 415** O. g¿i Feb, Mar ?0 0.0,4 3.895** O'.jj; Apr, May 50 -6.642 0.072 i8 -e .fit 0.030 3.602* O. CiA -5.263 3.109 o.1qe 62 -4.695 0.01 1 2.765 0.359 Age > 0 -J- Jun. Jul Aug, Sep 53 -5.798 0.009 3.376** 0.226 Oct, Nov 2_0 -5.267 0.027 3.084 0.580 Dec, Jan 39 -5.551 0.017 3.238* 0.205 Feb, Mar 33 -5.637 0.014 3.276** 0.158 Apr, May 23 -5.515 0.013 3.230* 0.183 28 -5 .475 0 01 5 3 .236 0.253 t'Significant deviation from 3.0 at the 5,'A probability level. x*Significant deviation from 3.0 at the I il probabiliry level. In the age 0* fish' numbers for october ancl November males anrl-nrrl-ir'?irl. femaìes are combined; the regression line is based on the weights of 7 samoles each ol tno*n rr,eã"'l.ngtr,'1tot"r nu'"rr.ì-oi N"*b¿;.-f"i iemales are also t;";;ìù;. and January males and ôombined; thJ;¿g*s;l;;'ji"î'Ì.'uàrè¿ o"-trrÉ-*.iànir"of til's".pl"s each offish is 130). of known mean rength (totat number

26 (Hile 1936). Thus, the regression line is affected by differential changes in weight in fish of different lengths. Coefficient å differed signiûcantly between 2-monthly samples for each group of G. tulgaris, and for both age groups of each sex å was negatively correlated with a throughout the sampling period (r : (males, age 0f); r : (males, -0.972 -0.995 : age >0f); r : -0.974 (females, age 0f); r Therefore, for each -0.984 (females, age >0*)). group changes in coefrcient ø could not be used to o) compare condition at various times of the year, and 7 so changes in condition were considered in terms o) of 30, :õ of predicted weight values for hypothetical fish 5 80, and 100 mm TL (from the regression coefficients given in Table 5) and are presented (Fig. l0)'

The condition of age 0f frsh increased throughout the growing season. In older fish both males and females showed similar trends, with an increase in condition in late spring and autumn and a reduction in condition in early sprìng and summer. Females were generally in better condition than males.

To eliminate the effects of changes in gonad weight' regression lines were fitt€d to length-somatic weight J ) A S O N DIJ MAÀ{ 1971 1970 Time (months)

Fig. 11 : Two-monthly that is, total weight) relati omatic weight relãtiônship femalc G. vulgaris of 100 mm TL.

(that is, total weight of flsh with gonad weight sub- tracted) data, and modified condition factors were calculated. The modified condition factors followed o) the same trends as the condition factors based on total weights, which indicates that somatic tissue undergoes -c, gonad development. .91o seasonal changes irrespective of Ì The 2-monthly variation in the length-weight (that is, total weight) relationship and the 2-monthly variation in the length-somatic weight relationship for hypothetical fish of 100 mm TL are shown (Fig. 1l)' Though both relationships follow the same trend in each TL = 30, oge 0+ F sex, the development of gonads has a greater effect on M than in males. ^ -/'^ condition in females "/--'\-\F The trends in condition referred to above were reflected in the fat deposits overlying the stomach and posterior part of the alimentary tract. A scheme ) AM on thaf of Prozorovskaia (cited by Nikolsky 1971 baseå 1970 Time (monlhs)

27 FACTORS AFFECTII\G THE GRO\ryTH OF G. VULGARIS

Growth results from the consumption of food, its empty stomachs in samples taken in spring and assimilation, and its transformatior-r into body con- summer may be regarded as an indication of an stituents. As indicated by Coche (1967), it may be increase in food consumption then. regarded as the energy sr"rrplus transformecl to tissue after the metabolic requirements for maintenance The effect of temperature on metabolism, food have been met. Thus, any factor which affects either intake, and growth has been demonstrated by a num- directly or indirectly the consumption and assimilation ber ol workers, for exarnple, Markus (1932), Brown of food ancl its transformation into tissue may be (1946b), Baldwin (1957), Frost and Kipling (1968), regarded as having an effect on growth. Elwood and Waters (1969), and McCormick, Hokan- son, and Jones (1972), and under natural conditions The importance of both the quar-rtity and quality changes in growth rate have often been correlatecl with of food in affecting the growth of fish has been changes in temperature (Swift 1961, Nikolsky 1963, demonstrated in a number of experimental situations, May, Pinhorn, Wells, and Fleming 1965, Frost and for example, Pentelow (1939), Brown (1957), palohei- Kipling 1967, Muth 1969, J. E. Johnson 1970). mo and Dickie (1965, 1966a, 1966b), and Warren and ,-ulgar¡s Davis (1967). However, it is difficult to demonstrate In G. rnost growth occurred when the highest water temperatures the inrportance of food availability as a factor influ- were recorded, and growth ceased when water encing growth under natural conditions unless detailed temperatures were low. The importance ofthe influence oftemperature on the length of the growing season was pointed out by Van Oosten (1944). Temperature appeared to affect the length of the growing season of G. vulgarls, particularly for age 0f fish, in that higher temperatures were correlated with earlier spawnings, so that those fish which hatched from early spawnings had a longer growing season than those hatched from late spawnings. The onset of spawning differed by up to 6 weeks in different parts of the Glentui River (Cadwallader 1976a), so In G. vulgans the proportion of empty stomachs that differences in length-frequency distributions in was highest in samples taken during autumn and win- samples from different parts of the river (Fig. g) were ter, and lowest in samples taken dr.rring spring and to be expected. summer, when most growth occurred. Metabolic As suggested by the work of L. Johnson (1966), the seasonal growth cycle may be influenced by light acting through changes in da¡r length. Gross, Roelofs, and Fromm (1965), working with green sunfish, Lepomis cyanellus Rafinesque, showed that photoperiod had a pronounced effect on food consumption, food conversion efficiency, and growth. Gilaxias vulgaris is usually active mainly 1936, Reimers 1957, Molnár and Tölg 1962a, 1962b, at night and obtains its food when it is active (Cadwallader 1975b). Tyler 1970, Elliott 1972), so that in summer food is How- ever, maximum growth occurred at generally digested more quickly than in winter. This a time when nights were shortest and consequently the potential would tend to increase the number of empty stomachs timJfor -of food consumption was reduced. found in summer if the same amount food was A number of workers, for example, Hile (1936) and Frost and Kipling (1967), have shown an inverse relationship between population density and growth rate. This is normally explained by different degiees of intraspecific competition for food. At high ãensity competition is severe and individual growth rate ii reduced. Various interspecific relationships may also 1ry9! t!" rate of growth (Nilsson 1967, WeatLerley 197 2). Furthermore, Brown (19 57), Ivlev ( I 96 I ), Chen and Prowse (1964), and Chen (1965) described experiments in which it was shown that space factors had

28 an effect on growth; both the total volume of water between ûsh. However, as indicatecl by Chapman and the degree of crowding of the individuals (acting (1966), since the supply of drift passing a given point independently of competition for food) were thought is proportional to water velocity, it is conceivable that to be important. Space factors and intraspecific drift feeders such as G. vulgaris require less space to competition for food may be significant for recently guarantee adequate food when velocities are high. hatched G. vulgaris, which are gregarious. With adults Since large boulders are usually associated with rapid the situation is different in that direct intraspecific flows, the two factors probably exert complementary competition for food is reduced by the establishment effects on the size and therefore the number of terri- of feeding territories. tories. Finally, the physico-chemical condition of the water may also affect growth (Hile 1936). Observations on adult G. vulgaris in a laboratory stream tank (Cadwallader 1975e) indicated that Growth in G. vulgøris under natural conditions may though they do not have distinct, patrolled territories, depend on any one, or a combination, of the factors each appears to have a feeding station or a "diffuse discussed above. Experimental analysis may indicate station territory" of the type described by Kalleberg the importance of some of the major factors in con- (1958). As Kalleberg pointed out, the number of such trolling the growth rate, but under natural conditions territories on the stream bed may depend on the topo- the growth rate must be considered to be the result of graphy of the bottom, with more territories poten- a number of factors, both internal and external, some tially available in areas where there are plenty of of which may have an effect only at certain stages of obstacles (such as boulders) to prevent visual contact the life cycle.

29 SUMMARY

In this study of the age, growth, and condition of I * age groups of each sex generally gained weight Galaxias vulgaris in the Glentui River, Canterbury, throughout the year, though most weight was added New Zealand, fish were aged by sagittal otoliths and during summer and autumn, at the time when fish length-frequency distribution analysis. Most G. in the 2l and 3-F age groups also gained most vulgaris belonged to the age groups 0+, l+, and2l, weight. a few to the 3+ and 4* age groups, none to the 5-| age group, and one to the 6f age group, in its The length-weight relationship was calculated seventh year. separately for males and females and for age 0f and older fish of each sex. Since the values of the å co- Annual growth in length was estimated by back efficient for the various groups differed significantly calculation from otoliths. Length reached at the end between 2-monthly samples, changes in coefrcient ø of the first year of life was not significantly different could not be used to compare condition at various between males and females. Holvever, after the first times of the year. Changes in condition \ryere therefore year, growth in length of males lagged behind that of considered in terms of predicted weight values derived females. The von Bertalanffy equation adequately from the 2-monthly regression coefficients calculated described annual growth in length of both males and females. for each group. The condition ofage 0+ frstr increased throughout the growing season. In older fish, both Seasonal growth in length was estimated from males and females, there was an increase in condition monthly changes in the mean length of each age in late spring and autumn and a reduction in condition group. Increase in length of both males and females in early spring and summer. Females were generally in occurred from November to May, and growth ceased better condition than males. Somatic tissue underwent from June to October, in winter and early spring. seasonal changes irrespective gonad Differences in growth in length were apparent be- of development. tween year classes and between fish sampled in The seasonal changes were reflected in the fat deposits different parts of the river. overlying the alimentary tract.

Wet weights were obtained from fish subsampled Possible factors affecting the growth of G. vulgaris from regular monthly samples. Fish in the 0* ãnd are discussed.

ACKNO\ryLEDGMENTS

This work was done at the Department of Zoology, for the provision of suitable facilities, Dr D. J. University of Canterbury, during the tenure of-a Staples for initial advice on the project, Dr V. L. re-search fellowship from the New Zealand Ministry Benzie for helpful discussion, and Dr V. M. Stout for of Agriculture and Fisheries. I am grateful to thê water analyses. I thank Professor p. A. Larkin, of following, whom all of are members or former the University of British Columbia, for the use of a members of the staff of the Department of Zoology, computer programme for fitting the von Bertalanffy University of Canterbury: Professor G. A. Knox equation.

30 REFERENCES

All¡¡r, K. R. 1966: A method of fitting growth curves of the C,lnr-morn, K. D. 1969: "Handbook of Freshwater Fisbery von Bertalanffy type to observed data. Journal of the Biology," Vol. l. Iowa State University Press, Ames. Fisheries Research Board of Canada 23: 163-79. 752 pp. BltowrN, N. S. 1957: Food consumption and growth of C.lssrr, R. M. 1950: The analysis of polymodal frequency brook trout at different temperatures. Transactions o.f distributions by the probability paper method. N.Z. the Amerícan Fisheries Society 86:323-8. Science Review 8 : 89-91. BENZIE, V. L. 1961 : A comparison of the life history and CnlrulN, D. W. 1966: Food and space as regulators of variation in two species of Galaxias, G. attenuatus and salmonid populations in streams. American Nafuralist G. vulgarís. (M.Sc. thesis, lodged in University of 100: 345-57. Canterbury library.) CnrN, F. Y. 1965: The living-space effect and its economic 1968' The life history of Galaxias vulgaris Stokell, implications. Appenclix 3, Tropical Fish Culture Research with a comparison with G. maculatus attenuatus. N.Z. Institute, Malacca, Report for 1965: 37-40. Journal oJ Marine and Freshwater Research 2: 628-53. CHru, F. Y-, and PRowsE, 1964: The -Btnr.ltlNFry, L. von. 1938: quantitative theory of organic G. A. effect of living A space on the growth rate of fish. Ichthyologíca growth (inquiries on growth laws. 2). Human Biology 3:, 1l-2O. I0: 181-213. CnucuNovr, N. L 1959: "Age and Growth Studies in Fish." BH.ÀTIA, D. 1931a: A critical study of the scales of two specimens Academy of Sciences of the U.S.S.R., Department of of starved and excessively fed trout-Salmo irideus. Biological Sciences, Board of lchthyology, Instirute of Journal du Conseil Permanent International pour Animal Morphology. (Translated from Russian and published for the National Science Foundation, I'Exploration de la Mer 6:266-72. \ùy'ashington, 1931b. On the production of annual zones in the D.C., by the Israel Program for Scientific scales of the rainbow trout (Salmo irídeus). 7. Journal Translations, Jerusalem, 1963, 132 pp. OTS 61-31036.) - of Experimental Zoology 59:.45-59. Cocnr, A. G. 1967: Production ofjuvenile steelhead trout in a 1932. Factors involved in the production of annual freshwater impoundment. Ecological Monographs 37: zones on the scales of the raìnbow trout (Salmo irideus). 201-28. 2. fournal ol Experimental Biology 9:6-ll. - DrNNnvrc, E. H. 1956: Chemical composition of the zones in Bouo, R., and ErooN, G. A. 1960: Investigation of large cod otoliths. Journal du Conseil Permanent International black shags in North Canterbury to determine feeding pour l'Exploration de la Mer 2I: 756-9. habits of shags in this locality. N.Z. Marine Department DnnrtNcroN, P. J., jun. 7957 : " Zoogeography : the Geographi- Investigation Report No. 21. 5 pp. cal Dist¡ibution of dnimals." Wiley, New York. 675 pp. BnowN, M. E. 1946a: The growth of brown trout (Salmo Dr BoNr, A. F. 1967: Some aspects of age and growth of fish in trutta Linn.). 2. The growth of two-year-old trout at a temperate and tropic constant temperature of 11.5'c. Journal of'Experimental "The Biological Basi Biology 22: l3O-44. pp. 67-88. Blackwell 1946b' The growth of brown trout (Salmo trutta Edinburgh. Linn.). 3. The effect of temperature on the growth of DEcENs, E. T., D¡uspn, W. G., and Hleonrcu, R. L. 1969: two-year-old trout. Journal of Experimental Biology 22: Molecular structure and composition of fish otoliths. - 145-55. Marine Biology 2: lO5-13. (Ed.) 1957: "The Physiology of Fishes", Vol. I and DrcrrNsoN, P. 1951: Stomach contents of New Zealand inland Vol. 2. Academic, New York. 441 pp. and 526 pp. shags. Australian Journal of Maríne and Freshwater Re- -Clpw¡.r-r¡oen, P. L. 1974: Effect of formalin on the length search 2: 245-53. and weight of Galaxías vulgaris (Pisces: Salmoniformes: DuNceN, K. W. 1968: The food of the black shag (Påalacrocorax Galaxiidae). Mauri Ora 2:63-6. carbo novaehollandiae\ in Otago inland waters. ?r¿zs- 1975a. Feeding habits of two ûsh species in relation actions of the Royal Sort"r, of N.2., Biological Sciences, to invertebrate drift in a New Zealand river. N.Z. Journal I I: 9-23. of Marine and Freshwater Research 9: 1l-26. - Er-ltorr, J . M. 1972:. Rates of gastric evacuation in brown trout, 1975b' Spontaneous locomotory activity of Galaxias Salmo trutta L. Freshwater Biology 2: l-18. vulgaris Stokell (Pisces: Salmoniformes). N.Z. Journal Er-wooo, J. W., and MrEns, T. F. 1969: Effects of floods on - of Marine and Freshwater Research 9:27-34. food consumption and production rates of a stream 1975c' The food of the New Zealand common river brook trout population. Transactions of the American galaxias, Galaxias vulgaris Stokell (Pisces: Salmoni- Fisheries Society 9B : 253-62. formes). Australian Journal of Marine and Freshw,ater Frrr¡, R. 4., and Srorrll, G. 1945: Investigation of the - Research 26: 15-30. stomach contents of New Zealand fresh-water shags. Proceedings 1975d: Feeding relationships of galaxiids, bullies, eels Transactions and of the Royal Society of and trout in a New Zeala¡d rivet. Australian Journal of N.2.74:320-31. - Marine and Freshwater Research 26:299-316. Fnosr, W. E., and Kru-INc, C. 1967:. A study of reproduction, 1975e' A laboratory study of interactive segregation early life, weightJength relationship and growth of pike, between two New Zealand st¡eam-dwelling frsh. Journal Esox luciusL., in Winderme¡e. Journal of Animal Ecology -- of Anímal Ecology 44:865-75. 36: 651-93. l976ai Breeding biology ofa non-diad¡omous galaxiid, 1968. Experiments on the efect of temperature on the Galaxias vulgaris Stokell, in a New Zealand river. growth of young pike, Esox lucius L. Salmon and Trout Magazine 184: 170-8. - Journal of Fish Bíology B: 157-77. - 1976b; Home range and movements of the common GrvnErr-, R., and MEssronrr, J.1964: Age determination in thc river galaxias, Galaxias vulgaris Stokell (Pisces: Salmoni- whiting (Merlangius merlangus L.) by means of the formes), in the Glentui River, New Zeala¡d. Australian otoliths. Journal du Conseil Permanent Internationol poat - Journal of Marine and Freshwater Research 27:23-33. l'Exploration de la Mer 28:393-4A4.

3r Cn¡co, D. R. I f New Zealand, I :250 000, McConurcr, J. H., HorlNsoN, K. E. F., and JoNes, B. R. 1972: Sheet 18 nt of Scientífic and Indus- Effects of temperature on growth and survival of yollng rrial Res brook trout, Salvelinus fontinalis. fournal of the Fisheries Cnoss, W. L., Ro*ors, E. W., and Fnorrau, P.O. 1965: Influence Researclt Board of Canada 29: ll07-12. of photoperiod on growth of green sunfish, Lepomis McDow.lrr, R. M. 1967; New landJocked fish species of the cyanellus. Iournal of the Fisheries Research Board of genus Galaxias from North Auckland, New Zealand. Canada 22:1379-86. Breviora 265:1-ll. HrnrrEy, P. H. T. 1947l. The natural history of some British l9ó8' Galaxias maculatus (Jenyns), the New Zealand freshwater frshes. Proceedings of the Zoological Society whitebait. Fisheries Research Bulletin, N.Z. Marine of London 7 I I : 129--206. - Depqrtnrenl, No. 2.84 pp. Hrucr<, F. R. 1949 : Some harmful effects of the electric shocker _ 1969l. Relationships of galaxioid frshes with a further on large rainbow trolt. Transactions of the Anterican discussion of salmoniform classification. Copeia 1969; FÌsheries Society 77 : 6l-4. 796-824. Hrr-E, the ci edi l97O' The galaxiid fishes of New Zealand. Bulletin of the n ds, the Museum o.f Comparative Zoology, Harvard Universít1', Unit of - 139:341-431. 1972'The species problem in freshwater fishes and the I{ororN, M. J. 1955: Ring formation in the scales of Tilapia -- taxonomy of diadromous and lacustrine populations of variabilís Boulenger and, Tilapia esculenta Graham frõm Galaxias maculatus (Jenyns). fournal ofthe Royal Society Lake Victoria. Appendix C, East African Fisheries of N.Z. 2: 325-67. Research Organization Annual Report Jbr 1954/1955: 36-40. MlcruEr, C. 1960: Postlarval development and diet of the largescale sucker, Catostomus tnacrocheilas, in ldaho. F[orxINs, C. 1,. 19'71: l,ife history of Galaxias tliveyg¿ns Copeia 1960: 119-25. (Salnronoidea: Galaxiidae). Journal Marine N.Z. of M.rnrus, C. 1932: The extent to which temperature changes and Freshwater Research 5:41-5'1. fI. influence food consumption in largenrouth bass (flaro HorxrNs, C. L., and McDownr, R. M. 1970: A review of florídnna). Transactions of the American Fisheries Socíety present knowledge of fishes in New Zealand fresh waters. 62:202-10. Proceedings Part of N.Z. TVater ConJÞrence t970: I S. 196l : Age growth 10. 1-10. 14. Mlrsuuna, and of flatfish, Ganzö-birame, Pseudorhombus cinnamoneus (Temminck et Schlegel). HvNEs, H. B. N. 1970: "The Ecology of Running Waters." Records of Oceanographic Works in Japan, Special Liverpool University Press, Liverpool. 555 pp. Number 5: 103-10. Inrn, T. 1955: The crystal texture of the otolith of a nrarine M.nv, A. W., PrNnonN, A. T., WEns, R., and Fr¡urNc, A. M. feleost Pseudosciaena. Journal of the Faculty of Fisheries 1965: Cod growth and temperature in the Newfoundland and Animal Husbandry, Hiroshima University, 1: l-8. atea. Internat ional Commi ssion for t he Northwest A tlantic Fisheries Special Publication 545-55. Ivuv, \y'._S. 1961 : "Experimental Ecology of the Feeding of 6: Fishes." Yale University Press, New Haven. 302 pp. MrN,r, M. V. 1968: A note on a problen in the visual qualitative JnusoN, A. C. 1965: A standard terminology and notation for evaluation ofotolith zo¡es. Journal du Conseil Permanent InternaÍional pour l'Exploration de Ia Mer 32: 93-'7. otolith readers. Research Bulletín, International Commis- sion fo r t he Nor t hw e s t At lant íc Fi s heríe s, 2 : 5-7 . Mro, S. 196l: Age and growth of red sea bream, Evynnis japoníca Tanaka. Records Oceanographic in JeNsrN, K. W. 1957 h of Titapia of Works nilotica L., and Laies Japan, Specíal Number 5: 95-101. niloticus C. . Kongelige MolNÁn, G., and Törc, I. 1962a: Relation between water Norslce Vide 30; ls}-'t. temperatllre and gastric digestion of largemouth bass (M JoHNsoN, J. 1970: Age, on dynamics icropterus salmo ides Lacépède). Journa I of the Fisher ies E. of Research Board o-f 19: threadfin shad, Doro ther), in central Canada 1005-12. Arizona reservoirs. American Fish- 1962b' Experiments concerning gastric digestion of eries Society 99 : 739-53. pike perch (Lucíoperca lucioperca L.) in relation to water temperature. Acta Biologica 13 : 231-9. JonNsoN,_ J. F d MrNcrcr-rv, W. L. 1970: - Selectivi f some commercial fishing Murs, K. M. 1969: Age and growth of broad whitefish, - devices reservojrs. Journal of thõ Coregonus nasus, in the Mackenzie and Coppermine Arizona 6:46-50. Rivers, N.W.T. Journal of the Fisheries Research Board Canada 26:2252-6. JoarsoN, L. 1966: Experimental determination of food con- of sumption of pike, Esox lucÌus, for growth and mainten- Nrxolsrv, G. V. 1963: "The Ecology of Fishes." Acadenic, ance. lournal of the Fisheries Research Board of Canada London and New York. 352 pp. 23:1495-505. Nrr-ssoN, N.-4. 1967: Interactive segregation between fish Knruerr

32 1966b. Food and growth of fishes.3. Relations amorrg 1949: The systematic arrangenìent of the New Zealand food, boCy size, and growth efficiency. Journal oJ the Galaxiidae. Part 2. Specific classiûcation. Transactions - Fisheries Research Board of Canada 23: 1209-48. - and Proceedings o.f the Roya! Socierlt of N.Z. 77; 472-96. P.rNN¡rr-r, G. 1971: Fish otoliths: daily growth layers and 1953. The distribution of the family Galaxiidae. periodical patterns. Science, New York, 173: ll24-7. Proceedings of the Seventh Pacific Science Congress oJ' the Pacific Science AssociaÍion 4: 48-52. Penrrlow, F. T. K. 1939: The relation between growth and - food consumption in the brown trout (Salmo trurta). 1959.Notes on galaxiids and eleotrids with descriptions Journal of Experimental Biology 16: 446-73. of new species. Transactions of the Royal Sociefy oJ N.Z. 87:265-9. Por-rnro, D. A. 1971 : The biology of a landlocked form of the - normally catadromous salmoniform fish Galaxias Swrrr, D. R. 1961 : The annual growth-rate cycle in brown trout maculatus (Jenyns). 1. Life cycle and origin. Australian (Salmo trutta Linn.) and its cause. Journal of Experi- Journal of Marine and Freshwater Research 22;91-123. ntental Bíology 38: 595-604. Punxnrr, C. A. 1958: Growth of the fishes in the Salt River, Tescu, F. W. 1968: Age and growth. 1z Ricker, W. E. (Ed.), Missouri. Transactions oJ the American Fisheries Society "Methods for Assessment of Fish Prorluction in Fresh B7: 116-31 . Waters", pp. 93-120. Blackwell Scientific Publications, Oxford and Edinburgh. Reruens, N. 1957: Some aspects of the relation between stream foods and trout survival. Caldornia Fish and Game 43: Tvun, A. V- 1970: Rates of gastric emptying in young cod. 43-69. Journal of the Fisheries Research Board o.f'Canada 27: 117'7-89. RtcrER, W. E. 195B: Handbook of computations for biological statistics of fish populalions. Bulletín of the Fisheríes VrN OosrnN, J. 1944: Factors affecting the go\À'th of fish. Research Board oJ Canada 119.300 pp. Transactions o.f the North American Wildlife Conference 9: 177-83. SruNorns, J. W., and Svrrn, M. W. 1954: The effective use of a direct current fish shocker in a Prince Edward Island \¡y'ARREN, C. E., and Dlvls, G. E. 1967: Laboratory studies on sfream. Canqdian Fish Culturist 16: 42-9. the feeding, bioenergetics, and growth of ûsh. 1n Gerkr'ng, S. (Ed.), "The Biological SNrorcon, G. W., and CocHRAN, 1967: "Statistical D. Basis of Freshwater Fish W. G. Production," pp. 175-214. Blackwell Scientiûc Methods." 6th ed. Iowa State Unìversity Press, Ames. Publica- 593 pp. tions, Oxford and Edinburgh. WEarnnnrEv, A. H. 1972: "Grorvth and Ecology Fish Sorlr, R. R., and Ronrr, F. J. 1969: "Biometry. The of Princip'les Practice Statistics Populations." Academic, London and New York. and of in Biological 293 pp. Research." W. H. Freeman, San Francisco.776 pp. Sororov, N. P., and Cnvrr-rovr, WnNr, A. E. J., and Fnosr, W. E. 1942: River Liffey Survey 5. M. A. 1936: Nr"¡trition of (Salmo Gombusia affinis on the rice flelds Turkestan. Growth of brown lrour. trutta L.) in alkaline and of Journal acid waters. Proceedings o.f Animal Ecology 5: 390-5. oJ'the Royal lrish Academy, B 48: 67-84. So'vEnrN, V. D. van. 1950: The "winter check" on trout scales WrNoer-r, J. 1967: Rates in East Africa. Nature, London, 165:. 473-4. T. of digestion in fishes. 1z Gerking, S. D. (Ed.), "The Biological Basis of Freshwater Fish Sppncrn, S. L. 1967: Internal injuries of largemouth bass and Production", pp. 151-73. Blackwell Scientific Publica- bluegills causerl by electricity. Progressive Fislt-CulturisÍ tions, Oxford and Edinburgh. 29: 168-9. Wooos, C.,S. 1964: Fisheries aspects of the Tongariro power Sr.l'rr-ns, D. J. 1971: Methods of ageing red gurnard (Teleostei : development project. Fi,sheries Technical Report,' N.Z. Triglidae) by fin rays and otoliths. N.Z. Journal of Marine Marine Department, No. 10. 214 pp. and Freshwater Research 5:70-9. 1967: A systematic biology of Gobiomorphas (Pisces, Srorrrr, G. 1938: A new species of the genus Galaxias, with a Eleotridae) with sr.rpporting studies on Retropinna and note on the second occurrence of Galaxias burrowsii Galaxias. (Ph.D. thesis, lodged in University of Phillipps. Records of the Canterbury Museum (N.2.) 4: - Canterbury library.) 203-8. 1968' An improved fish measuring board. N.Z. fournal 1945i The systematic arrangenent of the Ner¡, Zea\artd o-f Marine and Freshwater Research 2: 678-83. Galaxiidae. Part 1. Generic and sub-gene¡ic classifica- -YuNoxrrvr, Y. 1961 : On the age and growth of Chelidonichth¡,g tion. Transactions and Proceedings of the Royal Society krrnru (Lesson et Garnot). Records of Oceanographíc - of N.Z. 75: 124--3'l . Works in Japan, Special Number 5: 1ll-6.

33 INDEX

Age,9, 10, 13. Feeding territories, 29. Macrophytes, 11, 12, Age determination, I 5-20. Flooding, 11. Megaloptera, 12. Air temperatures, 11. Food,13, 18,28,29. Merlangius mer langus, 15. Anguilla australis schmidtii, 72. Food conversion efficiency, 28. Metabolism, 15, 18, 28. Ang uilla diefenbachii, 12. Methods, general, 13-14. Anguillids, 12. Micropterus salmoides, 13. Annual growth in length, 21. Galaxias,9. Mortality, 13. Annuli, 15, 16, 17, 18. Galaxias anomalus,9, Mountain beech, 11. Arthropods, terrestrial, I 2. Galaxías dívergens, 15. Myriophyllum, ll,12, Ashley River, 11. Galaxias maculatus, 9, 75. Australia, 9. Galaxiidae,9. Galaxiids, 9, 12. Nematomorpha, 12. Geological deposits, 11. Neochanna,9. Bald Hills Stream, 10, ll. Glentui River, 10, 1l-12, 13, 18, 19, Nothofagus solandr i, 1 7. Bass, large-mouth, 13. 20, 21,22,23,28. Beech, mountain, 11. Go bíomorp hus br ev icep s, 12. Biological characteristics of study Gomphonema,12. Oligochaeta, 12. area,72. Gonads, 14,18,27. Otoliths, 15-18, 20, 21. Black shags, 12. Gordius,72. Bluegill, 13. Green sunfish, 28. pH Breeding biology, 9, 13. Growing season, 27, 28. of water, 12. Breeding season, 15, 18,26,27. Growth, 9, 10, 13, 15, 18,20,2l-25, Phalacrocorax carbo, 12. Brown trout, 12. 2819. Photoperiod, 28. Buller, Upper, River System, 9. Plecoptera, 12. Bully upland, 12. Population density, 28. Hemiptera, 12. Post-spawning stage, 14. Home range and movement, 13. Potamopyr gu s ant ip odar um, 72. Calcifrcation, rate of, 18, Pre-spawning, 28. Cass River, 22. Prey, abundance of, 28. Ictalurus punctatus, 13. Catfish, channel, 13. Primary rings in otolith, 15. c alo s to mus macr oche il u s, 26. Insects, larval,72. Channel catfish, 13. Interrelations of G. vulgaris with other frsh, 13. Climate, 11. Rainbow trout, 13. Invertebrates, 12,28. Coleoptera, 12. Rainfall, 11. Condition, 9, 10, 13, 18,26-27. Reproduction, 14, 18. Conductivity, 12, 13. Resting stage, 14. Large-mouth bass, 13. Rings in otolith, 15. Large-scale sucker, 26. Ripening stage, 14. Diatoms, 12. Larval insects, 12. Diel activity pattern, 18. Length, annual growth in, 21. Diptera, 12. Length-frequency analysis, 15, 20. Salmo gaírdneri,73. Length-frequency distributions, 19, Salmo trutta, 12. 20,22,28. Salmonids, 12. Early-ripening stage, 14. Length, measurement of, 14. Sampling pr ogramme, 13, 22, Eels, 12. Length, seasonal growth in, 22-23. Seasonal growth cycle, 28. Electric fishing, 13. Length-somatic weight relationship, Seasonal growth in length,22. Eleotrids, 12. 27. Seasonal growth in weight, 23-25. Embryological development, 9. Length-weight relationship and condition,26-27. Secondary rings in otolith, 15. Ephemeroptera, 12. Lepomis cyanellus,28. Sex determination, 14. Lepomis macrochirus, 13. Shags, black, 12. Fat deposits, 13, 18, 27. Long-finned eel, 12. Sho¡t-finned eel, 12. Feeding behaviour, 13, 28. Lumbricids, terrestrial, 12. Somatic tissue, 27.

'Bulletins of the Fisheries Research Division

No. l. The New Zealand Cetacea. By D. E. Gaskin. Published in 1968. 92 pp. 2. Galaxias maculatus (Jenyns), the New Zealand whitebait. By R. M. McDowall. Published in 1968. 84 pp. 3. Phytoplankton productivity in Tomahawk Lagoon, Lake Waipori, and Lake lldahins¡¿¡gi. By S. F. Mitchell. Published in 1971.87 pp. 4. Some aspects of the bionomics of fish in a brown trout nursery stream. By C. L. Hopkins. Published in 1970. 38 pp. 5. Reproduction, early life history, and age-growth relationships of the New Zealatd pilchard, Sardinops neopilchardus (Steindachner). By Alan N. Baker. Published in 1972. 64 pp. 6. The biology of the New Zealand tarakihi, Cheilodactylus mauopterus @loch and Schneider). By L. J. Tong and C. M. Voo¡en. Published in 1972. 60 pp. 7. Aq- analysis of the statistics on the fishery for tarakihi, Cheilodactylus macropterus (Bloch and Schneider), in New Zealand waters from 1936 to 1969, with notes on the trawl fishery in general. By C. M. Vooren. Published in 1974. 44pp. 8. Lakes Rotorua and Rotoiti, North Island, New Zealand: Their trophic status and studies for a nutrient budget. By G. R. Fish. Published in 1975. '12 pp. 9. A survey of the tarakihi, Cheilodactylus mauopterus (Bloch and Schneider), in the East Cape area, New Zealand426-30 March 1971. By C. M. Vooren and L. J. Tong. Published in 1973.28 pp. 10. The relation between primary productivity, nutrients, and the trout environment in some New Zealand lakes. By A. M. R. Burnet and Denise A. Wallace. Published Á 1973.28 pp. 11. Biology and distribution of the toheroa, Paphies (Mesodesma) ventricosa (Gray). By P. Redfearn. Published in 1974. 51 pp. 12. Studies on age and growth, reproduction, and population dynamics of red gurnard, Chelidonichthys kumu (Lesson and Garnot), in the Hauraki Gulf, New Zealand. By R. D. Elder. Published in 19'16.77 pp. 13. A study on age2 growth, and population structure of the snapper, Chrysophrys auratus (Forster), in the Hauraki Gulf, New Zealand. By L. J. Paul. Published in 1976. 62 pp- 14. Variations in growth, mortality, and population density of snapper, Chrysophrys ouratus (Forster), in the Hauraki Gulf, New Zealand. By C. M. Voóren and R. F. Coombs. Published in 1977.32 pp. 15. The commercial fishery forsnapper, Chrysophrys auratus (Forster), in the Auck- land region, New Zealand, from 1900 to 1971. By L. J. Paul. Published in 1977. 84 pp. 16. A bibliography of the literature about New Zealand's marine and freshwater commercial frsheries, 184V1975. By L. J. Paul (in press). 17. 4C9, growth,and condition of the common river galaxias, Galaxias vulgaris Stokell, in a Canterbury river, New Zealand. By P. L. Cadwallader. Publishèd in 1978. 35 pp. 18. Tag_ging experiments on the sand floundet, Rhombosolea plebeia (Richardson), in Canterbury, New Znaland,1964 to 1966. By J. A. Colman (in press). t9. fþç üfe history of Neochanna apoda Günther (Pisces: Galaxiidae). By G. A. Eldon (in press). 20. Th,e ecolo-gy 9f whitelait migrations (Galaxiidae: Galaxias spp.) By R. M. McDowall and G. A. Eldon (in press).